ENCYCLOPEDIA OF CONSCIOUSNESS
ENCYCLOPEDIA OF CONSCIOUSNESS EDITOR-IN-CHIEF
WILLIAM P. BANKS Pomona College, Claremont, CA, USA
Amsterdam • Boston • Heidelberg • London • New York • Oxford Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo Academic Press is an imprint of Elsevier
Academic Press is an imprint of Elsevier The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK 525 B Street, Suite 1900, San Diego, CA 92101-4495, USA First edition 2009 Copyright ã 2009 Elsevier Inc. All rights reserved FOLK THEORIES OF CONSCIOUSNESS Copyright ã 2009 B F Malle VISUAL IMAGERY AND CONSCIOUSNESS Copyright ã 2009 N J T Thomas No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+44) (0) 1865 843830; fax (+44) (0) 1865 853333; email:
[email protected]. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material Notice No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein, Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Catalog Number: 2009920709 ISBN (print): 978-0-12-373864-6 ISBN (online): 978-0-12-373873-8 For information on all Elsevier publications visit our website at books.elsevier.com
07
08
09
10
11
10
9
8
7
6
5
4
3
2
1
FOREWORD
Consciousness is a scientific problem that is unlike any other. Our own consciousness, as Descartes noted, is the most indubitable feature of our existence. It is the most precious one, as well: consciousness is life itself, and for most people having their bodies kept alive in a vegetative state is no better than dying. The major religions are defined by their theories of consciousness: whether a person’s essence consists of his consciousness (his soul) or his body; how that consciousness ultimately fares as the result of its choices in life (whether it goes to a special place, or melds into a global mind); and whether the world contains forms of pure, disembodied consciousness in the form of gods, demons, angels, and spirits. And the conviction that other people can suffer and flourish as each of us does is the essence of empathy and the foundation of morality. This encyclopedia is thus a celebration of a momentous recent development in the history of science. Consciousness has not only become a respectable scientific topic, but it is one that has seen tremendous progress. And it enjoys a surprising degree of consensus for such an elusive subject. My hunch is that a majority of consciousness researchers would assent to the following propositions. Human consciousness is entirely a product of the physiological activity of the brain. Consciousness does not depend on language. Nor is it the same thing as self-conscious rumination. The vast majority of information-processing taking place in the brain is unconscious. The parts that are conscious have no privileged access to the unconscious parts, and are more likely to confabulate about how they work than to report those workings accurately. Consciousness is not a unitary process that is located in a single part of the brain or that has complete control of behavior. The neural correlates of consciousness are defined both over space (in terms of connections to the frontal lobes and other cortical areas) and over time (in terms of oscillatory patterns in cortico-thalamic loops). Consciousness is intimately tied to– perhaps the same thing as—attention, imagery, and working memory. The cognitive representations that can become conscious tend to be found at intermediate levels of processing hierarchies, corresponding neither to information in the sensorium nor to the most abstract categories and goals. Consciousness plays a critical role binding together distributed activity over the brain that corresponds to a single entity in the world, and its main computational function is to serve as a ‘‘blackboard’’ on which the output of one cognitive process can be made available to many others. If I am correct that this is a rough emerging consensus among consciousness researchers, it would represent an extraordinary degree of progress. The modern science of consciousness is barely two decades old. It has progressed in the teeth of objections from behaviorist philosophers and psychologists that the very idea of consciousness is a category mistake, and has had to distance itself from a horde of obfuscators, mystics, kibitzers, and kooks. For all that, it is possible that our knowledge about consciousness will never be complete. Consciousness researchers do not agree that we will inevitably have an intellectually satisfying explanation of the ‘‘hard problem’’ of consciousness. This term, from the philosopher David Chalmers, is, of course, something of an in-joke. There is nothing easy about the so-called ‘‘Easy Problem’’ of distinguishing conscious from unconscious mental computation, identifying its correlates in the brain, and explaining why it evolved. v
vi
Foreword
But when it comes to the Hard Problem—why the conscious portion of brain activity consists of firstperson, subjective experience rather than pure information processing of the kind that takes place in thermostats and calculators–no one knows what a solution might look like, or even whether it is a genuine problem in the first place. I am among those who suspect that the hard problem is a genuine problem but one that we will never solve. It is an idea that goes back at least to Hume, and has also been embraced by the biologist Gunther Stent, the linguist Noam Chomsky, and most extensively, the philosopher Colin McGinn. The theory holds that our vertigo when pondering the Hard Problem is itself a quirk of our brains. The brain is a product of evolution, and just as animal brains have their limitations, we have ours. Our brains can’t hold a hundred numbers in memory, can’t visualize 7-dimensional space, and perhaps can’t intuitively grasp why neural information-processing observed from the outside should give rise to subjective experience on the inside. I have found that most scientists hate this theory (which McGinn calls ‘‘cognitive closure’’), acknowledging that that the Hard Problem remains unsolved for now, but assuming that it will eventually succumb to research that chips away at the Easy Problem. I have also found that most scientists don’t understand the theory. I have been told that the cognitive closure theory tries to put the study of consciousness off-limits to science, a bizarre misinterpretation (the theory encourages the study of consciousness, while offering an empirical prediction of how it will turn out). I have been told that the theory is unfalsifiable. In fact the theory could be demolished if a genius came up with a flabbergasting new idea which made it all clear to us (in contrast, it is the conviction that some day in the future, we don’t know when, we don’t how, the problem will succumb, that is unfalsifiable). I have read that the theory is an endorsement of mystery over science, whereas in fact it seeks to explain of the phenomenon of mystery—what it is about the human brain that makes it susceptible to feelings of mystery. Moreover, it pinpoints the aspect of cognitive processing that is responsible for our mystification: the fact that our explanations exploit the combinatorial powers of the brain, which explains complex phenomena in terms of rule-governed interactions among simpler elements, and thus is frustrated when it runs across problems that have a holistic flavor, such as the hard problem of consciousness. And I have been told that research that characterizes consciousness (for example, by crossmodal similarity judgments) solves the hard problem, whereas such research just addresses the easy problem, documenting patterns of representation and information processing. Whether or not my prediction withstands attempts at empirical falsification, it would matter little to the actual science of consciousness. The infuriating thing about the hard problem is that nothing hangs on it scientifically. As soon as you start to study it, you are studying one of its objective, third-person manifestations, which as far as we know can always be explained as information processing in a physical system. If we ever could trace all the neurocomputational steps from perception through reasoning and emotion to behavior, the only thing left missing by the lack of a theory of the hard problem of consciousness would be an understanding of the hard problem of consciousness itself. That is why, for all the unsatisfied intellectual curiosity we might have as to why subjective experience exists, the absence of an explanation takes nothing away from the exhilarating ideas and findings presented in this encyclopedia. Steven Pinker Harvard University
FOREWORD BIOGRAPHY
Steven Pinker is Harvard College Professor and Johnstone Family Professor of Psychology at Harvard University. He has also taught at Stanford and MIT. His research on visual cognition and the psychology of language has won prizes from the National Academy of Sciences, the Royal Institution of Great Britain, and the American Psychological Association. He has also received six honorary doctorates, several teaching awards, and numerous prizes for his books The Language Instinct, How the Mind Works, and The Blank Slate. He is currently Honorary President of the Canadian Psychological Association and Chair on the Usage Panel of the American Heritage Dictionary, and writes frequently for The New Republic, The New York Times, and other publications. He has been named Humanist of the Year, and is listed in Foreign Policy and Prospect magazine’s ‘‘The World’s Top 100 Public Intellectuals’’ and in Time magazine’s ‘‘The 100 Most Influential People in the World Today.’’ His latest book is The Stuff of Thought: Language as a Window into Human Nature.
vii
PREFACE
The study of consciousness, like psychology, has a long past and a short history, but the history part is shorter and the past much longer than psychology’s. Topics now included in the study of consciousness have been matters of fundamental existential human concern since prehistoric times. The earliest ceremonial burial sites, which suggest a belief in a sentient essence independent of the body, may be 70,000 years old. Only very recently has the approach to these issues been substantially different from those of the previous millennia. The difference between now and then is an empirical approach to consciousness and mental activity. This difference comes from the change in worldview that led to what we now call science and was a long time coming to the study of the mind. The application of an empirical criterion for evidence brought consciousness into the domain of ‘‘normal science.’’ Not everything from the older approach gets into normal science, however, only the few parts that follow the rules. The rules include no supernatural explanations and no untestable ones, of which the supernatural is a subset. The anti-supernatural rule constrains us to physical explanations, which are for the most part biological. The result is that some questions unthinkable in the prescientific tradition, such as neural correlates of consciousness, become important. Another previously unthinkable idea is that thought is a product of neural activity. (Imagine that! Reason founded on neurons?) Still another unthinkable idea is that some means is needed to maintain the coherence and continuity of experience. If the self is a spiritual, supernatural entity, continuity is in its nature. If it is a product of brain activity we need a neurophysiological explanation of how continuity and unity can come from the massively parallel operation of the brain. Articles by Schmidt, Vallacher, Bermudez, Olson, and Gallagher (Perception: The Binding Problem and the Coherence of Perception; Self: The Unity of Self, Self-Consistency; Self: Body Awareness and Self-Awareness; Self: Personal Identity and Consciousness of Time and the Time of Consciousness respectively) cover theories of the self. These are just a few of the previously unthinkable explanations needed in a science of consciousness. One of the developments that puts the study of consciousness on a solid empirical footing is the array of neural imaging tools that let us see relations between awareness and brain activity. The images themselves are exciting because they give a glimpse of what the brain is doing when we are consciously seeing, remembering, thinking our own thoughts and so on. We can see changes in brain activity as the focus of attention switches from one image to another, for example. In addition to the scientific value of such findings, there is what could be called rhetorical value: the techniques drive home the point that what is going on in the mind is going on in the brain. Articles by Haynes, Kouider, and Smythes, Goodale, and Sterzer (The Neural Basis of Perceptual Awareness; Neurobiological Theories of Consciousness; The Neurochemistry of Consciousness; Perception, Action, and Consciousness and Bistable Perception and Consciousness) and others in this collection report some of the scientific progress made through these tools. Several generalizations can be drawn from these contributions. One is that perception, and presumably other conscious processes, has a division of labor among many special-purpose analyzers. Some of these operate on sensory input and others on very high-level analysis such as face recognition. Their results are not by themselves conscious but if coordinated with other special-purpose analyzers can result in a conscious perception. The evidence that ix
x
Preface
sophisticated representations can be created unconsciously would support theories that postulate an active unconscious. Control and coordination is one of the many processes subsumed under the rubric of attention. Change blindness and inattentional blindness illustrate the central importance of attention to consciousness. Change blindness is found when a scene or part of a scene is changed while the observer is not attending. The changes in the scene can literally be not seen at all if not attended. Rensink (Attention: Change Blindness and Inattentional Blindness) notes that such a gap in perception would have been classed as a momentary aberration before it had theoretical significance. Change blindness indicates that at a phenomenal level our impression of seeing the whole scene before us is not correct. The entirety of the scene may be available, but only what is attended will be perceived. Inattentional blindness is similar to change blindness, but it is not noticing something that is present but unattended in the visual field. These ‘‘blindnesses’’ both suggest that attention is a gateway to awareness. Unattended objects may be analyzed to some depth outside of attention, as when emotionally laden words are seen in intentional blindness experiments. These words had to have been processed to the level of meaning without attention, or else they would not have been recognized. This finding is one of many that show analysis progressing quite far without attention or awareness. Breitmeyer (Perception: Unconscious Influences on Perceptual Interpretation) discusses in depth the relations between conscious and unconscious processes in perception. Snodgrass (Perception: Subliminal and Implicit) analyzes effects of stimuli that cannot be consciously represented, and covers many of the methodological problems in this area. Processing of unattended material brings us to the larger question of what kind of processing can take place unconsciously. Consciousness is the tip of the iceberg of mental activity, the rest being unconscious, but the nature of the unconscious activity is a question. Some theorists assume that unconscious thinking is essentially like conscious thinking, just not conscious. Others, like Freud, theorized that unconscious mental activity goes on but it has different characteristics than conscious thinking has. It’s an associative ‘‘primary process’’ rather than the rational thought of the conscious, waking mind. The articles by Kihlstrom (Unconscious Cognition) and Macmillan (Psychodynamic Theories of the Unconscious) discuss the active unconscious. Still others do not require an active unconscious but only action patterns or habits that can be automatically called into action. Approaches like this are covered in several articles, including those by Aarts (Habit, Action, and Consciousness) and Dijksterhuis (Unconscious Goals and Motivation). Schneider (Automaticity and Consciousness) contrasts automatic processing, which is unconscious, fast, and virtually effortless with controlled processing, which has properties that might classify it as conscious. Any or all of these versions of unconscious activity may take place in some situations. Implicit cognition does much of our daily mental work, just as automatic processing spares us the effort of controlling skilled actions. Cleeremans (Implicit Learning and Implicit Memory) describes a wide range of studies that show how implicit memory can develop and influence performance, usually for the better, without conscious recall of the learning event or events. Bar-Anan and Nosek (Implicit Social Cognition) describe the Implicit Attitude Test (IAT) for measuring implicit attitudes that can influence opinions and social behavior but may be in contrast to admitted, explicit attitudes. Vallacher (Self: The Unity of Self, Self-Consistency) describes how the IAT is used to assess implicit self-esteem. Self-esteem measured by overt techniques such as questionnaires may differ from implicit measures. The result can be dysfunctional uncertainty about selfconcept and conflicting and erratic behavior in different situations. As Steven Pinker discusses in his foreword, scientific models of functions that have a conscious consequence run into what the philosopher David Chalmers termed the ‘‘hard’’ problem. For example, a complete model of color vision may predict such things as the outcome of color mixtures very well, but where in the explanation is the experience of the resulting color? We would like to have it explain not only the result, that a paper reflecting a certain spectral distribution is called ‘‘red,’’ but also why it has the qualitative look it has, the ‘‘redness’’ of red. There is a disjoint here. Scientific models have not explained subjective experience, and there seems no way to do it with the concepts they employ. Seager (History of Philosophical Theories of Consciousness) and Rowlands (The MindBody Problem) refer to the
Preface
xi
problem of explaining experience with a scientific model as intractable. We seem to be attempting to connect two things, the wavelength distribution, which is a physical quantity, and the experience, which is known only to the observer. The models don’t have a meaning for the expression, ‘‘It looks red to me.’’ It is though we have returned to the dualistic universe of material and spiritual (supernatural) substances. One can’t hook a neuron to a feeling. They are different kinds of things. The scientific enterprise goes about its business of explaining psychological phenomena without a worry about the intractability of the connection between explanation and experience. It is a problem seemingly at the center of the discipline that stands like an elephant in the room, yet science goes on as if it were not an issue. Perhaps it is enough to work only on the ‘‘easy’’ problems and ignore that old hard one. However, there are reasons to ask for some account of the experiential side of the process. The first is mentioned by Rowlands. It is unsatisfactory to leave this account out of the story. The perception of red is not explained to our satisfaction until the perception itself is explained – wasn’t that the point, after all? Another reason to have an answer about subjective states comes from the other hard problem, volition. How does the decision to act result in action? Again we seem to have two different things: the decision to act, which is an idea, and the action, some sort of motion of a group of muscles. How can an idea move a muscle? Here we do not have a passive sensation of red, we have an idea that seems actually to do something. It is as though the action is an effect of the idea and clear evidence that a connection is made between the idea and the material world. Articles by Pockett (Brain Basis of Voluntary Control), Hommel (Conscious and Unconscious Control of Spatial Action), Goodale (Perception, Action, and Consciousness), Jeannerod (Neuroscience of Volition and Action), and Custers (Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness) constitute together a reference volume on volition and action. These accounts show that action is much more complex than the simple formula ‘‘idea causes action.’’ It is possible that our clear impression of a causal relation comes from an overly simplistic description. Pinker leans toward one resolution of the hard problem in his forward. This is that there is a resolution, but our simian minds, evolved for other things, are simply not capable of finding it, or of understanding it if it were presented to us. I would strive for the negative capability that the poet John Keats admired. That is the ability to resist the urge to choose one option or another and remain in uncertainty. I should add that I have a suspicion that the problem may be formulated wrongly. The history of science shows many cases in which when a problem has gone unsolved for many years, and a number of very skilled problem-solvers have failed with it over decades or centuries, the problem has been improperly formulated, often because a wrong theory, implicit or explicit, has been applied. Our language betrays an implicit dualism in many locutions: a thought becomes ‘‘available’’ to consciousness (as if consciousness were outside the system); a certain neural process ‘‘reports’’ to consciousness; attention brings something ‘‘into’’ consciousness. These metaphors can constrain in subtle ways. In the dictionary ‘‘mental’’ and ‘‘physical’’ are given as antonyms, and they lurk in my personal lexicon as such. Again, language tips us off to the implicit metaphor. I don’t think I’m alone in having this problem. Francis Crick called the idea that consciousness is a product of brain events ‘‘the astonishing hypothesis.’’ It is astonishing because we accept it as a scientific matter but some small part of us still finds it an affront to common sense. Clearly, the old way of thinking is very deep and still may haunt us, so to speak. Perhaps a generation that grows up with results of the sort reported in this handbook will have an informed intuition that will not find the hypothesis astounding. Perhaps a member of that generation will come up with a formulation that will lead to a scientific answer. As a final note, consciousness research has developed subdisciplines remarkably quickly for such a young field. These include analysis of attention, memory, the self, unconscious cognition, emotion, social perception, volition, psychopathology, and more. Just scan the table of contents of this encyclopedia to see the richness and variety of this work. It is as though researchers in many areas suddenly discovered that awareness was an important aspect of the topic they studied. There is more than just growth of interest in consciousness here. Awareness has become a central question in many areas. In order to understand perception, emotion, memory, volition, and many other
xii
Preface
human functions we need to consider the role of awareness. The result is a transformation that ranks with the Copernican revolutions that have come at critical times in the history of science. This encyclopedia depended on the good work of a number of people. I thank Nikki Levy, the Publisher, for conceiving of this project and getting it through the inevitable rough spots gracefully and effectively. Maria Turnock, Development Editor, for Major Reference Works, and Edward Taylor, as Project Manager, deserve thanks for the daily operation of this project and dealing with hundreds of academics. The Editorial Assistants that have tirelessly worked as support on this project are Victoria Sayce, Caroline Phipps and Milo Perkins. Finally, the editors, Bernard Baars, Mahzarin R. Banaji, Bruce Bridgeman, Shaun Gallagher, Geraint Rees, Jonathan Schooler, and Daniel Wegner, are responsible for assembling a stellar group of authors and obtaining the superb set of articles you see here. William P. Banks
EDITOR-IN-CHIEF
William P. Banks received his BA, cum laude, from St. John’s College, Annapolis, Maryland in 1964, and his MA in 1966 and PhD in 1968 from Johns Hopkins University. He took a post-doctoral fellowship at the Institute of Human Learning at the University of California, Berkeley, for the academic year 1968–1969. He has been on the faculty at Pomona College since 1969, and was promoted to Professor in 1979, and in 2007 became the Edwin F. and Margaret Haynes Professor of Psychology. He was head of the psychology department 1974–1990 and again 2001–2004. His courses have included perception and cognition, consciousness, the psychology of evil, and a currently ongoing high-impact aerobics class that students have taken for PE credit every semester since 1985. He has twice received the Pomona college award for distinguished teaching. Among some eighty publications of his are four co-authored books. His publications cover a variety of topics, including cognition of magnitude, Gestalt factors in perception, the psychophysics of alcohol intoxication, signal-detection theory and memory, the recovered memory debate, and of course consciousness. Support for his research has come from the National Institutes of Mental Health, the National Science Foundation, the Irvine Foundation, the Fetzer Foundation, and the National Institute on Alcohol Abuse and Alcoholism. Between 1986 and 1992 he was a Project Director and co-PI under two Sloan New Liberal Arts (NLA) grants. The motivation of the NLA project was to counter what appeared to be a growing neo-Luddism among liberal arts students. His main contribution to the project was to develop a problem-solving program that gave students the opportunity to tackle real-world problems. Teams of students spent an academic year working on paid contracts with government and business. Thirty-five contracts were completed with clients as diverse as city governments, McDonnell Douglas, and NASA. He has been Psychology Editor for the International Encyclopedia of Science and Technology since 1988. From 1984 to 1987 he was Associate Editor for the Journal of Experimental Psychology: Human Perception & Performance. He has been a member of the panel on Memory and Cognitive Processes,
xiii
xiv
Editor-in-Chief
National Science Foundation; the Air Force Office of Scientific Research, Cognitive Science Panel; the National Science Foundation Graduate Fellowship Selection Panel; and the American Psychological Foundation Teaching Award Committee. He is a Fellow of the Association for Psychological Science. In 1990 he and Bernard Baars founded the journal Consciousness and Cognition. The journal has grown from 377 pages in the first volume, published in 1992, to 1400 pages in the most recent, volume 17. He began serving as Editor-in-Chief in 2005.
EDITORIAL BOARD
Bernard J. Baars is former Senior Research Fellow in Theoretical Neurobiology at the Neurosciences Institute in San Diego (www.nsi.edu), Adjunct Professor at the Institute for Intelligent Systems at the University of Memphis, and Distinguished Consulting Faculty member at Saybrook Graduate School in San Francisco. His PhD is in Cognitive Psychology from UCLA. He is interested in human language, the brain basis of consciousness, volition, and a variety of related topics including the history of scientific studies of consciousness and neuroethics. Baars pioneered a cognitive theory of consciousness called Global Workspace Theory, which is widely cited in philosophical and scientific sources. Together with William P. Banks, Baars has edited the journal Consciousness & Cognition for more than fifteen years. (From Academic Press/ Elsevier). He has written an introductory text for cognitive neuroscience, called Cognition, Brain & Consciousness: An Introduction to Cognitive Neuroscience (Baars & Gage, Eds. San Diego, Calif.: Elsevier/ Academic Press, 2007). Baars was founding President of the Association for the Scientific Study of Consciousness and has given presentations internationally (see www.nsi.edu/users/baars and http:// bernardbaars.pbwiki.com, and http://en.wikipedia.org/wiki/Bernard_Baars).
xv
xvi
Editorial Board
Mahzarin R. Banaji was born and raised in India, in the town of Secunderabad, where she attended St. Ann’s High School. Her B.A. is from Nizam College and her M.A. in Psychology from Osmania University in Hyderabad . She received her Ph.D. from Ohio State University (1986), was a postdoctoral fellow at the University of Washington, and taught at Yale University from 1986 until 2001 where she was Reuben Post Halleck Professor of Psychology. In 2002 she moved to Harvard University as Richard Clarke Cabot Professor of Social Ethics in the Department of Psychology and Carol K. Pforzheimer Professor at the Radcliffe Institute for Advanced Study. Banaji is a Fellow of the American Association for the Advancement of Science, the American Psychological Association (Divisions 1, 3, 8 and 9), and the American Psychological Society. She served as Secretary of the APS, on the Board of Scientific Affairs of the APA, and on the Executive Committee of the Society of Experimental Social Psychology. She was elected fellow of the Society for Experimental Psychologists in 2005. Banaji has served as Associate Editor of Psychological Review and of the Journal of Experimental Social Psychology and is currently Co-Editor of Essays in Social Psychology. She serves on the editorial board of several journals, among them Psychological Science, Psychological Review, Journal of Personality and Social Psychology, and The DuBois Review. Her research has been funded by the National Science Foundation, the National Institute of Mental Health, and the Third Millennium Foundation. Banaji was Director of Undergraduate Studies at Yale for several years, chaired APS’s Task force on Dissemination of Psychological Science, and served on APA’s Committee on the Conduct of Internet Research. Among her awards, she has received Yale’s Lex Hixon Prize for Teaching Excellence, a James McKeen Cattell Fund Award, and fellowships from the Guggenheim Foundation and the Rockefeller Foundation. In 2000, her work with R. Bhaskar received the Gordon Allport Prize for Intergroup Relations. With Anthony Greenwald and Brian Nosek, she maintains an educational website that has accumulated over 3 million completed tasks measuring automatic attitudes and beliefs involving self, other individuals, and social groups. It can be reached at www.implicit.harvard.edu, and details of the research may be found at www.people.fas.harvard.edu/~banaji. Banaji studies human thinking and feeling as it unfolds in social context. Her focus is primarily on mental systems that operate in implicit or unconscious mode. In particular, she is interested in the unconscious nature of assessments of self and other humans that reflect feelings and knowledge (often unintended) about their social group membership (e.g., age, race/ethnicity, gender, class). From such study of attitudes and beliefs of adults and children, she asks about the social consequences of unintended thought and feeling. Her work relies on cognitive/affective behavioral measures and neuroimaging (fMRI) with which she explores the implications of her work for theories of individual responsibility and social justice.
Editorial Board
xvii
Bruce Bridgeman is a professor of psychology and psychobiology a the University of California, Santa Cruz. He holds a BA in psychology from Cornell University and a PhD in physiological psychology from Stanford University. He was a post-doctoral fellow at the Free University of Berlin with a Humboldt Foundation fellowship, and at the UC Berkeley School of Optometry with a NIH fellowship. His research has been supported by the National Institutes of Health, the National Science Foundation, the Air Force Office of Scientific Research, and NASA. For six years he was a Guest Professor at the Max-Planck Institute for Psychological Research, Munich, and has served on the advisory committees of two Max-Planck Institutes. He has been an invited faculty member at summer schools in Germany and Holland, and held research appointments at the Universities of Bielefeld, Germany, and Padua, Italy. His research concentrates on spatial aspects of vision - how information is coded and processed in the brain during activities such as pattern recognition and visually guided action. Professor Bridgeman is the author of The Biology of Behavior and Mind (Wiley) and Psychology and Evolution: The Origins of Mind (Sage), and more than 120 scientific journal publications, as well as book chapters and international conference contributions. His most recent academic honor is the Chemers award for excellence in research.
Shaun Gallagher is Professor and Chair of the Philosophy Department at the University of Central Florida and Research Professor of Philosophy and Cognitive Science at the University of Hertfordshire. He received his Ph.D. in philosophy from Bryn Mawr College, an MA in economics from the State University of New York, and an MA in philosophy from Villanova University. He has been an invited Visiting Professor at the Ecole Normale Supe´riure, Lyon (2007), occasional invited Visiting Professor at the University of Copenhagen (2004–06), and Visiting Scientist at the Medical Research Council’s Cognition and Brain
xviii
Editorial Board
Sciences Unit at Cambridge University (1994). His research interests include phenomenology and philosophy of mind, cognitive sciences, hermeneutics, theories of the self and personal identity. His books include How the Body Shapes the Mind (Oxford University Press, 2005); The Inordinance of Time (Northwestern, 1998); Hermeneutics and Education (SUNY, 1992) and, co-authored with Dan Zahavi, The Phenomenological Mind: Introduction to Philosophy of Mind and the Cognitive Sciences (Routledge, 2007). He is co-editor of the interdisciplinary journal Phenomenology and the Cognitive Sciences, and a recent co-editor of Does Consciousness Cause Behavior? An Investigation of the Nature of Volition (MIT Press, 2006).
Geraint Rees is a Wellcome Senior Clinical Fellow and Professor of Cognitive Neuroscience at University College London. Trained as a neurologist and experimental psychologist at University College London and the California Institute of Technology, his group uses functional neuroimaging to study the neural basis of perceptual awareness. He is an Associate Editor of Brain and has published extensively on the neural basis of human consciousness.
Jonathan W. Schooler Professor of psychology at the University of California at Santa Barbara, Jonathan W. Schooler pursues research on consciousness, memory, the relationship between language and thought, problem-solving, and decision-making. A cum laude graduate of New York’s Hamilton college, Dr. Schooler earned a Ph.D. in psychology at the Univeresity of Washington in 1987. He joined the faculty of the University of Pittsburgh as assistant professor, and in 2004 accepted and held the position of full professor and Canada Research Chair in Social Cognitive Science until 2007 when he accepted his current position. A fellow of the Association for Psychological Science, Dr. Schooler has been the recipient of three Akumal Scholar Awards from the Positive Psychology Network, an Osher Fellowship given by the Exploratorium Science Museum, and a Lilly Foundation Teaching Fellowship. His work has been
Editorial Board
xix
supported, among others, by the National Institute of Mental Health, the Office of Educational Research and Canada’s Social Sciences and Humanities Research Council. He is currently on the editorial boards of Consciousness and Cognition and Social Cognitive and Affective Neuroscience. Dr. Schooler is the author or coauthor of more than one hundred papers published in scientific journals and the editor (with J.C. Cohen) of Scientific Approaches to Consciousness.
Daniel M. Wegner is Professor of Psychology at Harvard University. A Ph.D. of Michigan State University (1974), he has held professorships at Trinity University in Texas and at the University of Virginia. Wegner has published over a hundred articles on the role of thought in self-control and in social life, pioneering the study of thought suppression (why we have trouble keeping unwanted thoughts out of mind), transactive memory (how we remember things cooperatively with others), and apparent mental causation (what gives us the sense that we are consciously causing our actions). Among Wegner’s books are The Illusion of Conscious Will (2002) and White Bears and Other Unwanted Thoughts (1989). His research has been funded by the National Institute of Mental Health and by the National Science Foundation, and he has been a Fellow of the Center for Advanced Study in the Behavioral Sciences. A Fellow of the American Psychological Association and of the American Association for the Advancement of Science, he has served as associate editor of Psychological Review, and on the Board of Reviewing Editors for Science. His recent research focuses on how people perceive the minds of others.
CONTRIBUTORS
H Aarts Utrecht University, Utrecht, The Netherlands
M Campbell National University, La Jolla, CA, USA
M T Alkire University of California, Irvine, CA, USA
J M Chin University of British Columbia, Vancouver, BC, Canada
M C Anderson University of Oregon, Eugene, OR, USA L Y Atlas Columbia University, New York, NY, USA B J Baars The Neurosciences Institute, San Diego, CA, USA Y Bar-Anan University of Virginia, Charlottesville, VA, USA A J Barnier Macquarie University, Sydney, NSW, Australia J J Van Bavel The Ohio State University, Columbus, OH, USA J L Bermu´dez Washington University in St. Louis, St. Louis, MO, USA J Blascovich University of California, Santa Barbara, CA, USA
A Cleeremans Universite´ Libre de Bruxelles, Bruxelles, Belgium M A Cohen Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA V Cologan University of Lie`ge, Lie`ge, Belgium M A Conway University of Leeds, Leeds, England N Cowan University of Missouri, Columbia, MO, USA W A Cunningham The Ohio State University, Columbus, OH, USA R Custers Utrecht University, Utrecht, the Netherlands
M Boly University of Lie`ge, Lie`ge, Belgium
T T Dang-vu University of Lie`ge, Lie`ge, Belgium; Centre Hospitalier Universitaire de Lie`ge Sart Tilman, Lie`ge, Belgium
K C A Bongers Radboud University Nijmegen, Nijmegen, the Netherlands
A Demertzi University of Lie`ge, Lie`ge, Belgium
B G Breitmeyer University of Houston, Houston, TX, USA
A Dietrich American University of Beirut, Beirut, Lebanon
xxi
xxii
Contributors
A Dijksterhuis Radboud University Nijmegen, Nijmegen, the Netherlands
T S Horowitz Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA
D B Edelman The Neurosciences Institute, San Diego, CA, USA
P D L Howe Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA
J T Enns University of British Columbia, Vancouver, BC, Canada K K Evans Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA M-E Faymonville Centre Hospitalier Universitaire Sart Tilman, Lie`ge, Belgium J I Fleck The Richard Stockton College of New Jersey, Pomona, NJ, USA S Gallagher University of Central Florida, Orlando, FL, USA; University of Hertfordshire, Hatfield, UK
D D Hutto University of Hertfordshire, Hatfield, UK M Jeannerod Institut des Sciences Cognitives, Lyon, France N Kalaida University of Michigan Medical Center, Ann Arbor, MI, USA J F Kihlstrom University of California, Berkeley, CA, USA S Kouider De´partement d’Etudes Cognitives, CNRS/ EHESS/ENS-DEC, Paris, France J Kounios Drexel University, Philadelphia, PA, USA
D Gamez University of Essex, Colchester, UK
S Laureys University of Liege, Lie`ge, Belgium
V de Gardelle De´partement d’Etudes Cognitives, CNRS/ EHESS/ENS-DEC, Paris, France
B J Levy University of Oregon, Eugene, OR, USA
M A Goodale University of Western Ontario, London, ON, Canada
M Macmillan University of Melbourne, Melbourne, VIC, Australia
P Gordon University of Columbia, New York, NY, USA
B F Malle University of Oregon, Eugene, OR, USA
J-D Haynes Charite´ – Universita¨tsmedizin Berlin, Berlin, Germany
P Maquet University of Lie`ge, Lie`ge, Belgium; Centre Hospitalier Universitaire de Lie`ge Sart Tilman, Lie`ge, Belgium
W Hirstein Elmhurst College, Elmhurst, IL, USA O Holland University of Essex, Colchester, UK B Hommel Leiden University, Leiden, the Netherlands
A R Mele Florida State University, Tallahassee, FL, USA R Menary The University of Wollongong, Wollongong, NSW, Australia
Contributors
A D Milner Durham University, Durham, UK
A Sahraie University of Aberdeen, Aberdeen, UK
A Morin Mount Royal College, Calgary, AB, Canada
R Saxe Massachusetts Institute of Technology, Cambridge, MA, USA
M Moscovitch University of Toronto, Toronto, ON, Canada A B Newberg University of Pennsylvania, Philadelphia, PA, USA B A Nosek University of Virginia, Charlottesville, VA, USA
M Schabus University of Lie`ge, Lie`ge, Belgium T Schmidt University of Giessen, Giessen, Germany C Schnakers University of Lie`ge, Lie`ge, Belgium
D A Oakley University College London, London, UK; Cardiff University, Cardiff, UK
W Schneider University of Pittsburgh, Pittsburgh, PA, USA
E T Olson University of Sheffield, Sheffield, UK
J W Schooler University of California Santa Barbara, Santa Barbara, CA, USA
R Pedersini Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA
xxiii
W Seager University of Toronto Scarborough, Toronto, ON, Canada
S Pockett University of Auckland, Auckland, New Zealand
A K Seth University of Sussex, Brighton, UK
G Rees University College London, London, UK
S Sher University of California, San Diego, CA, USA
R A Rensink University of British Columbia, Vancouver, BC, Canada
J Smallwood University of California, Santa Barbara, CA, USA
A Revonsuo University of Turku, Turku, Finland; University of Sko¨vde, Sko¨vde, Sweden
J Smythies University of California, San Diego, CA, USA
H Roggenkamp University of Pennsylvania, Philadelphia, PA, USA
M Snodgrass University of Michigan Medical Center, Ann Arbor, MI, USA
D M Rosenthal City University of New York, New York, NY, USA
S A Spence University of Sheffield, Sheffield, UK
M Rowlands University of Miami, Coral Gables, FL, USA
P Sterzer Charite´ Campus Mitte, Berlin, Germany
J-M Roy University of Lyon, Lyon, France
N J T Thomas California State University, Los Angeles, CA, USA
xxiv
Contributors
G Tononi University of Wisconsin, Madison, WI, USA
M R Waldman University of Pennsylvania, Philadelphia, PA, USA
C T Trevethan University of Aberdeen, Aberdeen, UK
E S Winer University of Illinois, Chicago, IL, USA
R R Vallacher Florida Atlantic University, Boca Raton, FL, USA
P Winkielman University of California, San Diego, CA, USA
K Valli University of Turku, Turku, Finland; University of Sko¨vde, Sko¨vde, Sweden
J M Wolfe Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA
H Veling Utrecht University, Utrecht, the Netherlands
D Zahavi University of Copenhagen, Copenhagen, Denmark
T D Wager Columbia University, New York, NY, USA
P D Zelazo University of Minnesota, Minneapolis, MN, USA
HOW TO USE THE ENCYCLOPEDIA
The Encyclopedia of Consciousness is intended for use by students, research professionals, and interested others. Articles have been chosen to reflect major disciplines in the study of consciousness, common topics of research by academics in this domain, and areas of public interest. Each article serves as a comprehensive overview of a given area, providing both breadth of coverage for students, and depth of coverage for research professionals. We have designed the encyclopedia with the following features for maximum accessibility for all readers. Articles in the encyclopedia are arranged alphabetically by subject in the table of contents. The index is located in volume 2. Some topics are covered in a multitude of articles from differing perspectives, while other topics may have only one entry. We encourage use of the index for access to a subject area, rather than use of the table of contents alone, so that a reader has a full notion of the coverage of that topic. Coverage for automaticity, for instance, may be found both in ‘‘Automaticity and Consciousness’’ but also under ‘‘Habit, Action, and Consciousness.’’ A reader searching under A for automaticity in the table of contents would easily find one article but would miss the other.
Each article contains a glossary, cross-references to other related encyclopedia articles, and suggested readings and relevant websites for additional information. The glossary contains terms that may be unfamiliar to the reader, with each term defined in the context of its use in that article. Thus, a term may appear in the glossary for another article defined in a slightly different manner or with a subtle nuance specific to that article. For clarity, we have allowed these differences in definition to remain so that each article is fully understandable on its own. Each article has been cross-referenced to other related articles in the encyclopedia at the close of each article. We encourage readers to use the cross-references to locate other encyclopedia articles that will provide more detailed information about a subject. The suggested readings include recent secondary sources to aid the reader in locating more detailed or technical information. Review articles and research articles that are considered of primary importance to the understanding of a given subject area are also listed. These suggested readings are not intended to provide a full reference listing of all material covered in the context of a given article, but are provided as next steps for a reader looking for additional information.
xxv
CONTENTS
Foreword
v
Foreword Biography
vii
Preface
ix
Editor-in-Chief
xiii
Editorial Board
xv
Contributors
xxi
How to use the Encyclopedia
xxv
VOLUME 1 Aesthetics and the Experience of Beauty
W Hirstein and M Campbell
Altered and Exceptional States of Consciousness Animal Consciousness
A Revonsuo
9
D B Edelman
23
Artificial Intelligence and Consciousness
O Holland and D Gamez
Attention: Change Blindness and Inattentional Blindness Attention: Selective Attention and Consciousness T S Horowitz, J M Wolfe, and M A Cohen Automaticity and Consciousness
37
R A Rensink
P D L Howe, K K Evans, R Pedersini,
Autobiographical Memory and Consciousness
M A Conway P Sterzer and G Rees
93 107
S Pockett
Cognitive Theories of Consciousness
123
V de Gardelle and S Kouider
Coma, Persistent Vegetative States, and Diminished Consciousness S Laureys, and M Boly Concepts and Definitions of Consciousness
61 83
C T Trevethan and A Sahraie
Brain Basis of Voluntary Control
47
77
W Schneider
Bistable Perception and Consciousness Blindsight
1
D M Rosenthal
Conscious and Unconscious Control of Spatial Action
B Hommel
135 A Demertzi,
147 157 171
xxvii
xxviii
Contents
Consciousness and Memory in Amnesia
M Moscovitch
Consciousness of Time and the Time of Consciousness The Control of Mnemonic Awareness
S Gallagher
B J Levy and M C Anderson
Development: Consciousness From Birth to Adulthood Emotion and Consciousness
221
S Sher and P Winkielman
Folk Theories of Consciousness
231 C Schnakers,
B F Malle
243 251
A R Mele
265
Functions of Consciousness General Anesthesia
193 205
P D Zelazo
Ethical Implications: Pain, Coma, and Related Disorders M-E Faymonville, and S Laureys Free Will
183
A K Seth
279
M T Alkire
295
Habit, Action, and Consciousness
H Aarts and R Custers
315
History of Consciousness Science
B J Baars
329
History of Philosophical Theories of Consciousness Hypnosis and Suggestion
339
A J Barnier and D A Oakley
Implicit Learning and Implicit Memory Implicit Social Cognition
W Seager
351
A Cleeremans
369
Y Bar-Anan and B A Nosek
383
Inner Speech and Consciousness
A Morin
389
An Integrated Information Theory of Consciousness Intentionality and Consciousness
G Tononi
R Menary
417
Intuition, Creativity, and Unconscious Aspects of Problem Solving Language and Consciousness
403 J I Fleck and J Kounios
P Gordon
431 447
VOLUME 2 Memory: Errors, Constructive Processes, and Conscious Retrieval
W Hirstein
1
Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness R Custers and H Veling
13
Mental Representation and Consciousness
19
Meta-Awareness
D D Hutto
J M Chin and J W Schooler
The Mind–Body Problem
M Rowlands
Mind Wandering and Other Lapses Neglect and Balint’s Syndrome
33 43
J Smallwood P D L Howe
57 71
The Neural Basis of Perceptual Awareness
J-D Haynes
75
Neurobiological Theories of Consciousness
S Kouider
87
The Neurochemistry of Consciousness Neuroscience of Volition and Action
J Smythies M Jeannerod
101 111
Contents
Perception, Action, and Consciousness Perception: Subliminal and Implicit
M A Goodale and A D Milner M Snodgrass, N Kalaida, and E S Winer
Perception: The Binding Problem and the Coherence of Perception Perception: Unconscious Influences on Perceptual Interpretation Phenomenology of Consciousness
159
W Seager A Dietrich
217 231
S A Spence
245
H Roggenkamp, M R Waldman, and J L Bermu´dez
Self: Body Awareness and Self-Awareness
273 289
E T Olson
301
Self: The Unity of Self, Self-Consistency Sensory and Immediate Memory
187 201
M Macmillan
Religious Experience: Psychology and Neurology A B Newberg Self: Personal Identity
B G Breitmeyer
175
Psychoactive Drugs and Alterations to Consciousness Psychopathology and Consciousness
135 147
L Y Atlas and T D Wager
Psychodynamic Theories of the Unconscious
121
T Schmidt
D Zahavi
Philosophical Accounts of Self-Awareness and Introspection The Placebo Response
xxix
R R Vallacher
313
N Cowan
Sleep: Dreaming Data and Theories
327
K Valli and A Revonsuo
Sleep: Implications for Theories of Dreaming and Consciousness V Cologan, and P Maquet Social Foundations of Consciousness
341 T T Dang-Vu, M Schabus,
J Blascovich
357 375
A Social Neuroscience Approach to Intergroup Perception and Evaluation J J Van Bavel and W A Cunningham
379
Subjectivity; First- and Third-Person Methodologies
389
Theory of Mind (Neural Basis) Unconscious Cognition
R Saxe
401
J F Kihlstrom
411
Unconscious Goals and Motivation
K C A Bongers and A Dijksterhuis
Visual Experience and Immediate Memory Visual Imagery and Consciousness
J T Enns
N J T Thomas
William James on the Mind and Its Fringes Index
J-M Roy
B J Baars
423 435 445 459 469
PERMISSION ACKNOWLEDGMENTS
The following material is reproduced with kind permission of Nature Publishing Group Figure 3 of General Anesthesia Figure 3 of Perception, Action, and Consciousness Figure 1 of Sleep: Implications for Theories of Dreaming and Consciousness http://www.nature.com/nature The following material is reproduced with kind permission of Oxford University Press Figure 5 (e) of General Anesthesia Figure 1 of Hypnosis and Suggestion http://www.oup.co.uk The following material is reproduced with kind permission of American Association for the Advancement of Science Figure 4 of The Control of Mnemonic Awareness Figure 5 of The Control of Mnemonic Awareness http://www.sciencemag.org
Aesthetics and the Experience of Beauty W Hirstein, Elmhurst College, Elmhurst, IL, USA M Campbell, National University, La Jolla, CA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Neuroaesthetics – A research paradigm whose proponents attempt to learn more about the human aesthetic experience by studying the brain, in particular the perceptual system.
Introduction What is required for aesthetic experience to take place? Who or what is capable of having aesthetic experience and of judging something to be beautiful? A wild wolf surveying the canyon floor for small moving prey is in an exquisitely sensitive perceptual state, but he is not having an aesthetic experience. Is aesthetic experience solely the province of humans, then? Is there no animal model of our appreciation of art? Or what about the nest of the bowerbird, discussed by Ernst Gombrich in his study of the psychology of decorative art? Gombrich compares the sorts of repetitive patterns and color arrays with no underlying meaning beyond their visual impact found in decorative art with products found in the animal world. He points to the example of the nest of the bowerbird, known for creating highly decorated nests made complex (and beautiful?) with shells, brightly colored objects such as berries, feathers, flowers, and even found items of human manufacture such as plastic or glass. Not only does the male bird spend many hours building, creating, and crafting the nest, but he will also replace moved or disturbed elements of a seemingly intentional design. Moreover, each bower built is completely original and reflects decisions made by its creator on the type, color, and structural array of objects collected for decorating the nest. Should we say that these birds are artists? Should
we attribute the capacity of aesthetic judgment to the female bowerbirds who select their mate on the basis of the brilliance of his bower? In the human realm, our aesthetic drive seems to predate history at the very least. Paleolithic people imprinted cave walls with figures, symbols, and what seem to be projected forms of internal images. It is not known whether the underground paintings at Chauvet cave or Lascaux, France, are truly works of art, products of an ‘aesthetic impulse,’ or just primitive attempts at magical wish-fulfillment, but they clearly arouse aesthetic responses in modern-day viewers who marvel at their beauty and elegance. Principled reflection on the concept of the aesthetic response and on its relation to cognition probably starts with the ancient Greek philosophers Plato and Aristotle, who each had something to say about aesthetics (from the Greek, aisthanomai and aesthetikos, perception or apprehension through the senses). Plato holds beauty itself in the highest regard, as an eternal aspect akin to virtue and truth. However, Plato has something quite different to say about the experience of beauty as effected through perceptual activity, and we get a sense of his displeasure by noting his censure of those who would attempt to play with our sensory intake and manipulate the quality of perceptual experience – artists – who deal in images, representations, and analogies, but not, as Plato would claim, unfiltered truth. Unlike Plato, Aristotle sees a place for the arts and the development of aesthetics as worthwhile imitations of life and its passions: The arts can provide an occasion for katharsis, a purging of emotions through vicarious experience. The field of aesthetics comes into its own in the eighteenth century when Alexander Baumgarten, a German rationalist philosopher, resurrected the term ‘aesthetics’ from the earlier Greek usage to serve as the label for a philosophical discipline dealing with knowledge from sensory perception.
1
2
Aesthetics and the Experience of Beauty
Another eighteenth-century thinker, Immanuel Kant, realized that sensory experience, in all its complex and concept-defying richness, requires its own methods of discovery and analysis. Turning his thinking to aesthetics, Kant distinguishes three types of aesthetic, or reflective, judgments based on feelings of pleasure or displeasure: judgments of the agreeable, judgments of the beautiful (or of taste), and judgments of the sublime. In deeming something agreeable, one is merely attributing to it a subjectively determined pleasure-producing quality. We do not expect others to find it so; we would call something agreeable ‘beautiful’ only in this limited and nonbinding sense. Notions of the sublime go beyond the concept of beauty: The sublime is aweinspiring, fearsome, overwhelming. Turning to judgments of beauty, Kant mentions four aspects that are essential to such judgments: (1) disinterested pleasure (we find pleasure in the object because it is beautiful, not the other way around); (2) universal validity (but this universality is not based on any kind of rigid conceptual categorization); (3) something he calls ‘purposeless purposiveness’ (the result of ‘free play’ of the imagination); and (4) necessity (a subjective attribution of beauty exemplifies how the object ought to be judged).
Aesthetic Perception Aesthetic judgments, although grounded in sensory cognition, differ in the role they play in human experience. Beyond enabling the attribution of properties such as shape, color, odor, texture, and so forth to particular things, aesthetic judgments allow us to ascribe meaning and value to experiences. Not only is a sample of Chateau Lafite wine pungently fragrant, purplish red, tart, and cool, a sip of it delivers a highly complex blend of aromas and flavors that can be named only by comparison to mundane substances: oak, cherry, smoke, chocolate. The experience of the connoisseur supervenes on the base elements of taste and smell, but there is more: He transforms these elements into something new through the prism of his cognitive stock – his knowledge that the wine is rare, outrageously expensive, and coveted by experts – along with his memories of having
drunk other rare wines and his feelings of anticipation fueled by imagining the pleasure to come. Qualities in nonhuman creatures that attract mates (or in some cases frighten away predators) may be compared with those qualities which, at first glance, comprise human beauty or attractiveness. But again, at least in the case of humans, the aesthetics of beauty require more than recognition of forms and sounds and perception of colors and odors. The appreciation of beauty involves conscious activity in which areas of the brain are engaged in the production of a complex mental state – a hybrid of sensory and cognitive components. Recent research on the facial characteristics we consider beautiful has found that symmetric faces are judged to be more beautiful than asymmetric ones. Of course, perfect faces can be boring, and introducing asymmetries or flaws can improve on perfection. We call a single dark mole on one side of a woman’s mouth a ‘beauty mark,’ but correct the asymmetry by placing a matching mole on the opposite side, and we see the pair as a sort of disfigurement. The single mole reorganizes our perception of her face by accentuating certain features and drawing attention away from others. Our understanding of how we read emotions from faces is also progressing nicely, and much of this information is relevant to the artistic expression of emotion through the depiction of faces, such as the Mona Lisa’s intriguing smile. By steadily understanding the processes behind our perception of emotion and attractiveness in faces, we can both come to understand those judgments and teach the artist the dimensions along which faces vary. Many visual artists, particularly painters of the Impressionist and post-Impressionist periods of the late eighteenth and early twentieth centuries, are interested in understanding and exploiting the workings of the eye and brain in perceiving light and constructing colors.
Aesthetics and Neurophysiology The early findings of the Gestalt psychologists demonstrate that our visual systems will complete gaps in lines and corners, fill in colors, and so on in order to produce a more unified and coherent perceptual end-product. We have reached the point where we can provide neuroscientific
Aesthetics and the Experience of Beauty
explanations of these Gestalt phenomena, something that seems to invite the possibility of our gaining a neuroscientific understanding of more complicated perceptual events such as those involving the perception and aesthetic appreciation of works of art. The new science of neuroaesthetics investigates perception and the way the brain generates aesthetic responses. How are forms and patterns processed? Are certain forms ‘preferred’ over others by the perceptual system? Neuroaesthetics has produced two early theories of how the brain generates aesthetic responses. Semir Zeki has developed a theory of aesthetics that appeals to neurophysiological structures and events in the human brain, and he goes some distance beyond well-established claims that they are essential in expounding or expanding traditional ideas about painting as an art. Zeki asserts that artists, because they study the brain and the mechanisms of perception and the causal avenues for producing aesthetic responses, are themselves neurologists. Where a painter differs from a scientist who studies visual perception is in methodology, not in subject matter, according to Zeki. Viewing the work of visual artists from Zeki’s perspective provides a deeper insight into why an artist does what he does in constructing a painting, why he chooses the exact shapes, colors, and specific juxtapositions of painterly elements such as line, shading, perspective, and so on. The artist is exploring, exploiting, challenging, and manipulating the brain’s ability to absorb and interpret sensory data in novel, unexpected, and exciting ways. Therein lies the basis of the aesthetic response; not only does the spectator see shapes, lines, and colors of various tints and shades when he or she looks at a painting, but there is also the essential component of any successful work of art that is the production of a type of mental state that outstrips ordinary perception. Great, or even good, art will arouse subtle, or perhaps strong, emotions. It will jostle and arouse memory and effect numerous associations with other mental states. It can induce a sense of joy or of infinite sorrow; it may even lift the spectator to heights of spiritual or existential transcendence. And the artist accomplishes all this without electrodes, probes, or chemicals. The artist, in Zeki’s view, is a virtuoso scientist using instruments much simpler
3
and less invasive, but requiring great ingenuity and immense perspicacity in their application. Zeki emphasizes the analogy between the way that our perceptual systems function to extract the more permanent patterns in the flow of energy reaching them with the way that artists often strive to capture the essence of objects they depict. Ramachandran argues further that the representational artist is presenting us with a sort of caricature of the represented object, in which certain perceived features have been exaggerated while other features have been deemphasized. A skilled artist can evoke a particular person or place with a few deftly drawn lines, using the art of caricature to capture a sort of ‘formal essence’ of the represented person or thing. What we are able to see through such caricatures depends, in scientifically describable ways, on the structure of our visual system and its connections to the different memory systems. Much of visual art can be captured by rules of form, many of which are derived from the findings of the Gestalt psychologists, such as symmetry, balance, and grouping of forms. Artists combine these basic form-primitives in new and evocative ways, according to Ramachandran.
Consciousness and Aesthetic Experience What is the connection between consciousness and aesthetic experience? Are aesthetic experiences necessarily conscious, for instance? While it may be no easier to show the necessity of consciousness to aesthetic experience than to any other sort of experience or mental function, aesthetic experiences are often included in lists of paradigm conscious experiences. It is also interesting to note that several of the classic puzzles that theorists of consciousness are tasked to solve either make reference to aesthetic reactions themselves, or to their essential ingredients, such as the ability to perceive color, as in Frank Jackson’s ‘Mary case.’ Mary is a future neuroscientist who knows about all of the physical processes involved in human color vision, but has never seen any colors other than black and white (she is kept in a special room, only given certain foods, etc.). When Mary is
4
Aesthetics and the Experience of Beauty
finally allowed to see colors, she seems to learn something new, that is, what it is like to consciously experience red. But she already knew about all the physical events involved, and so there appears to be a problem for theorists who claim that the mind is physical, in that the conscious experience of red seems not to be among the physical events that we agreed initially that Mary knew. Another classic puzzle, Dennett’s ‘Chase and Sanborn problem,’ involves something close to aesthetic perception – in this case, coffee tasting. Chase and Sanborn have been brewing their coffee for decades, but neither of them likes the taste any longer. Chase says that his taste preferences have not changed over time, but that the way the coffee tastes to him has changed. Sanborn says that the coffee tastes the same to him, but his preferences have changed so that he no longer enjoys that flavor. Dennett claims that no physicalist theory can capture the difference between Chase and Sanborn, because he believes that the taste and the preference for the taste cannot be teased apart. Delving further into the distinction between perception and preference, and whether and how it might be made, will also reveal information of interest about art. Indeed, to pose a final neuroaesthetic theory of the experience of art, the scientist may have to wait until these more basic ontological questions are answered. But must aesthetic experiences be conscious? Could there be a zombie artist, for instance, a being who looks and acts just like a normal person, creating and appreciating works of art but not capable of consciousness? But, is not the whole point of creating works of art to produce those special conscious experiences? What is the point of his making these objects, we want to know. We might look more closely into our normal aesthetic experience for unconscious events that are at least important to our experiences, if not outright components of it. One candidate for an unconscious aesthetic response is the sort of case where, for example, one realizes some time after seeing a film an interesting subplot of which he had not been conscious aware on first seeing the film. In more complicated artworks, especially of the narrative variety such as the novel, how the author handles the main plot and its relations to various subplots is an important part of how we aesthetically
evaluate such works. In cases like the film viewer who only later fully appreciates the work, the brain seems to be continuing the attempt to understand the artwork all on its own. Certainly many artists have reported being inspired by dreams, although it is debatable whether dreams themselves are a type of conscious state. Consciousness relates to aesthetic experience at two basic points, which then can permute and recombine to produce more complex aesthetic experiences. The first point of contact concerns the artist’s act of creation, including the intentions specifying how the artwork will be made and what the end result will (ideally) be like. The second point of contact concerns the conscious perception of the finished artwork by a human perceiver. One of the first recombinations of these basic ingredients occurs in the brain of this viewer when she considers the intentions of the artist. There is also the fact that many artists attempt to emulate the mind of a viewer, or often more specifically, viewers of different sorts: critics, peers, average people, etc. So the viewer is modeling the intentions of the artist, while the artist is modeling the understanding of the viewer. This sort of activity falls under the category of ‘theory of mind’ or ‘mindreading’; in this case intentions and perceptual states are modeled. Another kind of reading we do of other people occurs when we watch their intentional activities while employing mirror neurons to understand their actions. Rizzollati, who is the first to attribute the familiar imitative behavior we observe not just in monkeys but in humans as well to these specialized ‘mirror neurons,’ notes that mirror neurons are at their most effective when the observer is face-toface with the observed. Stafford suggests that this fact explains why so many visual artists focus on the face as the center of a work, sometimes presenting the viewer with nothing but a face. Here the artist is playing up the natural emphasis on frontality and the intuitive recognition that some of our most deep-seated reflections about who people are, about our own identity, arise from contemplating the face of the other, or of the face in the mirror. Several neurological patients have recently been described who have become obsessed with art after brain lesions, or repeated epileptic seizures, very often in the temporal lobes. These obsessions typically focus on a particular medium, often painting
Aesthetics and the Experience of Beauty
or music. The obsessions can compel the patient to spend every waking moment creating and performing art, and some of the patients have become quite accomplished and are even able to sell paintings for high prices, or impress music critics with their composition or performances. Apparently, these patients find thinking about and creating works of art extremely rewarding, so much so that they sometimes forget to eat while in the throes of creation. This suggests an explanation for aesthetic responses that relates them to consciousness in general. There is little point in assembling the expensive and complicated machinery needed to achieve consciousness if we do not trouble ourselves to use it. Perhaps because conscious thinking requires effort, there needs to be a reward for engaging in conscious thought. Nature’s way of enticing us toward conscious activity, according to this view, is to make conscious experience itself rewarding. Contemplation of the contents of consciousness is not, as we might have thought, basically neutral, with any emotional response being traceable to the particular contents of that mental state and not to consciousness itself, but rather pleasurable in itself. Sometimes, naturally, the particular contents of consciousness are unpleasant enough to cancel out this basically positive reaction, as usually happens in the case of pain experience, for instance. One perhaps sees a similar ‘enticement’ mechanism at work in the case of sex; nature entices us to use our sexual organs so that the race proliferates by making their use highly rewarding. It seems clear that conscious sensations play a vital role in this case.
Conclusion Once art is brought into the realm of science, several classical claims about it can be seen to be empirically testable. In his landmark work, Art and Experience, for example, John Dewey argues that aesthetic experience requires dissolution of self and object as separate existences. This stands in contrast with scientific observation, he notes, which calls for as much separation from subjectivity as possible in perceptual judgments. Empirical observation for the purposes of science idealizes a form of perception cleansed of the observer’s
5
expectations, personal desires, and preexisting beliefs. Any significant level of projection or interpretation skewed toward personal preferences, tastes, emotions, or remembrances is considered as tainting, or even invalidating, scientific observations. On the contrary, aesthetic experience depends, both ontologically and epistemologically, on such subjective contributions. Conscious or not, these sorts of differences should be quite measurable as differences in brain activity. The hypothesis that specialized brain activity can be directly correlated with specific sorts of aesthetic responses received some confirmation by a recent experiment conducted by Kawabata and Zeki using functional magnetic resonance imaging (fMRI) to locate and quantify what they term ‘neural correlates of beauty.’ Prior to being scanned, their subjects classified paintings of different genres (abstract, still life, portrait, and landscape) as beautiful, neutral, or ugly. Subjects then viewed the same paintings while in the fMRI scanner. Distinct, specialized areas of the brain were found to be activated when perceiving different genres of painting. More interestingly, independent of the genre of the perceived painting, the orbitofrontal cortex was found to be differentially activated when the subject was perceiving beautiful or ugly stimuli, while the motor cortex was found to be activated differentially during perception of paintings as beautiful or ugly. Activity in the orbitofrontal cortex increased during the perception of beautiful stimuli and decreased in the case of ugly stimuli. The reverse was seen to be true of the motor cortex. Stimuli perceived as ugly produced the most activity while stimuli judged to be beautiful the least. That finding perhaps helps explain why aesthetic perception can sometimes have a curiously paralyzing effect on us, as Stendahl noted. Typically conscious experience induces one to act, whereas aesthetic experiences ‘stop us in our tracks’; we end bodily motion to engage in deeper contact with the artwork. Aesthetic appreciation of this sort is not goal-oriented. If any action is motivated by aesthetic appreciation, it is to move closer to the art object, or at least to prolong the experience. One often-heard objection, especially from the art establishment, is that neuroaesthetics is ‘reductionist,’ in the sense that its proponents intend to
6
Aesthetics and the Experience of Beauty
claim that the aesthetic experiences we are aware of are somehow unreal or unimportant, whereas the underlying brain events, as described by brain scientists, contain all the reality there is to them. The critics are forgetting, however, that science need not work this way. There is no reason in principle why understanding the phenomena underlying something should make our existing ways of understanding that thing go away. This only tends to happen when serious flaws are discovered in the existing ways of thinking and speaking, and this is not in general what is being proposed by the neuroaestheticians. This sort of noneliminative reduction can serve to ground aesthetic experience more firmly in scientific perceptives, making it more solid and real rather than less.
Suggested Readings Aristotle (1984) In: Barnes J (ed.) Poetics, The Complete Works of Aristotle. Princeton, NJ: Princeton University Press. Collingwood RG (1995) Principles of Art. Oxford: Clarendon Press. Dewey J (1934) Art as Experience. New York: Perigee Books. Gombrich E (1956) Art and Illusion: A Study in the Psychology of Pictorial Representation. New York and London:
Andrew William Mellon Lectures in the Fine Arts, 5; Bollingen Series, 35: 5. Gombrich E (1979) The Sense of Order, A Study in the Psychology of Decorative Art. Oxford: The Wrightsman Lectures. Goodman N (1976) Languages of Art: An Approach to a Theory of Symbols. New York: Hackett. Jackson F (1982) Epiphenomenal qualia. Philosophical Quarterly 32: 127–136. Jackson F (1986) What Mary didn’t know. The Journal of Philosophy LXXXIII 5: 291–295. Kawabata H and Semir Z (2004) Neural correlates of beauty. Journal of Neurophysiology 91: 1699–1705. Kant I (1790) In: Nicholas W (ed.) The Critique of Judgment. Oxford: Oxford University Press. Onians J (2007) Neuroarthistory: From Aristotle and Pliny to Baxandall and Zeki. New Haven and London: Yale University Press. Plato (2002) In: Cooper JM (ed.) Complete Works. New York: Hackett. Ramachandran VS (2004) A Brief Tour of Human Consciousness. New York: Pi Press. Rizzolatti G and Craighero L (2004) The mirror neuron system. Annual Reviews of Neuroscience 27: 169–192. Sacks O (2007) Musicophilia: Tales of Music and the Brain. New York: Knopf. Stafford BM (2007) Echo Objects: The Cognitive Work of Images. Chicago, IL: Chicago University Press. Wollheim R (1980) Art and Its Objects. Cambridge: Cambridge University Press. Zeki S (2000) Inner Vision: An Exploration of Art and the Brain. Oxford: Oxford University Press.
Biographical Sketch
William Hirstein is professor and chair of the philosophy department at Elmhurst College, in Elmhurst, Illinois, USA. He received his PhD from the University of California, Davis, in 1994. His graduate and postdoctoral studies were conducted under the supervision of John Searle, V.S. Ramachandran, and Patricia Churchland. He is the author of several books, including On the Churchlands (Wadsworth, 2004), and Brain Fiction: Self-Deception and the Riddle of Confabulation (MIT, 2005).
Aesthetics and the Experience of Beauty
7
Melinda Campbell is associate faculty in arts and humanities at National University and also teaches courses in philosophy and women’s studies at San Diego Mesa College. She is an artist and the author of several articles on the metaphysics of color. She studied with Richard Wollheim and received her PhD from the University of California, Davis, in 1993.
Altered and Exceptional States of Consciousness A Revonsuo, University of Sko¨vde, Sko¨vde, Sweden; University of Turku, Turku, Finland ã 2009 Elsevier Inc. All rights reserved.
Glossary Hypnagogic state – The transitional state between wakefulness and sleep when falling asleep. Hypnopompic state – The transitional state between sleep and wakefulness when waking up from sleep. Lucid dream – A dream during which the dreamer realizes that the experience is a dream. Muscular atonia – The lack of muscle tension in voluntary muscles, causing temporary inability to move. NREM sleep – Non-REM sleep, stages of sleep without rapid eye movements and less frequent dreaming, or no dreaming at all. Psychedelic drugs – Chemical substances that affect the central nervous system and cause radical changes in overall mental function, such as hallucinations and delusions. REM sleep – Rapid eye movement sleep, a stage of sleep when the eyes move rapidly behind the closed eyelids, and when the subject is frequently dreaming.
Introduction Consciousness, the stream of subjective experiences, manifests itself in many different forms, such as waking perception, dreaming, and mental imagery. Under some circumstances our subjective consciousness becomes so different from our usual experience that the state of our consciousness can be said to be qualitatively altered. The notion of altered state of consciousness (ACS) presupposes that there is some definable normal or baseline state of consciousness that is temporarily lost or
somehow transformed in ACSs. What exactly is the normal state of consciousness and what counts as an altered state is difficult to define precisely and therefore continues to be a matter of debate. In the following, different definitions of altered state are first considered and then the major phenomenological features of the most significant ASCs are reviewed. Some altered states are called exceptional or higher states of consciousness. The notion of a ‘higher’ state of consciousness suggests that different ASCs can be ordered in a hierarchical manner such that some altered states are lower or deficient in comparison with the ‘normal’ state of consciousness (say, hallucinations induced by high fever or the delirium caused by heavy drinking), whereas others, the higher states, are in some sense better, more desirable, or perhaps spiritually more advanced than the normal state. The so-called higher states are usually experienced as positive and personally significant. Many of them are closely related to religious and mystical experiences and thus are among the most intriguing ASCs. The major phenomenological features of exceptional and higher states will be reviewed and possible explanations for them explored. In the scientific study of consciousness, ASCs may shed light on aspects of consciousness that remain otherwise hidden. Therefore, they constitute a unique source of empirical evidence for theories and models of consciousness. They may also reveal the underlying brain mechanisms of some aspects of consciousness that would be difficult or impossible to study by focusing only on the normal state of consciousness.
What Is an ASC? All definitions of ASC take for granted that there is a normal baseline state of consciousness, and that any ASC only temporarily deviates from this 9
10
Altered and Exceptional States of Consciousness
normal state in some crucial manner. Hence, an ASC is a temporary, reversible state that typically lasts from a few minutes to at most a few hours. Permanent, irreversible changes in conscious experience, as in psychiatric or neurological disease, are usually not counted as ASCs. In an ASC, what exactly is the nature of the deviation from the normal state? Can it be precisely defined so that a distinguishing feature (or a set of such features) can be pointed out that separates all the ASCs from the typical normal state of consciousness? Attempts to define the concept of ASC and the empirical criteria that separate ASCs from the normal state have been put forward, but it remains unclear if any of these definitions is successful in drawing a clear line between normal and ASCs. Typically, the ‘normal,’ unaltered state of consciousness (henceforth NSC) is taken to be the state where we are awake, alert, perceptually aware of our own self and of the physical and the social environment, and capable of rational thought and smooth behavioral interactions with the environment. Our thought processes, sensory and perceptual representations, emotional experiences and reactions, as well as current beliefs reflect our current situation in a relatively realistic manner. When in the NSC our thoughts and beliefs are not delusional, our sensations and perceptions are not hallucinatory, our emotional experiences and expressions are not overly exaggerated, and our voluntary mental activities and external behavior is under control and not entirely inappropriate for the physical and the social situation in question. If any of these features are however present in our experience or behavior, they may be signs of an ongoing ASC. Still, this list of potential signs of an ASC is hardly a good definition of the concept. We should expect the definition of the concept of ASC to draw a clear line between the paradigmatic NSC and all the different ASCs. Thus, the ideal definition should focus on the common core of all the different types of ASCs, and also connect the notion of an ASC to an overall conceptual or theoretical framework in the science of consciousness. In the following sections we will consider different attempts to define the common core of all ASCs.
ASCs as Defined by an Overall Changed Pattern of Experience One way to define an ASC is to say that in an ASC, the overall pattern of subjective experience is significantly different from the baseline NSC. The idea behind this definition is that the change that has happened in consciousness is global in nature and therefore affects several different dimensions of experience, cognition, or behavior. The dimensions of experience where the changes take place are many and varied, for example, attention, perception, mental imagery, inner speech, memory processes, thought processes, meaningfulness of experience, time experience, emotional feelings and expressions, level of arousal, self-control, suggestibility, body image, and personal identity. While such a list is surely illustrative of aspects of experience that may be altered in ASCs, it still fails to offer a crisp general definition of the concept of ASC. Clearly, different ASCs may involve more or less global changes in subjective experience (i.e., involve changes in only a few or almost all of the above-mentioned dimensions). Thus, it remains unclear how covering or drastic the changes should be to count as an ASC. The borderline between NSC and ASCs thus remains rather vague. A related problem is that these same dimensions of consciousness vary widely also within the boundaries of the baseline NSC. Watching a film in a movie theater, giving a lecture, taking a shower, playing football, driving a car, singing in a choir, diving into a swimming pool, walking in the woods, making love, joking with friends, sitting in a dentist’s chair while one’s tooth is being drilled – all these experiences usually take place in the NSC, yet we may say that the overall pattern of subjective experience (and cognition and behavior) surely is drastically different when we move from one situation to another in the above list. The richness and sheer variety of the contents of experience within the scope of the NSC makes it difficult to draw a sharp line between the NSC and ASCs only on the basis of changes in the patterns of experience. Furthermore, in some ASCs the contents or overall patterns of experience are not much different from the NSC. For instance, a dream-related
Altered and Exceptional States of Consciousness
ASC called false awakening is known to occur during sleep, often in the morning at the time when one should get up. This dream experience may be almost indistinguishable from the corresponding events in the NSC. In false awakening, the sleeper dreams about waking up, getting out of bed, and engaging in normal morning routines, only to wake up again after awhile to discover that he or she is still in bed and that the first experience was but a realistic dream of getting up. During the false awakening the sensations, percepts, emotions, thought processes, etc. may be almost identical to a real awakening – no wonder then that the sleeper does not realize he or she has not woken up at all! Thus, the overall patterns of experience change radically within the NSC, whereas in some ASCs, the overall pattern of experience remains identical to a corresponding NSC. Therefore, the definition of ASC as an altered pattern of experience may not be able to demarcate the NSC from ASCs in a reliable manner. ASCs as Defined by the Recognition of a Change in the Overall Contents or Patterns of Experience Perhaps a better definition could be reached by adding a further condition: the subject having the ASC must feel or recognize that his or her experience is remarkably different from the normal state. This definition adds the requirement that an ASC must be reflectively recognized or recognizable by the subject having it. Clearly, this is not true about many ASCs during the time they take place. For example, during dream experiences, we are usually oblivious to the fact that we are dreaming, and it is extremely difficult to arrive at the conclusion that ‘this is a dream’ while we continue to dream. Also, sometimes in the NSC we may mistakenly believe we are in an altered state, if something utterly unexpected or shocking suddenly happens so that we have difficulties in believing what we see before our very eyes, but might for a moment think that we must be dreaming. However, if we allow that the recognition of the ASC as an ASC may take place also afterward when the altered experience is already long gone,
11
then we will be able to correctly classify most ASCs, such as dream experiences, as ASCs. Still, it seems we do the classification not on the basis of the changed pattern of experiences as such – in some mundane dreams and especially in false awakenings, the experiences themselves are identical to waking experiences in NSC – but rather on the basis of the observation that we must have been asleep and therefore only dreaming. Here, the crucial difference between NSC and ASC is not in the content of experience, after all, but rather in the relation between the experience and the real world that the experience represents. ASCs Defined as the Changed Relation between the Content of Consciousness and the Real World This leads us to a third potential definition ASCs. Perhaps the core of all ASCs is not the change in the overall patterns of experience (which of course may also happen) but rather the fact that while in an ASC, the content of our experience relates differently to the real world. According to this definition, the causal or representational relation between contents of experience and their typical sources breaks down in such a way that (at least some) contents of experience in consciousness carry false information about (some aspects) of the world, or of ourselves. The contents of experience, in other words, in some way misrepresent external reality or the self. This definition can easily handle all cases of hallucinatory sensations or percepts. Positive hallucinations are experiences that have no corresponding stimulus as cause in the external world. Our experience thus misrepresents such causes as being out there, but the pattern of experience need not be out of the ordinary in any way. Negative hallucinations, conversely, are omissions of stimuli from experience that would become experienced in the NSC but are not accessible to consciousness because of the ASC. For example in hypnosis or deep meditative states, subjects may not see or hear some stimuli at all if the direction of attention and the contents of consciousness are strongly dominated by other things. Also altered time experiences are covered by this definition: in an
12
Altered and Exceptional States of Consciousness
ASC, our conscious experience misrepresents objective time so that it seems to flow either more slowly or more quickly than it really does. During meditation or hypnosis, an hour may pass in a time that feels like a few minutes only. This definition also seems to include delusions, changes in memory functioning and rational thought. Delusions are false beliefs held in the face of contrary evidence or despite knowledge of the actual state of affairs. In dreams we are often deluded into believing things we would never accept during wakefulness. Psychedelic drugs inducing ASCs may have similar effects. Delusions thus misrepresent the world by interpreting some aspects of the world or ourselves as being in some sense something quite different than what we would take them to be in the NSC. An ASC may involve both retrograde and anterograde amnesia. During the ASC we may not be able to access memories we would have no difficulties in retrieving in the NSC. Moreover, the events that happen during the ASC may not be normally encoded into memory and thus our memories of the experience may be vague and quickly forgotten. When we dream, we usually cannot recall our true circumstances in the real world: that we went to bed, what happened during the day, or the place where we are sleeping, although we immediately remember such things if we wake up for a moment, even in the middle of the night in total darkness. During dreaming memory encoding does not work normally: we tend to forget our dream experiences more quickly and more easily than corresponding waking experiences. Some subjects show spontaneous amnesia about events during hypnosis. Furthermore, false or confabulated memories may be spontaneously formulated, so that during an ASC we seem to remember something that in actual fact has never happened and that we would never take seriously in our NSC. Thus, in these cases our memory (or our conscious access to our memory) misrepresents our past experiences in one way or another, and our critical thinking is not working properly to recognize the fact that we are deluded, amnesic, or confabulatory. Still, also the definition of ASCs as involving misrepresentations has its difficulties. Trivial illusions and misperceptions sometimes occur in the NSC, thus not all types of misrepresentations
count as criteria for an ASC. Furthermore, the notion of misrepresentation of reality may not apply so easily to changes in attention or emotional experience and expression. But if we take emotional states as normally being caused by some external circumstances and/or internal beliefs, then having strong emotions such as fear or ecstasy without any appropriate external or internal cause may be comparable to a kind of misrepresentation of our true situation. Definition of ASC Perhaps the most workable definition of an ASC could be arrived at by combining the above ideas: An ASC is any temporary, reversible state of consciousness in which the relationship between at least some of the patterns of experience, and their typical, appropriate causes has been changed so that some patterns of experience tend to occur without their appropriate causes, or some patterns of experience do not occur despite the presence of their appropriate causes, or both. An individual should be able to recognize either during the ASC or after it that an ASC is or was occurring, that the pattern of experience was different and its connection to the external world was not equally accurate as in the normal waking state. An ASC may be brought about by multiple different circumstances such as perfectly normal changes in brain physiology and neural activity (sleep and dreaming), chemically induced changes in brain physiology (alcohol, psychedelic drugs), special patterns of stimulation in the physical and/or social environment (hypnotic inductions and suggestions, sensory deprivation, strong rhythmic or repetitive patterns of stimulation) or in internal control of attention (meditation). In the following we will take a closer look at the major types of ASCs.
Typical ASCs Sleep-Related ASCs While the most typical NSC is an alert state of wakefulness, the most typical ASCs are associated with sleep. When we fall asleep, the patterns of experience and the connection between experience and the external world change suddenly and
Altered and Exceptional States of Consciousness
drastically. At the borderline between wakefulness and sleep, we may experience an ASC that involves a mixture of perception and dream images. This transitional state from wakefulness to sleep is called hypnagogia (literally, leading to sleep) and the internally generated images in this state are called hypnagogic hallucinations. Conversely, hypnopompic hallucinations (meaning literally: leading out of sleep) occur in the transitional state from sleep back to wakefulness. Hypnagogic and hypnopompic hallucinations are regarded as intrusions of rapid eye movement (REM) sleep imagery into the borderline between wakefulness and sleep. During these hallucinations some degree of perceptual or bodily awareness remains, but other aspects of experience consist of hallucinatory images. Most typical are visual hallucinations of various kinds, from simple geometric forms to objects, faces, animate characters, or entire landscapes. Also auditory phenomena are common: noises, sounds, music, or human voices. Other sensory modalities may also be involved, such as bodily feelings of various kinds or tactile sensations. Hypnagogic hallucinations give way to brief sleep-onset dreams where sensorimotor events from the previous days are briefly revisualized. If you have studied intensely the whole day you may see similar figures or words dancing in front of your eyes at sleep onset, or if you have played sports or computer games, you may see brief excerpts of the game. Also other kinds of brief dreams with complex content may be experienced at sleep onset. Another ASC that often takes place in the hypnagogic or hypnopompic state is sleep paralysis. It is a mixture of wakefulness and REM-sleeprelated muscular atonia: the subject feels awake, but cannot move any part of his or her body. This may be accompanied by difficulties in breathing or by the feeling that something heavy is pressing against one’s chest. Sometimes this is perceived to be an evil character that is sitting on the chest (this is called ‘the old hag experience’). Sleep paralysis is also often associated with the sense of an evil presence or the strong feeling that there is another person or being present close by, observing the subject and having some sort of evil intentions toward him or her. Thus, sleep paralysis may be a frightening experience emotionally. It has been speculated that the
13
true origin of many experiences taken to be paranormal in the vernacular (ghosts, apparitions, UFO abductions) is to be found in sleep paralysis and hypnagogic hallucinations. Such phenomena typically happen during the night when the subject is lying in bed in a dark room and likely in a transitional state between sleep and wakefulness. As the experiential content of these ASCs may be intense, realistic, and extremely frightening, the subject who has never heard about sleep paralysis or hypnagogic hallucinations may interpret the experiences as representing real but paranormal events in the world. During sleep, subjective experiences of some kind occur most of the time, constituting the contents of further sleep-related ASCs. About 85% of REM sleep awakenings and about 25%–50% of non-REM (NREM) awakenings lead to reports of subjective experience. Subjective experiences during sleep can be divided into two categories: sleep mentation and dreaming. The difference between these two is in the complexity of experience. Typical sleep mentations consist of a single image that occurs in a single sensory modality and remains static or repeats itself in the same form. An image of a visual object, a word or sentence or sound heard repetitively, or a thought that runs through the mind again and again are common types of sleep mentation. By contrast, dreaming involves complex, organized, and animated imagery in multiple sensory modalities that shows progression and change through time. Thus, dreams depict a sensory–perceptual world with objects and characters, and simulate events that take place in such a world. Systematic research on the content of dreams has shown that all of our sensory modalities may be involved in dreams, vision and auditory experiences being the most common however, that most dreams have a central character or a dream self, who is a representation of the dreamer in the dream; that most dreams also include other human (or animal) characters; and that social interaction and communication between dream characters is common. Negative events and emotions are more common in dreams than positive ones, and some activities we often engage in the real world are almost totally absent from the dream world, such as reading, writing, typing, working
14
Altered and Exceptional States of Consciousness
with computer, or calculating. Dreams often include bizarre contents and events that would be impossible or highly unlikely in the real world, but we suffer from disorientation, amnesia, and a lack of critical thinking in dreams, and therefore we are only rarely able to recognize the peculiarity of the dream events while they go on, or able to recognize the dream for what it is – a full-scale hallucination or simulation of a world. Lucid dreaming is the name for the special ASC in which we become reflectively aware of the fact that our ongoing experience is a dream while the dreaming continues. Thus, lucidity renders the dream into a pseudohallucination: a hallucination whose subject is fully aware of the hallucinatory nature of the current experience. Lucid dreaming can be considered a higher state of consciousness and thus will be handled in more detail with other higher states below. Further sleep-related ASCs include bad dreams or extremely unpleasant dreams that do not wake the dreamer up, nightmares or extremely unpleasant dreams that wake the dreamer up, and sleepwalking or nocturnal wandering which involve a mixture of NREM sleep and wakefulness: the sleepwalker’s eyes are open and he or she usually pursues some unreasonable goal without realizing that he or she is still sleeping and that the goal does not make sense. All in all, sleep-related ASCs are perhaps the most common types of ASCs. Dreaming and sleep mentation are states that occur basically every night for every one of us even if we cannot recall the experiences we had after we wake up. While they are rightly classified as ASCs, there is nothing pathological about them: they are not abnormal phenomena, but perfectly normal concomitants of physiological sleep. Hypnosis as an ASC Hypnosis occurs in a situation where a hypnotist gives first a hypnotic induction to a subject, telling the subject to intensely focus on something (such as the hypnotist’s voice, a light), relax, and let one’s eyes close. After the induction to hypnosis, the hypnotist gives more specific suggestions about specific changes in experience for the subject. The suggestions typically concern changes in how one feels
one’s own body or one’s own actions (e.g., feelings of heaviness in one arm, inability to open one’s eyes, numbness or painlessness of a part of the body) or changes in sensation, perception, memory, or thinking (e.g., seeing or hearing something that is not there, not remembering something one normally remembers, not being able to think straight, or believing uncritically some statement and acting as if it were true). Different people respond differently to hypnotic inductions and suggestions. Some experience almost nothing at all, whereas others report that they really experienced all the things that the hypnotist suggested to them. Among hypnosis researchers, there is a long-lasting controversy whether hypnosis involves an ASC or whether hypnosis is just a peculiar social situation where subjects behave according to their expectations and play along with the rules (just like in any other social situation) but do not enter any kind of ASC in the process. This controversy has been difficult to resolve empirically as there have been no universally accepted definitions or measurable criteria for what would count as an ASC. One possibility is that only a small proportion of subjects truly enter an ASC after getting the hypnotic induction. These very highly hypnotizable people, also called hypnotic virtuosos, may, due to the suggestions given to them by the hypnotist, experience vivid hallucinations in different sensory modalities, amnesia such that they suddenly cannot remember something that they obviously do remember outside hypnosis. They may also forget totally what has happened during the hypnosis session, and have an altered sense of time, thinking that they were under hypnosis only a few minutes when an hour has passed. Hypnotic suggestibility or the ability to experience the suggested changes is normally distributed in the population, so that most people are moderately suggestible, whereas a small proportion of people is not suggestible at all and feels no changes in experience, and an equally small proportion is highly suggestible and feels many kinds of changes in experience if such changes are suggested to them. Thus, most people, when hypnotized, experience at most only similar things as one is expected to experience with guided mental imagery in a relaxed state. What happens in hypnosis
Altered and Exceptional States of Consciousness
for most people, especially the low and moderately hypnotizable ones, does not necessarily involve any ASC, but only mental imagery, expectations, and playing voluntarily along with the hypnotists’ suggestions.
Exceptional or Higher States of Consciousness Exceptional states, also called higher states of consciousness, are considered deeply meaningful, satisfying, and desirable, but also difficult to reach or maintain. They go beyond the NSC in the sense that in them subjective experience reaches extreme attentional, emotional, or cognitive levels. An exceptional or higher state characterized by changes in attention involves total absorption with the object in the narrow focus of attention (onepointedness of mind), or alternatively the widening of attention to simultaneously cover the entire sensory–perceptual field (full awareness or mindfulness). Higher attentional states are often characterized by the absence of reflective thoughts, especially negative ones, and a deep inner peace or calmness of mind, which can be experienced as highly satisfying or even blissful. Higher emotional states typically involve strong positive feelings of well-being, contentment, lovingkindness, compassion, joy, elation, or bliss. The quality of inner emotional experience is thus characterized by happiness: the presence of intense positive affect and the absence of negative affect. Higher cognitive states involve feelings of deep understanding, sudden revelation or insight into the nature of things, glimpses of higher knowledge about the order of the universe, or feelings of being directly connected or absorbed into the cosmos or with higher spiritual realms or beings, such as god. In these states, one seems to get in touch with deeply meaningful information about the nature of reality or have direct knowledge of it. However, it is unclear whether such information or knowledge is actually possessed, or is it only a feeling of deep insight without any actual informational content. In any case this knowledge is often impossible to express precisely in words or it is easily lost once one returns to the ordinary state and level of consciousness. Even when not lost but recalled, it
15
may lose its original significance and seem a trivial platitude when reconsidered in the NSC. This is especially true of the deep insights and supposed truths obtained in drug states, like under the influence of LSD. The attentional, emotional, and cognitive components of exceptional or higher states of consciousness may appear separately or in various combinations in different ASCs. Next, we will review some paradigm examples of exceptional or higher states. Meditation Meditation as such is not a higher or ASC, but rather a set of various techniques and practices that aim at controlling and altering consciousness. Thus, meditation may lead to an altered state, and sometimes this is the explicit goal of meditation. There are far too many radically different meditation techniques and traditions to cover in this context; thus, only some of the most central principles and techniques can be mentioned here. In one way or another, different meditation techniques involve deliberate control or manipulation of attention. In concentrative types of meditation, the scope of attention is kept narrow and highly selective, as only a particular content of consciousness (an object, a mental image, a word or a sentence, a repetitive action such as breathing) is fixed into the focus of attention for prolonged durations whereas everything else, all distractions, are driven out of consciousness. In mindfulness types of meditation, by contrast, the scope of attention is widened to encompass all available sensations, percepts, emotions, bodily feelings, etc., to be vividly aware of all of them in as much detail and intensity as possible. In Buddhist and Yoga literature, the term ‘samadhi’ refers to a higher state of meditative consciousness in which perfect concentration is reached and where the distinction between the object of meditation and the subject who meditates totally disappears. This state is characterized by mental one-pointedness and a merging together of the object and the subject. According to some traditions, the systematic practice of meditation to reach samadhi states can lead to progressively higher mystical states of consciousness, such as
16
Altered and Exceptional States of Consciousness
nirvana or enlightenment, which will be discussed below in connection with mystical states. Some forms of meditation combine the attentional and the emotional components of higher states and deliberately focus on diminishing negative emotions or strengthening positive emotions, such as loving-kindness or compassion. Thus, such meditative practices may lead to higher states of consciousness with intense positive emotional experiences. There is evidence from recent brain imaging studies that Buddhist monks, who have practiced this type of meditation for decades, actually do reach a state of consciousness unreachable by beginners or laymen, evidenced by the strong changes in brain activity that correlates with their meditative state. Optimal Experience and Flow Optimal experience is a higher emotional (and attentional) state of consciousness which characterizes the best moments of our lives, moments when we feel deep enjoyment, exhilaration, or happiness and forget about everything else. Empirical studies have revealed that people reach this kind of ‘flow’ state of mind when there is a certain kind of order in consciousness: attention is firmly focused on reaching a meaningful, challenging goal; we are intrinsically motivated to reach the goal for its own sake, and our skills and resources are just sufficient to reach the goal. When we struggle toward such goals, we momentarily forget about everything else, including the sense of time and our own selves; we become fully immersed into the actions necessary to reach the goal. During such moments, experience simply flows onward, we feel in control of the situation, and our minds are free of worries. Experience becomes one with the actions leading to the meaningful goal, and everything else disappears from consciousness. In some ways, intense flow experiences are similar to meditative samadhi states where self-awareness disappears and experience becomes one with the focus of attention. Flow states may emerge in almost any kind of activity, such as listening or performing music, engaging in games or sports, hiking in nature, immersing oneself in a conversation with an interesting person, reading a good book, sailing, and so on.
Runner’s High A higher state of consciousness that sometimes occurs during endurance running is known as runner’s high. It has to some extent similar phenomenological features as flow and samadhi experiences. This is understandable because, like some forms of meditation, endurance running is associated with highly regular, long-lasting rhythmic patterns of action and breathing. And like typical flow-producing activities, it is challenging but not anxiety arousing, and it involves physical activity where awareness and action can become merged together. In runner’s high, reflective or analytical thoughts disappear and subjective experience becomes immersed in the here and now. Intense feelings of pure happiness, timelessness, unity with nature, inner harmony, boundless energy, and floating may emerge. At the same time, there is reduced awareness of one’s surroundings and reduced sensitivity to bodily discomfort or pain. A similar state may emerge also in connection of other types of endurance training. Lucid Dreaming Lucid dreaming is primarily a higher cognitive state of consciousness. Sleep laboratory studies have confirmed that lucid dreaming takes place during REM sleep. The defining feature of lucidity is the cognitive realization or reflective consciousness of the fact that ‘This is a dream!’ When this realization takes place, the dream changes from an ordinary one to a lucid dream, and lucidity lasts as long as the dreamer is aware of the fact that he or she is dreaming. Lucidity is like an awakening within the dream, possessing the revelatory knowledge that the whole world around me right now is unreal or hallucinatory, none of the objects or persons around me really exist, they are mere inventions of my dreaming mind. Although the defining feature of this higher state is cognitive in that it constitutes of possessing knowledge that one ordinarily does not have during dreaming (lucid dreams have also been called ‘dreams of knowledge’), there are also attentional and emotional components to this state. Once lucidity ensues, the dreamer can deliberately pay attention to features of the dream world, make deliberate plans of action and carry them out within the dream, or explicitly recall the facts of waking life from long-term memory.
Altered and Exceptional States of Consciousness
The ability to carry out deliberate and even preplanned actions was the key to the laboratory studies in which it was shown that lucidity occurs during continuous REM sleep; no disruption of sleep or brief awakening is involved. Highly trained lucid dreamers are able to give preplanned eye-movement signals in the dream when lucid. The eye-movement recordings show that these incontestable objective signs of lucidity are clearly recognizable and that the EEG at the same time was typical of uninterrupted REM sleep. The heightened attentional state in lucid dreaming is often accompanied by a heightened realism of the dream world where the sensory–perceptual features seem almost unnaturally vivid, clear, radiating, and beautiful. Emotionally, lucid dreaming is often characterized by a positive tone, a feeling of full control, freedom and well-being, sometimes even elation. Although many people may have been briefly lucid during dreaming, for most people lucidity happens only very rarely, if ever. In dream samples, lucidity occurs on average only in one or a few dream reports out of a hundred. Only about 20% of people report having lucid dreams at least once per month. However, lucidity is a learnable skill and various training programs and tips exist that increase the probability of becoming lucid. These include frequent reality testing while awake (asking yourself frequently during the day: Is this a dream?), paying attention to impossible oddities and bizarre features of the dream world which reveal that it must be a dream, and reminding oneself before going to sleep of the intention to become lucid. Also technical devices exist that give signals (a flashing red light, for example) to the dreamer in REM sleep that are supposed to be perceived within the dream without waking up the dreamer. The signal is expected to intrude into the dream and be noticed by the dreamer (why is the world suddenly flashing in red?). This is supposed to immediately lead to the realization by the dreamer that ‘This is a dream!’ Exceptional and Mystical Experiences Out-of-body experiences (OBEs) OBE is an experience where the subject has a visual perspective or spatial location, which
17
seems to be outside the subject’s physical body. The thinking, acting, and perceiving subject or self seems to have left its physical body behind, and may see its body from the outside, usually from above. The subject often feels that the perceptual environment seen in this state is identical with the actual environment. The subject may feel that although the physical body has been left behind, he or she still possesses some kind of ghostly body. In the old parapsychological literature this is known as the ‘astral body.’ In some cases the subject has no clear body image at all, but constitutes a vague cloud or only a formless point of view. About 15%–20% of people report having experienced an OBE. In most cases, OBEs occur when the person is lying down but apparently in the waking state rather than sleeping. OBEs may occur at any time and under any circumstances, however, also during intense physical and mental activity, and sometimes in response to life-threatening situations. An OBE usually lasts for a few seconds to a minute. OBEs often have features that are similar to other higher and mystical states of consciousness. The subject may have the impression of being able to see distant events, or to be able to travel at will to any place. A sense of freedom and control reminiscent of lucid dreaming may occur, as well as feelings of exhilaration or elation, resembling mystical experiences. People typically interpret OBEs as evidence that something – a spirit or a soul – actually did leave the body during the experience. There is little objective evidence that this would ever have been the case. Experiments where the OBE subject’s task has been to retrieve some otherwise inaccessible information from the world (e.g., a number written on a piece of paper and placed out of ordinary sight) while out of the body have not produced any convincing results. Cognitive and neuropsychological theories try to explain the phenomenon by referring to hallucinatory dissociations between visual perspective and body image. In recent studies, OBEs and other similar distortions of body image and visuospatial perspective have been correlated with and induced by activity in particular cortical areas (the temporoparietal junction). Thus, one explanation for
18
Altered and Exceptional States of Consciousness
OBEs may be a temporary failure to bind the body image and the visuospatial representation of the world coherently together in the temporoparietal cortex. Near-death experiences (NDEs) Near-death experiences occur when a person’s life is physically threatened (e.g., cardiac arrest, drowning), when the person perceives that death is imminent even without or before any fatal physical damage (e.g., falling from a height), and sometimes in connection of nonlife threatening events (e.g., general anesthesia). The core features of typical NDEs are, in the order in which they are usually experienced: (1) peacefulness and weightlessness, (2) an OBE, (3) a dark tunnel into which the subject is drawn and through which the subject feels moving, (4) seeing a brilliant light at the end of the tunnel, and (5) entering the light or another world at the end of the tunnel. This last stage may be associated with meeting dead relatives, religious figures, and with a review of one’s life. Also, at this stage the experience becomes very difficult to describe, reminiscent of other mystical experiences. The estimations of the incidence of NDE in people who have become near to death vary from 10% to 50%. Among people who have experienced NDE, most have reported only the first stage of feeling peacefulness (60%). Only 10% report proceeding through to the stage of entering the light or an otherworldly realm. The core content of NDEs is remarkably similar across cultures, times, and different study populations, although only few subjects have experienced all the typical features of NDE. Age, sex, personality, or religious beliefs do not separate people who have had NDEs from those who have not, even though being equally close to death. Explanations of NDE can be roughly divided to supernatural (dualistic) and natural (physiological, psychological and neurocognitive). According to the first type of explanation, which can also be called the afterlife hypothesis, what happens in an NDE is that a nonmaterial soul or self is detached from the body, it travels through the tunnel into another spiritual realm where it is met with deceased relatives, and angelic or godlike being or beings radiating unconditional love. There, the person’s life is reviewed like a film and
some sort of self-judgment takes place, as well as a decision whether to go back to earthly life. After returning, the subjects themselves often feel profoundly transformed and regard the afterlife hypothesis as a self-evident explanation to their experience. According to the naturalistic explanations, also called the dying brain hypothesis, changes in physiological processes and brain function can account for NDEs. First, the feeling of peacefulness, positive emotion and bliss could be brought about by increased endorphin release in the brain under stress. Endorphins may also trigger abnormal or seizure-like activity in the temporal lobe. Epileptic seizures and direct stimulation of the temporal lobe (or the temporoparietal junction) may induce a variety of anomalous experiences, such as OBEs, distortions of body image, realistic memory images, and feelings of the sense of presence of some other conscious being. Anoxia (lack of oxygen) of the brain might lead to the release of cortical inhibition that is known to induce visual hallucinations in other conditions (e.g., drugs, neurological damage of visual pathways). Tunnels are one of four most common types of visual forms typically experienced when visual hallucinations are induced by drugs, seizures, or other causes. It goes without saying that the ‘afterlife hypothesis’ is impossible to integrate together with the current worldview of science. The naturalistic explanations are able to account for many of the core features of NDEs, but only by using speculative and indirect evidence, leaving many open questions. There is no direct evidence that during NDEs the hypothesized physiological or neurocognitive mechanisms would actually be at work and would therefore correlate with specific aspects of the experience or cause them. Thus, we have no direct evidence during NDEs of increased endorphin levels, cerebral anoxia, or seizure-like cortical activity in the occipital, parietal, and temporal lobes. This is not to say that such things do not occur, only that it is extremely difficult to get direct measures of them while a person is having an NDE. Another feature that may be difficult to explain by referring to abnormal or pathological brain activity is the well-organized nature and relative universality and uniformity of NDE. Hallucinations induced by epileptic seizures in
Altered and Exceptional States of Consciousness
the temporal lobe, drug states, or dreamlike states have enormous variability of experiential content both within and between subjects. A uniform and seemingly well-organized experience such as NDE would seem to be based on some mechanism that is widely shared and activated in a roughly similar manner and order in different people, rather than by a variety of processes running wild in a brain under high metabolic stress, very low arousal (unawareness and unresponsiveness to the external world), and burdened by pathological electrophysiological seizures. Thus, for the time being we have to admit that there is no explanatory model that would satisfactorily predict the existence or occurrence of NDEs and that would account for the remarkably universal and seemingly well-organized phenomenological features of NDE. Mystical experiences Mystical experiences are perhaps the ‘highest’ of all the higher states of consciousness. They involve many similar features as some of the other higher states, but in an extreme form. Also the effects of mystical states on the subsequent life of the person are often deep and long-lasting. Such experiences, even if relatively brief, are vividly recalled for years and they may be regarded as among the most significant moments of life. William James, the father of American psychology, regarded mystical states as closely related to personal religious experience. Mystical states are difficult to describe in words or communicate to other people. James took this feature, ineffability, as one of the defining features of mystical states. Mystical states involve both emotional and cognitive components. Emotionally, mystical states are intensely positive, involving overwhelming feelings of peace, calmness, harmony, joy, love, elation, awe, or bliss. Cognitively, mystical states seem to communicate highly significant information for the subject about the true nature of the world, revealing the underlying, hidden order of the universe and its guiding principles. Perceptually, mystical states may involve unusual visions or other forms of imagery, or seeing the ordinary perceptual world as unusually bright, clear, radiant, and beautiful. Mystical experiences are characterized by a sense of heightened
19
reality and significance, and the sense of time may be distorted. The experiences are usually brief, from a few seconds to one hour at most, but their aftereffects may last throughout life. They happen unexpectedly and suddenly and cannot be summoned by will, although certain practices (such as yoga or meditation) or drugs (hallucinogens) enhance the likelihood of their occurrence. Cosmic consciousness is a term introduced by Canadian psychiatrist R.M. Bucke in early 1900s to describe a paradigmatic mystical experience (quoted by William James in Varieties of Religious Experience): The prime characteristic of cosmic consciousness is a consciousness of the cosmos, that is, of the life and order of the universe. Along with the consciousness of the cosmos there occurs an intellectual enlightenment which alone would place the individual on a new plane of existence – would make him almost a member of a new species. To this is added a state of moral exaltation, an indescribable feeling of elevation, elation, and joyousness, and a quickening of the moral sense, which is fully as striking, and more important than is the enhanced intellectual power. With these come what may be called a sense of immortality, a consciousness of eternal life, not a conviction that he shall have this, but the consciousness that he has it already.
Cosmic consciousness entails a widening of consciousness to encompass the entire universe and its deeper working principles. Although such insights are experienced as being absolute truths by the subject of the experience, outsiders may remain doubtful, and rightfully so. The conviction of the subject and the felt authority of the experience are no guarantee of that the insights gained during the mystical state of consciousness carry any truth or validity in the objective sense. Enlightenment is an ultimate form of mystical (and religious) experience, and an ultimate or highest conceivable state of consciousness, usually associated with Eastern religions such as Buddhism. Enlightenment is an experience where one reaches, through meditative practices, complete and total understanding of the nature of reality, and of the nature of oneself in relation to reality. In those traditions, the terms ‘bodhi’ and ‘budh’ refer to awakening, wisdom, and brightness; thus ‘Buddha’ literally means ‘the awakened one.’ Enlightenment thus is a mystical experience that
20
Altered and Exceptional States of Consciousness
awakens ordinary consciousness into seeing the true nature of reality and thereby transforms consciousness into a qualitatively different, higher form that transcends normal consciousness, perhaps in a somewhat similar manner as becoming lucid reveals the true nature of the dream world to the dreamer whose conscious state thereby transcends the ordinary dreaming mind. According to Buddhist thought, enlightenment entails the cessation of all selfish desires and all clinging to material possessions, sensory pleasures, human relationships, and other external passing things. The true nature of everything is seen to consist in impermanence and emptiness; thus even one’s own self is seen to be a mere illusion. The meditative state samadhi, discussed above, involves the mystical union of subject and object, or disappearance of self, an important step toward full enlightenment. These revelatory insights and experiences are supposed to bring about an absolute emotional calmness, peace of mind, cessation of suffering, and deep compassion and unconditional love for all the unenlightened beings who continue to suffer. It is unclear, though, whether enlightenment once achieved subsequently persists, or can be lost and again regained later.
Altered and exceptional states of consciousness reveal the richness of different forms and varieties of our subjective existence. Any theory of consciousness should be able to explain not only the typical features and mechanisms of normal waking consciousness, such as sensory and perceptual experiences in response to physical stimuli, but also the features and mechanisms of altered states. This may be a challenge to the scientific study of consciousness, because many altered states are difficult or impossible to control experimentally, they are highly subjective in that their occurrence or content is impossible to verify by outsiders, and sometimes their precise experiential nature is impossible to describe verbally. Nonetheless, there is growing evidence from neurocognitive studies of ASCs such as dreaming, OBEs, and hypnotic hallucinations that ASCs are real in the sense that they have specific, objectively measurable neural correlates and mechanisms in the brain. Brain stimulation studies of the temporal lobe have furthermore established that OBEs and even mystical experiences can be triggered by simply stimulating the brain in the appropriate location. There is thus some hope that even the most mysterious of ASCs are not entirely beyond the reach of scientific experimentation.
Summary
See also: The Neurochemistry of Consciousness; Psychoactive Drugs and Alterations to Consciousness; Religious Experience: Psychology and Neurology.
ASCs can be defined as temporary states of consciousness when the relationship between inner conscious experience and the world outside of consciousness has been changed. Thus, in altered states, the patterns of thought and the sensory–perceptual and emotional experiences in some way represent the world or ourselves in a way that does not match with the actual state of affairs. Often, though not always, ASCs also involve highly unusual patterns of thought and experience. Furthermore, we usually are able to recognize when we are in an ASC and when not, or at least after returning to the normal state we can infer that we were in an altered state. Exceptional or higher states are a variety of altered states that involve very positive, desirable and insightful experiences that are felt to be personally deeply meaningful, sometimes leading to profound and long-lasting transformations of personal beliefs and experiences afterward.
Suggested Readings Blackmore SJ (1992) Beyond the Body. An Investigation of Out-of-the-Body Experiences. Chicago: Academy Chicago Publishers. Blackmore SJ (1993) Dying to Live: Near Death Experiences. Buffalo, NY: Prometheus Books. Bu¨nning S and Blanke O (2005) The out-of-body experience: Precipitating factors and neural correlates. Progress in Brain Research 150: 331–350. Cheyne JA, Rueffer SD, and Newby-Clark IR (1999) Hypnagogic and hypnopompic hallucinations during sleep paralysis: Neurological and cultural construction of the night-mare. Consciousness and Cognition 8(3): 319–337. Farthing WG (1992) The Psychology of Consciousness. New York: Prentice Hall. Green C and McCreery C (1994) Lucid Dreaming. The Paradox of Consciousness During Sleep. London: Routledge.
Altered and Exceptional States of Consciousness Hobson JA (2001) The Dream Drugstore. Chemically Altered States of Consciousness. Cambridge, MA: MIT Press. James W (1902) The Varieties of Religious Experience. New York: Longman, Green. Kallio S and Revonsuo A (2003) Hypnotic phenomena and altered states of consciousness: A multilevel framework of description and explanation. Contemporary Hypnosis 20(3): 111–164. LaBerge S (1985) Lucid Dreaming. New York: Ballantine. Mahowald MW and Schenck CH (1992) Dissociated states of wakefulness and sleep. Neurology 42(supplement 6): 44–52.
21
Mavromatis A (1987) Hypnagogia. The Unique State of Consciousness between Wakefulness and Sleep. London: Routledge. Strauch I and Meier B (1996) In Search of Dreams. Results of Experimental Dream Research. New York: SUNY Press. Tart CT (ed.) (1990) Altered States of Consciousness. New York: Harper Collins. Vaitl D, Birbaumer N, Gruzelier J, et al. (2005) Psychobiology of altered states of consciousness. Psychological Bulletin 131(1): 98–127.
Biographical Sketch
Antti Revonsuo is a professor of psychology at the University of Turku, Finland, and a visiting professor of cognitive neuroscience at the University of Sko¨vde, Sweden. He is an associate editor of the journal Consciousness and Cognition, the director of the consciousness research group at the Center for Cognitive Neuroscience, University of Turku, and a founding member of the faculty of an undergraduate degree program on consciousness studies at the University of Sko¨vde, Sweden. He has done research on consciousness since the early 1990s, focusing mostly on dreaming, hypnosis, visual consciousness and its neural correlates, and theories of consciousness. His major publications include a new theory of the function of dreaming, the threat simulation theory, published in Behavioral and Brain Sciences in 2000, a theory of hypnosis as an altered state of consciousness (together with S. Kallio) in Contemporary Hypnosis in 2003, and an overall theoretical framework for the explanation of consciousness, published in the book Inner Presence: Consciousness as a Biological Phenomenon (MIT Press, 2006).
Animal Consciousness D B Edelman, The Neurosciences Institute, San Diego, CA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Accurate report – A first-person account, through language or other types of voluntary response, of what an individual is experiencing. Accurate report has provided a critical means of investigating conscious states in humans. In the case of nonhuman animals, certain motor channels other than language might also be used to gauge the ability to make higher-order discriminations that might bear on a given animal’s conscious states. Behaviorism – A school of psychology that emerged in the nineteenth century and was developed and famously elaborated upon by John Watson and B.F. Skinner in the early part of the twentieth century. Behaviorism takes the position that everything an animal does can be reduced to behaviors, and, in fact, the only legitimate object of psychological study is behavior. In the behaviorist framework, no recourse to internal, or mental, states is considered necessary or even legitimate. Cephalopods – The class of marine mollusks that includes octopuses, squid, cuttlefish, and nautiloids. Distinguishing features of the cephalopods include highly prehensile tentacles (eight in the case of octopuses; ten in the case of cuttlefish and squid); very large eyes (octopuses, cuttlefish, and squid have single compartment, camera eyes with lenses that closely resemble those of vertebrates); chromatophore organs in their skin, which allow adaptive camouflage; and, in many species, ink sacs for defense or escape. Commentary key – A behavioral index that enables an animal to indicate a choice of stimuli in a given task and at the same report the nature and/or contents of its conscious
experience during the performance of that task. The ‘commentary key’ paradigm was used effectively by Alan Cowey and Petra Stoerig to show that rhesus macaques with lesions to one-half of the striate cortex (V1) behaved very much like blindsighted humans in their responses to – and likely experiences of – stimuli presented in intact and occluded visual hemifields. Decussation – The X-shaped crossing of nerve fibers to opposite sides of the brain or spinal cord. In mammals with stereoscopic vision, the optic tract of each eye is partially decussated to both sides of the brain. In many species of birds, the optic tract of each eye is almost completely decussated to one side of the brain. Dynamic core – The neural basis of consciousness in mammals as proposed by Gerald Edelman and Giulio Tononi. The populations of neurons in the thalamus and the cortex that function as coherent groups, through richly degenerate – or reentrant – connections, which are proposed to underlie conscious states. The reentrant connectivity of the thalamocortical system is considered by Edelman and Tononi to be uniquely capable of generating the complex dynamic properties of conscious states. Higher-order consciousness – An explicit awareness of self. A frame of reference that seemingly encompasses past, present, and future scenes. Higher-order conscious states reify the contents of primary consciousness as objects. Metacognition – Thinking about thinking. Awareness of one’s own thoughts. The ability to monitor one’s own cognitive states. Metacognition involves both conscious and unconscious knowledge.
23
24
Animal Consciousness
Phylogeny – The evolutionary history of a group (e.g., phyla, genus, species) of organisms and/or their constituent characters or traits. Primary (sensory) consciousness – Awareness of a multimodal scene. Unification of percepts into episodic scenes. Telencephalon – The anterior forebrain in birds and mammals. The cerebral hemispheres and parts of the hypothalamus are derived from the telencephalon during brain development. Thalamocortical – The densely reentrant circuitry that connects the thalamus and cortex in mammals and possibly in birds. Complex dynamic neural activity in the thalamocortical system is believed by Edelman and Tononi to underlie conscious states.
Introduction There is much in our daily experience, as well as numerous anecdotal reports and indirect evidence from cognitive studies, that suggests that human beings are not the only animals capable of conscious states. We are, however, the only animals naturally capable of relating our conscious experiences via linguistic reports, and therefore the only animals for which there is an increasingly robust body of evidence for conscious states in a variety of forms. In the absence of accurate linguistic report, other means must be sought for establishing that certain nonhuman animals are conscious.
Defining Consciousness Across Phylogeny Critical to the investigation of animal consciousness is both an adequate definition of consciousness as it might manifest itself in nonhuman species and the recognition that consciousness is a process that is instantiated within biological substrates, that is, nervous systems, and as such, has therefore been subject to evolutionary forces.
Sensory- versus Higher-Order Consciousness: A Critical Distinction First, a necessary distinction must be drawn between sensory, or primary, consciousness – a capacity that might be common across a wide swath of phyla – and higher-order consciousness, which may be unique to Homo sapiens and perhaps a small group of intensively trained nonhuman animals, including some great apes and certain birds such as parrots, which are capable of sophisticated linguistic performance, and, in the case of the latter, both vocal learning and, apparently, accurate linguistic report. A variety of claims have been made that certain animals are self-aware since Gordon Gallup’s demonstration in 1970 that chimpanzees that had been anesthetized and marked with two red dots – a control dot on an inaccessible region of the animals’ bodies and a red dot on the forehead – would, when awakened, behave in a manner consistent with self-recognition if confronted with their reflection in a mirror (the control dot was ignored, indicating that sensory cues other than visual recognition of the dot on the forehead were not involved in the animals’ reactions). Although elaborations of Gallup’s ‘mirror test’ have been performed on a number of different species with varying degrees of success (all great apes, dolphins, killer whales, Asian elephants, and certain birds have been reported to show signs of self-recognition, but, interestingly, certain old world monkeys have not), the results of this experimental methodology remain difficult to interpret and somewhat controversial. Among the more trenchant criticisms of the mirror test is the contention that in animals in which vision is not as well elaborated as other sensory modalities (e.g., dogs, which may often rely more heavily on olfaction), visual self-recognition might not be expected, as it is in animals with acute stereoscopic color vision, such as humans. Given the controversial nature of the mirror test as well as the very nascent state of affairs in the development of a rigorous framework for the scientific study of animal consciousness, perhaps it is best to begin by investigating sensory consciousness in nonhuman species. Sensory consciousness may be defined as the construction of a unified perceptual ‘scene,’ that is, the integration of input from disparate sensory
Animal Consciousness
modalities into a seamless whole. Higher-order consciousness, or metacognition, is the awareness of a self that is embedded within a unified perceptual scene; it is the awareness of being aware. Arguably, higher-order consciousness is contingent upon language, which may be unique to Homo sapiens and perhaps some laboratory-trained animals. Sensory consciousness is contingent upon an ongoing link between perception, across disparate sensory modalities, and memory; in this regard, many mammals and some birds may be said to have sensory consciousness. For the purposes of the present discussion, references to animal consciousness will be ostensibly in regard to sensory consciousness. Consciousness as a Product of Natural Selection More than 125 years ago, Charles Darwin cast the emergence of the human mind clearly and unambiguously in the context of an evolutionary continuum, with roots planted deeply within nonhuman animal lineages. To Darwin, the human mind was embedded in the natural world, not something separate and apart from all other organic structures, and therefore a product of natural selection. Indeed, Darwin famously took Alfred Russell Wallace, the codiscoverer of natural selection, to task for Wallace’s insistence that the human mind resided outside of the evolutionary continuum. Unlike Darwin’s theory of natural selection, which took its place as the founding principle of modern biology with the new evolutionary synthesis of the 1940s, the notion of the animal mind, either as a product of natural selection or otherwise, languished throughout much of the twentieth century, particularly after John Watson, the founder of the American behaviorist school, dismissed it as irrelevant to the task of understanding animal – and ultimately human – behavior. The ascendancy of the radical behaviorism represented by Watson’s turn away from internal mental states and B.F. Skinner’s stringent behaviorist experimental paradigm influenced generations of academic psychologists and philosophers, and effectively quashed the rigorous study of animal consciousness for much of the twentieth century by
25
erecting a sort of early functionalist perspective in which external, or behavioral, manifestations of the reactions of animal nervous systems occupied central importance to the exclusion of all else. For much of the twentieth century, few dared to challenge this orthodoxy (one early exception was the biologist Donald Griffin, who held that there was an obvious continuum between the cognitive capabilities of nonhuman animals and the human mind). Curiously, although behaviorism has largely been repudiated, many anecdotal observations of the cognitive performance of animals are often dismissed by modern behavioral scientists professing caution, but sounding very much like latter day behaviorists in the tone of their methodological rhetoric. With the emergence of evolutionary psychology, it has largely been accepted that a variety of cognitive capacities are the products of natural selection. Given such recent attempts to naturalize cognition generally, perhaps it is not unreasonable to finally embed the process of consciousness in nature as well.
A Methodological Framework for the Study of Animal Consciousness A number of correlates of consciousness have been identified in humans, including reentrant thalamocortical signaling, a characteristic electroencephalography (EEG) signature comprising fast, irregular, low amplitude electrical activity ranging from 12 to 70 Hz, and widespread brain activity correlated with conscious contents. Using the human case as a benchmark and reference, it is now possible to search for such correlates of consciousness in nonhuman species. Human Consciousness Studies as Benchmarks for Investigating Animal Consciousness Critical to addressing the question of animal consciousness is the establishment of benchmarks based on the only reliable, reproducible source for data: studies of awake, behaving humans. Such data come from humans reporting their conscious states in a wide variety of experimental paradigms, often involving noninvasive imaging techniques.
26
Animal Consciousness
Among the most definitive examples are binocular rivalry tasks performed by humans during magnetoencephalography (MEG), in which unmistakable changes in brain activity were observed when consciousness of an object was reported. These studies generally implicated the thalamocortical system in the generation of conscious states. In addition, evidence from strokes and destruction of brain regions has indicated that structures such as the thalamocortical circuit and the mesencephalic reticular formation are critical neural substrates for consciousness. The identification of such an anatomical benchmark, together with neurophysiological recordings during a task requiring rich discriminatory behavior, might suggest the presence of conscious states in primates and other mammals, even in the absence of something like accurate report. Indeed, a binocular rivalry paradigm has been employed in which monkeys were presented with rivalrous and nonrivalrous stimuli, and, during rivalry, the majority of neurons (>90%) recorded in the inferior temporal cortex (IT) fired in response to the perceived stimulus; the monkeys were trained to report the stimuli they perceived by pulling a lever. The IT neurons were not responsive to the nonperceived stimulus. The results of this study suggest that the monkeys were probably conscious of certain visual stimuli. A theoretical proposal to account for the properties and evolution of consciousness has been elaborated over several decades by Gerald Edelman and his colleagues. Central to this proposal is the idea that consciousness in mammalian species emerged when reentrant connectivity evolved between brain areas for perception and those involved in memory. Evidence for such reentry in humans has been demonstrated using MEG. Also notable is the distinction this proposal makes between two varieties of consciousness: primary consciousness, in which percepts are merged into episodic scenes, each of which is of a piece; and higher-order consciousness, which involves selfawareness, the ability to reconstruct past scenes and formulate future scenes, and in humans the ability to represent both the internal state and the external world symbolically through language or other means.
Types of Evidence for Animal Consciousness In the absence of accurate verbal reports, how can we assess the conscious states of different animals? Although human language is by far the most elaborate mode of communication known, other sensorimotor channels are sufficiently complex that they might be exploited as means of gauging what a given animal is or is not aware of. Indeed, the ‘commentary key’ paradigm laid out by Alan Cowey and Petra Stoerig in their study of blindsight in rhesus macaques is one example of how certain sensorimotor channels might be exploited as forms of accurate behavioral report in certain nonlinguistic species. Cowey and Stoerig devised an ingenious experiment in which rhesus macaques with lesions to one-half of the striate cortex (V1) were trained to touch the area of a video display where stimuli would appear. These monkeys were able to learn to detect stimuli presented in their occluded visual hemifields, that is, those hemifields rendered ‘cortically blind’ by lesions in the contralateral half of V1. They could even make discriminations across stimuli presented in their occluded hemifields. However, the monkeys could not distinguish between stimuli presented in their occluded hemifields and blank regions of the video display (containing no stimuli) presented in their unaffected hemifields. The foregoing result could be interpreted as a report by the monkeys that they were not aware of the difference between the two hemifields. The extension of Cowey and Stoerig’s ‘commentary key’ paradigm to a variety of sensorimotor channels might provide a means of investigating consciousness in a number of nonhuman species. Moreover, the human benchmark could provide sufficient neuroanatomical and neurophysiological neural correlates for conscious states that might also be investigated in species far from the human, or even mammalian, lineages. Proceeding from the three principal properties of human consciousness mentioned earlier, it is clear that evidence for consciousness in nonhuman species might be gathered from three areas of study: neuroanatomy, neurophysiology, and behavior.
Animal Consciousness
Evidence for Consciousness in Nonhuman Species The broadly conserved architecture of the mammalian brain as well as ample evidence of complex behavioral repertoires are at least consistent with the possibility of conscious states in many mammals. But the recent revision of avian neuroanatomy supports a substantial structural homology to the mammalian brain. Moreover, reports of feats of avian memory, deception, vocal learning and reproduction, the capacity of some species to employ lexical terms in meaningful ways, and finally the ability of certain birds to make higherorder discriminations all suggest the elaboration of a nervous system as sophisticated as those of most mammals. In fact, the foregoing discussion would suggest that consciousness emerged in birds, probably independently of mammals. Birds and Mammals: Evolutionary Homology Lost and Found Birds exhibit a broad behavioral repertoire, which includes simple nest building in swallows, manufacture and use of tools by New Caledonian crows, food caching by scrub jays and other birds in perhaps a hundred or more sites (which certainly involved complex spatial memory capacity), seasonal migrations of thousands of miles, sophisticated song learning and production, stunningly accurate mimicry of natural and artificial sounds, and, in some birds, word comprehension, production, and naming. This last series of behaviors is based on the capacity for vocal learning, a capacity shared by perhaps seven animal groups: humans, cetaceans, bats, parrots, songbirds, hummingbirds, elephants, and perhaps even some rodents. Vocal learning, which is richly manifest in a number of forms in birds, might allow us the possibility to test for avian conscious states. In particular, song production in songbirds might act as a channel for something like accurate report, given an appropriately designed experiment. The studies of blindsight in lesioned monkeys by Cowey and Stoerig suggest a possible model for an experiment in which one-half of a bird’s visual cortex is lesioned, and the bird is then queried as to
27
whether it ‘sees’ a particular stimulus in the affected visual hemifield. Its response to a given stimulus, or lack thereof, would be through a vocalization previously entrained to that stimulus. Lesioning the ectostriatum (the avian brain area analogous to V1 in the mammalian tectofugal pathway) might disrupt early visual processing and therefore might lead to a form of avian blindsight. But it should be noted as well that the optic tracts of many birds are nearly completely decussated (>99%) to opposite hemispheres of the brain, and therefore in many birds would preclude the same type of hemifield arrangements found in monkeys and humans. Another approach might be an electrophysiological paradigm based on the finding by Nikos Logothetis that the majority of neurons in a behaving monkey’s IT fired in response to the percept (as defined by a behavioral report), while neurons in V1 responded to the signal. Such an experiment in birds would of course be contingent on identifying neurons functionally analogous to those found in the mammalian IT and striate (or visual) cortex. Some of the difficulties inherent in the experimental approaches suggested above might be obviated if the bird being tested is an African grey parrot capable of reproducing human speech. Irene Pepperberg and her colleagues have shown that African grey parrots are capable of performing naming tasks after acquiring vocabularies roughly equivalent to those of some of the chimpanzees employed in language-training projects over the past four decades. Notably, these chimpanzees had learned a large number of lexical terms after many years of rigorous training and reinforcement. By far, the most impressive nonhuman symbolic capacity was shown in captive bonobos by Sue Savage-Rumbaugh. One individual, Kanzi, acquired several hundred lexical terms and responded to requests or queries through the use of a keyboard containing arbitrary symbols. Although the concept of accurate report in humans or something analogous in monkeys (e.g., the commentary key paradigm developed by Cowey and Stoerig and the studies of the IT and striate cortex by Logothetis) was not explicitly invoked in the studies by Pepperberg and her colleagues of naming and categorization in African
28
Animal Consciousness
grey parrots, by naming objects, these animals appeared to be producing accurate reports of the discriminations they were capable of making. Moreover, Alex, the principal subject of many of Pepperberg’s experiments, seemed to be able to make a judgment to the effect that, ‘‘I know that something in this perceptual scene has changed, and here is what has changed,’’ when presented with an altered array of objects, suggesting the ability to make discriminations about discriminations. Although such a capability may in fact indicate the presence of higher-order consciousness, and indeed it has been suggested by others that such metacognition necessarily underlies animal consciousness, Pepperberg’s more modest conclusion – that parrots may possess primary or ‘perceptual’ consciousness – for now seems to be a more readily justifiable conclusion. The overall organization of the vertebrate central nervous system is part of a highly conserved body plan that emerged with the earliest chordates more than 500 million years ago. Identifying homologous neural structures is often not difficult because developmentally, many brain structures in lower vertebrates (i.e., reptiles and lizards), birds, and mammals can be traced to common origins in specific embryological tissues. In fact, there is substantial evidence from comparative embryological studies that suggests that much of the neuronal properties and circuitry that underlie the mammalian cortex were established within nucleated or clustered arrangements long before the evolutionary appearance of the distinct six-layered mammalian cortical mantle. In addition, the somatomotor circuitry of the avian dorsal pallium may be homologous to the mammalian basal ganglia–cortico–thalamic loop. Yet there remains little agreement as to which avian neural structure is the homologue of the mammalian isocortex. Indeed, among the most distinctive gross anatomical properties of the avian brains is the division of the telencephalon into structures resembling nuclei that lack the laminated cortical mantle characteristic of mammalian brains. The avian optic tectum and cerebellum are more elaborated than their mammalian homologues. More basal regions, such as the hypothalamus and preoptic area, are relatively easy to recognize. Mid-brain regions such as the amygdala and hippocampus cannot be identified easily through
structural similarities with mammalian homologues. In such cases, deeper homologies must be sought through investigation of the functional properties of neuronal populations within particular regions. For example, the claim that the avian anterior forebrain pathway is functionally analogous to the mammalian corticobasal ganglia–thalamocortical loop is supported both by observations that the medial nucleus of the avian dorsolateral thalamus (DLM) contains inhibitory as well as excitatory pathways and that neurons in the DLM exhibit functional properties similar to those of thalamocortical neurons. But the use of electrophysiological techniques to establish functional homology, while potentially fruitful, sometimes does not yield results easily. Electrophysiological studies have established common properties of mammalian thalamocortical neurons, such as low-threshold calcium (Ca2+) spikes and slow oscillations, but such studies have yet been adequately reproduced in birds. In instances where neurophysiological means are intractable or unavailable, molecular or histological techniques may be critical in the identification of certain regions of the avian brain. Molecular markers such as neurotransmitters, neuropeptides, and receptors that are specific to certain cells known to reside in particular regions of the mammalian brain can be targeted in avian brain regions by immunohistological techniques, as has been shown by Erich Jarvis and his colleagues. Gene expression patterns too can be compared in developing bird and mammalian brains. Such a comparison of homeotic genes involved in early brain development in chick and mouse embryos has indicated clear structural homology between parts of avian telencephalon and mammalian cortex. The evolutionary conservation of ancient vertebrate neurochemical systems has also been useful in the identification of avian brain regions homologous to known mammalian neural structures. In one instance, the identity of certain avian brain structures as components of the limbic system was established through the use of an antibody to the steroid metabolizing enzyme estrogen synthetase (aromatase), which in mammals is known to be specific to certain cells in the limbic system. Other molecular techniques, such as cloning and in situ hybridization, have been useful in the establishment of avian homologues of certain mammalian neural structures. For example, large numbers of glutamate
Animal Consciousness
receptors expressed in the mammalian CNS were successfully cloned from regions of songbird brain, and in situ hybridizations in mouse and zebra finch brains using probes based on these clones showed patterns of glutamate receptor expression that were highly conserved in cerebellum, midbrain, thalamus, and basal ganglia. Interestingly, on the basis of these findings, Kazuhiro Wada and his colleagues argue that avian and mammalian brains have become functionally similar through evolutionary convergence. Gross differences in topological organization, they point out, have led others to relegate the avian brain mistakenly to a lower functional status. Lower vertebrates, birds, and mammals have all been shown to possess certain homologous excitatory and inhibitory systems, including the serotonergic and GABAergic systems, and, in the case of birds and mammals, the dopaminergic system. Although neural structures have been shown to be highly conserved across the vertebrates, gross anatomical and histological data alone have been insufficient to establish this conservation. Where structural homologues have been difficult to establish through anatomy and histology, a variety of different techniques have been applied. For example, parsimony-based phylogenetic analyses of comparative embryological data, lesioning and behavioral response experiments, and comparative studies of gene expression patterns during development have yielded the identity of structures such as the amygdala and hippocampus in widely different nonmammalian vertebrates, including lampreys, goldfish, and birds. The extremely ancient origin of the hippocampus was established through comparative histological studies, which showed that the primordium that gives rise to the hippocampus in mammals is also present in the developing lamprey. Finally, comparisons of gene expression data have shown that the organization of the vertebrate brain was largely established in the earliest vertebrate ancestor. An increasingly sophisticated armamentarium of molecular techniques is enabling researchers to forge links between avian neuroanatomy and well-characterized regions of the mammalian brain. Such an approach may be instrumental in identifying structures that support conscious states in nonmammalian vertebrates in the same way as the thalamocortical circuit in mammals.
29
Although it is uncertain when consciousness appeared among the vertebrates, it seems likely that it emerged independently at least twice, sometime after the divergence, 300 million years ago, of the anapsid and synapsid reptilian lines that led to birds and mammals, respectively. Although it is impossible to say whether animals in either of the reptilian lineages preceding birds and mammals were conscious, it seems conceivable that consciousness could have emerged separately and independently in the two different vertebrate lines – a possibility that may find support as homologous structures or similar analogous arrangements in birds are discovered. The characterization of avian brain structures that are analogous or homologous to the mammalian cortex and thalamus is an important first step in investigating consciousness in birds. The key to identifying conscious states in birds, however, will be the discovery of neurophysiological signatures that resemble those of the mammalian conscious state. One such signature would be patterns of electrical activity that, as Anil Seth and his colleagues suggest, reflect ‘‘. . .widespread, relatively fast, low-amplitude interactions. . .driven by current tasks and conditions.’’ Similarities between the waking EEG patterns of birds and mammals, as well as slow wave electrical activity recorded during sleep in birds (albeit embedded within an overall EEG pattern that is distinctly different than that of mammals), are at least suggestive of generally analogous avian brain functioning that might support conscious states. Clearly, birds possess a number of the necessary substrates and conditions for primary consciousness. It remains for us to design experiments, based on the human benchmark and perhaps be informed by such work as that of Cowey, Stoerig, and others, that will make a convincing case that birds are indeed capable of conscious states. Cephalopods: An Alien Intelligence and a Challenge for Consciousness Research In invertebrate species, in which the organization of the nervous system diverges markedly from those of vertebrates such as birds and mammals, the case for consciousness becomes understandably quite tenuous. Of the many extant invertebrate
30
Animal Consciousness
groups, spiders and cephalopod mollusks would appear to have the most sophisticated behavioral capabilities and, on this basis, might be appropriate candidates to test for properties associated with conscious states. Exhaustive scientific accounts of cephalopod behavior suggest that some of these animals – particularly the octopus – exhibit a measure of behavioral sophistication and plasticity that is in many ways on par with certain higher vertebrates. Although exhaustive evidence has been presented previously by Jennifer Mather to make the case for sensory, or primary, consciousness in certain cephalopods, the overwhelming majority of this evidence has been of a behavioral nature. Behavioral or anatomical evidence suggesting the possibility of conscious states in these animals has not been forthcoming largely because much of the anatomy and physiology relevant to this problem has not yet been worked out. Since it is likely that a key substrate of mammalian conscious states is the reentrant circuitry between the cortex and the thalamus as suggested by Gerald Edelman and Giulio Tononi, it might not be unreasonable to suggest that a similar functional loop between analogous structures for relaying sensory input and those for different kinds of memory is also a critical contingency of consciousness in certain nonmammalian species. Although the technical hurdles represented by the task of demonstrating the functional neuroanatomical and neurophysiological substrates of consciousness in cephalopods are formidable, there is sufficient evidence of sophisticated sensory and cognitive capabilities to suggest that a search for conscious states is warranted. Over many decades, a number of behavioral studies have probed the cognitive capacities of cephalopods. Among the coleoid cephalopods, it is the rich cognitive capacity of octopuses that has been most thoroughly documented. For example, the pioneering work of J.Z. Young, as well as that of Martin Wells and Stuart Sutherland, has amply demonstrated the ability of the octopus to make sophisticated discriminations between different objects based on size, shape, and intensity. Perhaps suggestive of more general properties of learning in complex nervous systems, Sutherland noted that octopuses classify differently shaped objects in the same manner as vertebrates such as goldfish
and rats. More recent work on octopus learning and problem solving, such as the studies of Graziano Fiorito and his colleagues, have involved such tasks as finding the correct path to a reward in a plexiglas maze or retrieving an object from a clear bottle sealed with a plug. The most impressive capabilities of cephalopods have generally reflected an extremely plastic behavioral repertoire and, on the basis of reports of observational learning by Fiorito and Scotto, highly developed attentional and memory capacities. Researchers, including Veronique Agin, Raymond Chichery, and Graziano Fiorito, have documented evidence of distinct capacities for short-term and long-term memory in both the octopus and the cuttlefish. For example, in experiments in which an octopus was exposed to a maze containing obstacles that were changed ad libidum by the researcher, the animal could remember these changes and maneuver around shifting obstacles accordingly. Notably, the authors of this study, Tohru Moriyama and Yukio-Pegio Gunji, asserted that the octopus appeared to first consider the layout of the maze before proceeding. The formidable quality of octopus memory was also shown beautifully by the work of Fiorito and Scotto, who demonstrated the ability of these animals to solve problems through observational learning, and not merely through mimicry. Perhaps the most common task employed by researchers to assess the cognitive capacity of octopuses involves the presentation of a prey species, such as a crab, enclosed in a plugged jar. Commonly, as Fiorito and his colleagues have reported, the octopus’s first reaction to the jar is an immediate attack, as if it is confronting freely moving prey. In short order, after repeated exposure to the crabcontaining jar, the octopus learns to pull the plug and retrieve the crab. Prior exposure to the empty jar does not significantly reduce the time it takes the octopus to retrieve the crab, which might suggest that perhaps there is no transitive component in octopus learning. Moreover, it is notable that, even as the octopus appears to solve the prey-ina-jar problem, it nevertheless shows strong, speciesspecific, predatory behavior – the attack – which persists over repeated trials. Such invariance in the octopus’s behavioral repertoire, even as it solves the prey-in-a-jar problem, suggests the absence of
Animal Consciousness
the kind of inhibitory pathways that might be activated in a mammal’s brain as it becomes familiar with a particular task. Although such lack of inhibition does not necessarily bear on the possible presence or absence of conscious states in octopuses, it does suggest a nervous system markedly different – and perhaps more primitive – functionally than that of a mammal. The question of what the necessary conditions for cephalopod consciousness might be has not yet been addressed explicitly. It is conceivable that, if cephalopod neuroanatomy becomes sufficiently well characterized to identify functional analogues of mammalian visual cortex, it might be possible to design and perform experiments similar to those involving the ‘commentary key’ paradigm employed by Cowey and Stoerig on Rhesus macaques. The cephalopod chromatophore – the organ system comprising pigment saccules surrounded by radial muscles in the skin – is employed by cephalopods for what researchers Roger Hanlon and John Messenger have referred to as ‘adaptive coloration’ (i.e., camouflage, mimicry, sexual display), but could perhaps just as easily be co-opted as a channel for accurate report in such experiments. Perhaps an even more accessible type of psychophysical experiment, such as the presentation of a serial stream of stimuli with an embedded stimulus that can be temporally shifted in or out of perception (the so-called attentional blink), could be designed around the cephalopod chromatophore as a channel for reporting what part of the stimulus stream the animal is, or is not, aware of. One formidable challenge in the design of such an experiment has to do with the necessity of presenting streams of stimuli to animals by video to allow for the kind of rapid switching of stimulus type that would otherwise be impossible in a live presentation. Most cephalopods, including octopuses, are color blind, as John Messenger, Nadav Shashar, Thomas Cronin, and others have noted, and there is increasing evidence that many of these animals actually perceive polarized light and are able to make subtle discriminations on the basis of this capacity. Unfortunately, many video monitors currently available (e.g., LCD and CRT monitors) employ fluorescent backlighting and polarizing filters, both of which might create artifacts that could render video streams of stimuli effectively
31
unintelligible to most cephalopods. While not an insurmountable technical problem by any means, a solution would have to be found before proceeding with a program of ‘cephalopod psychophysics.’ Of all the cephalopods, the octopus has extraordinarily complex sensory receptors coupled to a nervous system that, normalized to body weight, rivals those of some vertebrates (smaller than those of birds and mammals, but larger than those of fish and reptiles, as noted by Roger Hanlon, John Messenger, Binyamin Hochner, Tal Shomrat, and Graziano Fiorito, among others). The brain of an adult octopus may contain between 170 million and 500 million cells, the vast majority of which are neurons. In contrast to the avian brain, which, despite early misprisions of nomenclature, closely resembles the mammalian brain, the organization of the cephalopod nervous system is utterly alien, and presents a profound set of problems for identifying necessary structural correlates of systems underlying consciousness. In this regard, the search for a functional cephalopod analogue of the reentrant thalamocortical circuitry found in mammals will present a major challenge. Where would one begin to search for such circuitry? Biochemical and molecular techniques that have been tremendously useful in identifying functional vertebrate neuroanatomies might be deployed to effectively address this question; certainly, at the level of constituent neurons and receptor cells, the similarities between the cephalopod nervous system and its vertebrate counterpart are clear. Indeed, the resemblance between giant squid axons and synapses and those of vertebrate neurons was an early watershed in modern neurophysiology and allowed many of the most important investigations of neuronal physiology that followed, including those by Alan Hodgkin, Andrew Huxley, Bernard Katz, Ricardo Miledi, and John Zachary (J.Z.) Young. Notwithstanding the profound gaps in our understanding of the functional circuitry of the cephalopod nervous system, some functional properties of neuroanatomy have been established over several decades. In the brain of at least one genus of squid, Loligo, Linda Maddock and J.Z. Young reported that at least 30 distinct nucleus-like lobes have been identified so far. The largest of these lobes, the optic lobes (which contain as many
32
Animal Consciousness
as 65 million neurons in some species of octopus), handle visual processing and memory establishment as well as some higher motor control. Although the optic lobes are not organized like the laminar sheets of the mammalian cortex, it has been suggested by Hanlon and Messenger that these lobes are functionally analogous to the vertebrate forebrain. Starting with the early work of J.Z. Young, the vertical, superior frontal, and inferior frontal lobes were also found to play important roles in the establishment of memories. In experiments by Fiorito and Chichery in which the vertical lobe of Octopus vulgaris was lesioned, the ability of animals to learn visual discriminations was severely impaired, whereas the consolidation of long-term memories remained unaffected. In another study by Wells and Young, when the median inferior frontal lobe was removed, learning was compromised due to memory impairment. If the faculty of recall was, indeed, what was affected in the foregoing studies, this would suggest that some regions of the octopus frontal and vertical lobes are functionally similar to regions of mammalian cortex. Carla Perrone-Capano and her colleagues have noted that considerably more genetic expression occurs in the squid nervous system than in other tissues or organs – an observation that is consistent with the notion that selection for highly complex nervous systems is phylogenetically ancient. It is therefore plausible that complex brains capable of rich behavioral repertoires began evolving in two very different lineages between 530 and 540 million years ago. Perhaps the most daunting quality of the cephalopod nervous system is its radically different overall organization. The peculiar parallel distributed nature of the octopus locomotor system is perhaps the signal exemplar of the alien character of the cephalopod nervous system. J.Z. Young has noted that the density of neurons located in the tentacles of the octopus, taken together, exceeds the total number of neurons in the central fused ganglia of the brain itself. Recent work by German Sumbre, Yoram Gutfreund, Tamar Flash, Graziano Fiorito, and Binyamin Hochner has shown that a detached octopus arm could be made to flail in a realistic manner when stimulated with short electrical pulses. Although the presence of semiautonomous
motor programs is not unique to this species, or even to cephalopods in general, there is ample evidence for the existence of central pattern generators (CPGs) in many invertebrates and vertebrates alike; in the case of vertebrates, CPGs seem to be located in the spinal cord, that is, within the central nervous system itself. Thus, it is not possible to produce the full suite of coordinated movements of locomotion in a detached vertebrate limb. This distinction makes the sophistication of the semiautonomous neural networks in octopus tentacles and their local motor programs all the more striking. Given this observation, notions of embodiment and bodily representation that might bear on the assessment of consciousness might necessarily be different for octopus than those set forth by cognitive scientists and philosophers when considering the vertebrate case. The identification of higher levels of neural organization in cephalopods poses even more profound challenges. Such functional concepts as cell assemblies, modules, cortical columns, thalamocortical loops, cytochrome oxidase-labeled blobs (first characterized in the cortex by Margaret Wong-Riley), and neuronal groups have been copiously and variously defined in mammals as groups of cells that share similar structure and/or function and populate in large numbers certain defined regions of the brain (i.e., cortex, nuclei, and ganglia). Cortical columns, as described more than 50 years ago by Vernon Mountcastle, might be the smallest functional modules of the mammalian neocortex. Esther Leise’s notion of the neural module comprises large concentrations of neuropil in invertebrates, and is perhaps functionally similar to the so-called minicolumn concept elaborated upon by Mountcastle and others. Although there is no experimental validation of such an invertebrate neural module, the concept might nevertheless provide a useful roadmap in the search for functional elements in the cephalopod brain. Certainly, the discovery of such neural modules might indicate an organization of brain circuitry that has functional properties similar to the mammalian cortex. At present, though, there is little insight into the organization of functional maps (perhaps akin to the topographic representations found in the mammalian cerebral cortex) in the cephalopod nervous system.
Animal Consciousness
Since cephalopod brains have been shown to contain many of the same major neurotransmitters that are found in the brains of mammals, it is conceivable that, as is now the case with functional avian neuroanatomy, it will eventually be possible to use immunohistochemistry and genetic manipulation to characterize those areas of the cephalopod brain that are analogous in function to neural regions showing correlated activity during waking, conscious behavior in mammals. Given the success of the cephalopod radiation and its occupation of so many marine niches, it is not surprising that there is substantial interspecies variation in sensorimotor capacities and behavioral repertoires. The depth and breadth of known ecological adaptations, as well as the emergence of large numbers of cells and cell-types and dense enough connectivity in certain species, at the very least suggests entirely plausible and sufficient preconditions for the appearance of the rich discriminatory capacities necessary for consciousness to emerge in some cephalopod species. Although cephalopod neurophysiology has only recently come into its own, perhaps the most suggestive finding in favor of precursor states of consciousness in at least some members of Cephalopoda is the demonstration of EEG patterns, including event-related potentials, that, as researchers Theodore Bullock and Bernd Budelmann have noted, look quite similar to those in awake, conscious vertebrates. This has been demonstrated in at least one other invertebrate, the fruit fly, by Bruno Van Swinderen and Ralph Greenspan. An obvious prerequisite to identifying cephalopod EEG patterns that reflect the signature of fast irregular activity, similar to that observed in human conscious states, will be to determine precisely where to record from. On the basis of earlier studies, it is possible that the optic, vertical, and superior lobes of the octopus brain are relevant candidates and that they may function in a manner analogous to mammalian cortex. In any event, they appear to be among the substrates of octopus learning and memory. Indeed, recent studies of slice preparations of octopus vertical lobe by Hochner, Fiorito, and colleagues have identified long-term potentiation (LTP) of glutamatergic synaptic field potentials similar to that found in vertebrates. Most intriguing is the finding
33
by Hochner, Shomrat, and Fiorito that, unlike the vertebrate form of neural plasticity, LTP in the octopus vertical lobe is NMDA independent, suggesting an entirely convergent evolution of this form of neural plasticity. Although we are far from making a strong case for sensory consciousness in cephalopods such as the octopus, present evidence from behavior, neuroanatomy, and neurophysiology does not preclude this possibility. Moreover, and tantalizingly, a variety of techniques are approaching a stage of maturity that holds the promise of substantive investigations of the necessary conditions for consciousness in these animals so phylogenetically distant from us.
Conclusion: Complex Nervous Systems, Evolutionary Convergence, and Universal Properties of Consciousness If consciousness did indeed arise in radically different nervous systems of animals such as mammals and birds on the one hand and cephalopods on the other, then the possibility that certain properties of consciousness are universal becomes highly likely. What might the implications of this be? A very important implication is that the elaboration of a capacity for higher-order discrimination, such as that which might underlie the process of consciousness, can arise in different kinds of neural substrates so long as the same functional interactions are in play within these substrates. In effect, the organization of structures, either over the course of development or in the sense of overall gross anatomy, may actually be secondary to issues of functional identity, that is, such elements of conscious states as the reentrant circuitry proposed by Edelman and Tononi, linking areas specialized for perception and those specialized for memory, which seems to be critical for mammalian consciousness (i.e., thalamocortical loops), could well have been instantiated a number of times in tissues with very different embryological – and evolutionary – histories. Finally, if consciousness indeed appeared independently at least three times during animal evolution, it seems likely that increasingly complex ecologies have
34
Animal Consciousness
tended to drive the nervous systems of highly motile animals – major structural differences notwithstanding – inexorably toward the capacity for higher-order discriminations and, ultimately, conscious experience.
Acknowledgments The work of the author was supported by the Neurosciences Research Foundation. See also: Ethical Implications: Pain, Coma, and Related Disorders.
Suggested Readings Aboitiz F (1999) Comparative development of the mammalian isocortex and the reptilian dorsal ventricular ridge: Evolutionary considerations. Cerebral Cortex 9: 783–791. Agin V, Dickel L, Chichery R, and Chichery MP (1998) Evidence for a specific short-term memory in the cuttlefish, Sepia. Behavior Processes 43: 329–334. Ayala-Guerrero F (1989) Sleep patterns in the parakeet Melopsittacus undulates. Physiology and Behavior 46(5): 787–791. Bachmann T, Breitmeyer B, and O¨gmen H (2007) Experimental Phenomena of Consciousness. Oxford: Oxford University Press. Boycott BB and Young JZ (1955) A memory system in Octopus vulgaris Lamarck. Proceedings of the Royal Society of London. Series B: Biological Sciences 143(913): 449–480. Bullock TH (2003) Have brain dynamics evolved. Should we look for unique dynamics in the sapient species? Neural Computation 15: 2013–2027. Bullock TH and Budelmann BU (1991) Sensory evoked potentials in unanesthetized unrestrained cuttlefish: A new preparation for brain physiology in cephalopods. Journal of Comparative Physiology 168(1): 141–150. Buxhoeveden DP and Casanova MF (2002) The minicolumn hypothesis in neuroscience. Brain 125: 935–951. Cheney DL and Seyfarth RM (1990) How Monkeys See the World: Inside the Mind of Another Species. Chicago: University of Chicago Press. Cowey A and Stoerig P (1995) Blindsight in monkeys. Nature 373(6511): 247–249. Darwin C (2004 [1871]) The Descent of Man, and Selection in Relation to Sex. London: Penguin Classics. Darwin C (1898 [1872]) The Expression of the Emotions in Man and Animals. New York: D. Appleton and Company. Edelman GM (1987) Neural Darwinism. New York: Basic Books. Edelman GM (1988) Topobiology: An Introduction to Molecular Embryology. New York: Basic Books. Edelman GM (1989) The Remembered Present. New York: Basic Books.
Edelman GM (2004) Wider than the Sky: The Phenomenal Gift of Consciousness. New Haven: Yale University Press. Edelman DB, Baars BJ, and Seth AK (2005) Identifying hallmarks of consciousness in non-mammalian species. Consciousness and Cognition 14(1): 169–187. Fiorito G and Chichery R (1995) Lesions of the vertical lobe impair visual discrimination learning by observation in Octopus vulgaris. Neuroscience Letters 192: 117–120. Fiorito G, Biederman GB, Davey VA, and Gherardi F (1998) The role of stimulus preexposure in problem solving by Octopus vulgaris. Animal Cognition 1: 107–112. Fiorito GP and Scotto P (1992) Observational learning in Octopus vulgaris. Science 256: 545–574. Gallup GG, Jr (1970) Chimpanzees: Self recognition. Science 167(3914): 86–87. Gardner RA and Gardner BT (1969) Teaching sign language to a chimpanzee. Science 165(894): 664–672. Grin DR (1976) The Question of Animal Awareness: Evolutionary Continuity of Mental Experience. New York: The Rockefeller University Press. Grin DR and Speck GB (2004) New evidence of animal consciousness. Animal Cognition 7(1): 5–18. Hanlon RT and Messenger JB (2002) Cephalopod Behavior. Cambridge: Cambridge University Press. Hochner B, Brown ER, Langella M, Shomrat T, and Fiorito G (2003) A learning and memory area in the octopus brain manifests a vertebrate-like long-term potentiation. Journal of Neurophysiology 90(5): 3547–3554. Hochner B, Shomrat T, and Fiorito G (2006) The octopus: A model for a comparative analysis of the evolution of learning and memory mechanisms. Biological Bulletin 210: 308–317. Jarvis ED, Ribeiro S, Da Silva ML, Ventura D, Vielliard J, and Mello CV (2000) Behaviorally driven gene expression reveals song nuclei in hummingbird brain. Nature 406: 628–632. Kardong KV (1995) Vertebrates: Comparative Anatomy, Function, and Evolution. Dubuque, IA: W.C. Brown. Karten HJ (1997) Evolutionary developmental biology meets the brain: The origins of mammalian cortex. Proceedings of the National Academy of Sciences of the United States of America 94: 2800–2804. Karten HJ, Hodos W, Nauta WJ, and Revsin AM (1973) Neural connections of the visual wulst’’ of the avian telencephalon: Experimental studies in the pigeon (Columbia livia) and owl (Speotyto cunicularia). Journal of Comparative Neurology 150(3): 253–278. Leise EM (1990) Modular construction of nervous systems: A basic principle of design for invertebrates and vertebrates. Brain Research. Brain Research Reviews 15(1): 1–23. Logothetis NK (1998) Single units and conscious vision. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 353: 1801–1818. Luo M and Perkel DJ (2002) Intrinsic and synaptic properties of neurons in an avian thalamic nucleus during song learning. Journal of Neurophysiology 88: 1903–1914. Luo M, Ding L, and Perkel DJ (2001) An avian basal ganglia pathway essential for vocal learning forms a closed topographic loop. Journal of Neuroscience 21: 6836–6845.
Animal Consciousness Mather J (2008) Cephalopod consciousness: Behavioral evidence. Consciousness and Cognition 17(1): 37–48. Medina L and Reiner A (2000) Do birds possess homologues of mammalian primary visual, somatosensory and motor cortices. Trends in Neuroscience 23(1): 1–12. Merker B (2005) The liabilities of mobility: A selection pressure for the transition to consciousness in animal evolution. Consciousness and Cognition 14: 89–114. Messenger JB (2001) Cephalopod chromatophores: Neurobiology and natural history. Biological Reviews 76: 473–528. Moriyama T and Gunji YP (1997) Autonomous learning in maze solution by octopus. Ethology 103(6): 499–513. Naftolin F, Horvath T, and Balthazart J (2001) Estrogen synthetase (Aromatase) immunohistochemistry reveals concordance between avian and rodent limbic systems and hypothalamus. Experimental Biology and Medicine 226: 717–725. Pepperberg IM (1998) Possible perceptual consciousness in grey parrots (Psittacus erithacus). American Zoologist 35 (5): 7A. Premack D and Premack AJ (1984) The Mind of an Ape. New York: W.W. Norton & Company. Puelles L (2001) Thoughts on the development, structure and evolution of the mammalian and avian telencephalic pallium. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 356(1414): 1583–1598. Sanders GD (1975) The cephalopods. In: Corning WC, Dyal JA, and Willows AOD (eds.) Invertebrate Learning. New York: Plenum Press. Savage-Rumbaugh S, McDonald K, Sevcik RA, Hopkins WD, and Rubert E (1986) Spontaneous symbol acquisition and communicative use by pygmy chimpanzees (Pan paniscus). Journal of Experimental Psychology. General 115(3): 211–235. Savage-Rumbaugh S, Sevcik RA, Rumbaugh DM, and Rubert E (1985) The capacity of animals to acquire language: Do species differences have anything to say to us. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 308(1135): 177–185. Seth AK and Baars BJ (2005) Neural Darwinism and consciousness. Consciousness and Cognition 14(1): 140–168.
35
Seth AK, Baars BJ, and Edelman DB (2005) Criteria for consciousness in humans and other mammals. Consciousness and Cognition 14(1): 119–139. Seyfarth RM and Cheney DL (2003) Meaning and emotion in animal vocalizations. Annals of the New York Academy of Sciences 1000: 32–55. Skinner BF (1938) The Behavior of Organisms: An Experimental Analysis. New York: Appleton-Century Company, Inc. Srinivasan R, Russell DP, Edelman GM, and Tononi G (1999) Increased synchronization of neuromagnetic responses during conscious perception. Journal of Neuroscience 19: 5435–5448. Sutherland NS (1969) Shape discrimination in rat, octopus, and goldfish: A comparative study. Journal of Comparative Physiology Psychology 67: 160–176. Tononi G and Edelman GM (1998) Consciousness and complexity. Science 282: 1846–1851. Van Swinderen B and Greenspan RJ (2003) Salience modulates 20–30 Hz brain activity in Drosophila. Nature Neuroscience 6(6): 579–586. Wada K, Hagiwara M, and Jarvis ED (2001) Brain evolution revealed through glutamate receptor expression profiles. Society for Neuroscience Abstract Vol. 27. Watson JB (1913) Psychology as the behaviorist views it. Psychological Review 20: 158–177. Weir AAS, Chappell J, and Kacelnik A (2002) Shaping of hooks in New Caledonian crows. Science 297: 981. Wells MJ and Young JZ (1969) The effect of splitting part of the brain or removal of the median inferior frontal lobe on touch learning in octopus. The Journal of Experimental Biology 56(2): 381–402. Wells MJ and Young JZ (1972) The median inferior frontal lobe and touch learning in the octopus. The Journal of Experimental Biology 56(2): 381–402. White SA, Fisher SE, Geschwind DH, Scharff C, and Holy TE (2006) Singing mice, songbirds, and more: Models for FOXP2 function and dysfunction in human speech and language. The Journal of Neuroscience 26(41): 10376–10379. Young JZ (1961) Learning and discrimination in the octopus. Biological Reviews 36: 32–96. Young JZ (1971) The Anatomy of the Nervous System of Octopus Vulgaris. Oxford: Clarendon Press. Young JZ (1983) The distributed tactile memory system of octopus. Proceedings of the Royal Society of London. Series B: Biological Sciences 218: 135–176.
36
Animal Consciousness
Biographical Sketch
Dr Edelman is an associate fellow in Experimental Neurobiology at The Neurosciences Institute, in San Diego, California, where his laboratory is engaged in the study of mitochondrial dynamics in the mammalian brain. He has longstanding interest in the behavioral, neuroanatomical, and neurophysiological correlates of animal consciousness. Recently, he has begun to collaborate with Dr Graziano Fiorito (Stazione Zoologica, Naples, Italy) on a psychophysical approach to studying octopus cognition. He is an active member of the Association for the Scientific Study of Consciousness and, together with Drs Anil Seth and Bernard Baars, has published a number of papers on consciousness, most recently in the journal Consciousness and Cognition.
Artificial Intelligence and Consciousness O Holland and D Gamez, University of Essex, Colchester, UK ã 2009 Elsevier Inc. All rights reserved.
Glossary Artificial general intelligence (AGI) – A system with AGI is capable of wide-ranging human-like intelligence that is not limited to the performance of a single task. Artificial intelligence (AI) – Systems with AI are capable of intelligent reasoning and behavior, and often have some form of learning mechanism. Frame problem – Every action has a large number of potential consequences and it can be difficult or impossible for a robot to know whether it has worked through enough of them to accurately update its knowledge of the world. Global workspace theory – A cognitive architecture that accounts for the difference between consciousness and unconsciousness. A number of specialized processors receive information from the global workspace, process it, and compete to output their result back to the global workspace, with the cycle being repeated until the problem is solved or a new problem is presented. Information integration – In a system with high information integration one element can influence many other elements in the system; in a system with low information integration the elements are mostly isolated from each other. Intentional state – Intentional states are states of the mind that are directed toward or about objects, properties or states of affairs. Machine consciousness – A research area that focuses on the development of machines with the behavior, cognitive characteristics and/or architecture associated with consciousness. This type of work may lead to the development of systems that are phenomenally conscious. Synthetic phenomenology – A research area that attempts to decide whether artificial systems are capable of conscious states and tries to describe the phenomenology of these states if they occur. Turing test – A test in which an interrogator communicates with a hidden person and a computer and attempts to identify which is which. A computer is said to have passed the Turing test if it cannot be distinguished from the person.
Introduction At the most general level artificial intelligence (AI) is concerned with the development of systems that are capable of intelligent reasoning and behavior, and often include some form of learning mechanism. AI systems may be standalone computer programs or they may process data from the world and produce behavioral output using a robot body. After a rather shaky start AI is now a major research area and it is used to perform tasks ranging from the processing of security camera data to the control of autonomous vehicles. During its development AI has influenced our theories about consciousness and provided a good way of testing them, and there has been an extensive debate about whether consciousness can be created in a machine. Several different ways of testing for consciousness in artificial systems have been put forward and the research on machine consciousness has raised a number of ethical and legal issues.
An Overview of Artificial Intelligence Research on AI is generally considered to have started in 1956 when a conference was held at Dartmouth with a number of the key figures. The early computer programs represented problems using logical symbols and solved them by carrying out manipulations that were defined and constrained by explicit formal rules. Some of this work was inspired by studying how humans claimed to solve similar problems, and the underlying idea that the intelligence of minds and machines could have the same basis was articulated as the physical symbol system hypothesis, which claims that a physical symbol system has the necessary and sufficient means for general intelligence. Although it was not explicitly a theory of consciousness, the physical symbol system hypothesis was held to account for a variety of cognitive abilities necessary for intelligence. Many of these cognitive abilities, such as understanding, would ordinarily be thought to imply some degree of consciousness. While the logic-based approach to AI had some initial success, by the early 1970s it was starting to run into difficult problems. To begin with, the programs that had been developed up to that point could only handle basic examples of the problems that they were supposed to represent, and the computers of that time did not have enough memory or processing speed to accomplish anything useful or to process significant amounts of data from the world. It also became clear that although computers
37
38
Artificial Intelligence and Consciousness
could solve problems that humans found difficult, such as theorem proving, it was very difficult to program them to carry out tasks that humans found easy, such as object recognition and manipulation. A third problem that emerged was that our interactions with the world depend on a background of common sense knowledge that constrains our selection of actions in a particular situation. Manually programming all of this knowledge into a computer would have been a large and almost impossible task and there was no other obvious way in which an artificial system could gain this type of information. Logic-based AI also encountered a difficulty known as the ‘frame problem,’ which is that even the most basic actions have a large number of potential consequences. It often takes a long time to calculate these potential consequences and it can be impossible for a robot to know when it has worked through enough of them to accurately update its knowledge of the world. A number of researchers responded to these problems by abandoning the logic-based approach in favor of robots that grounded their representations directly in the real world using sensors and actuators. The behavior of these systems was controlled by modules that processed the sensory data and produced a motor output, and these modules could communicate through the world and inhibit each other (to manage conflict resolution). The modules were typically organized in a series of layers connecting perception to action, with the higher layers taking the output of the lower layers into account – an arrangement known as the subsumption architecture. One problem with the robotics approach was that the difficulties and expense of working with real robots were so severe that many people used robot simulations, which often introduced misleading simplifications, such as a lack of noise. A second response to the problems with logic-based AI was for researchers to focus on well-specified problems within a particular domain that could be addressed more easily, such as chess playing or data mining. This type of specialized AI is often contrasted with work that is attempting to replicate human intelligence completely, which is known as artificial general intelligence (AGI). More narrowly focused AI has developed a number of technologies and had some impressive success – for example, a computer beat Gary Kasparov at chess in 1997 – and machine learning is now used routinely in areas ranging from computer games to stock market prediction. The rapid increase in computer power in recent years has made AI easier to implement and it has now developed a range of technologies that include artificial neural networks, agent systems and a variety of learning and reasoning techniques. Contemporary AI has also carried out some biologically inspired work based on simpler creatures, such as ants and salamanders. The resurgence of interest in consciousness in recent years has inspired a number of researchers to test theories of consciousness
using computer models, and there has been some speculation that this could eventually lead to more intelligent machines that might actually have phenomenal states. This type of research is gradually becoming known as ‘machine consciousness,’ although ‘artificial consciousness’ and occasionally ‘digital sentience’ have also been used to describe it.
Artificial Intelligence and Consciousness Research The Influence of Artificial Intelligence on Theories about Consciousness The early work on AI had a significant impact on our thinking about the mind because it highlighted the limitations of a purely logic-based approach and pointed to a number of issues, such as the frame problem and the background of common sense, that any thinking creature would have to solve. The work on robotics suggested that there is an important link between embodiment and intelligence, and this led researchers to speculate that physical embodiment may be necessary for consciousness. One of the most direct links between AI and theories about consciousness is Bernard Baars’ global workspace architecture, which is based on the blackboard architecture that was developed in the 1970s as part of research on AI. What the two have in common is that a large number of specialized processors analyze the same problem data and compete to output their result back to the blackboard/global workspace, with the cycle being repeated until the problem is solved or a new problem is presented. However, within global workspace theory the problem data are actively distributed to all processors by a broadcast mechanism, instead of using a blackboard that all processors can inspect, and global workspace theory compares the outcome of the competitive selection process to a spotlight on the stage of a darkened theatre, which lights up the message of the winning processor for the ‘audience’ of all the processors. While the blackboard architecture is an AI technology that was designed to solve specific problems, global workspace theory was put forward by Baars as a cognitive model of consciousness in which the spotlit broadcast message forms the contents of consciousness and the operations of the processors are unconscious. Applications of Artificial Intelligence to Research on Consciousness The work that has been carried out in AI, computer science and robotics can be used to test theories of consciousness by building models. This type of test is a good way of making a theory’s assumptions explicit since thought experiments can leave all kinds of things out, but a model will only work when everything necessary
Artificial Intelligence and Consciousness
has been included. While the usefulness of this approach is often limited by the fact that experimental systems are very basic, this type of modeling provides a very good way of bridging the gap between abstract theories and the real world and it can help us to understand how consciousness is implemented in the brain. A further benefit of this type of work is that it can help us to clarify our intuitions about consciousness. If someone claims that X, Y, and Z are necessary and sufficient for consciousness, then it is generally possible to develop a machine that has X, Y, and Z. If X, Y, and Z are functions, then the processing of X, Y, and Z in the machine can be done by any form of computation, and opponents of function-based theories of consciousness have pointed out that computation can be carried out by the manipulation of beer cans or by the population of China communicating with mobile phones and satellites. While it might have been easy to believe that X, Y, and Z were necessary and sufficient for consciousness in human beings, it is much harder to believe that these eccentric implementations are really conscious. Within AI many systems have been developed that have attributes associated with consciousness, such as imagination, a global workspace architecture or information integration (see later discussion), and yet researchers are wary of attributing phenomenal consciousness to them. This suggests that the presence of biological neurons has a big influence on our intuitions about consciousness, although the absence of consciousness in deep sleep indicates that this is not sufficient. AI has also been used in consciousness research to specify the contents of conscious experiences. It is generally acknowledged that infants and animals are likely to have very different phenomenal states from humans, and it is difficult or impossible to say in human language what it is like to be an infant or an animal. AI technology can help with this problem because it enables us to build systems whose states, interactions, and capacities can be used to describe the phenomenology of very young humans or nonhuman creatures. For example, some theories of consciousness link conscious states to our expectations about what will appear next in the visual field. The possible phenomenology of this type of system has been specified by the construction of a robot-based device, whose perceptions were based on its expectations about visual input, and not on the actual visual data that it received.
The Debate about Consciousness in Artificial Systems Positive Theories about Machine Consciousness Naturalistic accounts of consciousness
Many people believe that the brain is a complex mechanism that came about through natural selection and is
39
governed by well-understood principles that apply to all other physical processes in living things. If there is nothing special about the brain compared to the liver, for example, then once we have understood how consciousness works in the brain, we should be able to produce consciousness in other systems as well. Up to this point science has not identified anything in the brain that is correlated with consciousness and would be impossible to put into an artificial system. If such magical ingredients cannot be found, then it seems reasonable to suppose that silicon chips, wires, motors, and cameras could be assembled into a conscious robot without leaving anything out that is necessary for consciousness – although it might be difficult to match the processing speed and complexity of biological systems. Information theories of consciousness The information integration theory of consciousness was put forward by Giulio Tononi, who claims that the conscious parts of a system have the greatest capacity to integrate information. Tononi also maintains that the amount of consciousness in a system depends on the amount of information integration. Information integration is measured using the value F, which is calculated by dividing the system up in many different ways and evaluating the causal influence of the parts on each other. Tononi’s claim about a link between consciousness and information integration is entirely independent of the material out of which the system is made, and so any system that is capable of integrating information is conscious to some degree according to this theory. The information integration theory of consciousness has some overlap with David Chalmers’ claim that conscious experiences are realizations of information states, which predicts that systems as simple as thermostats are conscious because they process information. Cognitive theories Many people have conjectured that consciousness may be linked to cognitive characteristics, such as emotions, imagination, and a model of the self. If consciousness depends on functions at the cognitive level, then it should be possible to realize it on any piece of hardware that is capable of carrying out the appropriate processing. One example of a cognitive theory of consciousness is the axiomatic theory of Igor Aleksander, who claims that imagination, emotion, depiction, volition, and attention are minimally necessary for consciousness, and any natural or artificial system that implements these axioms is judged to be conscious according to this theory. Global workspace theory is another cognitive architecture that can be implemented on many different types of hardware, and Thomas Metzinger’s phenomenal self model and constraints on conscious experience are also largely independent of the physical system.
40
Artificial Intelligence and Consciousness
Other theories
Many other theories of consciousness have positive implications for the possibility of creating consciousness in artificial systems. To begin with, the pantheist claim that all matter is conscious to some degree suggests that computers and robots are conscious even when they are switched off. David Rosenthal’s claim that consciousness depends on a higher-order thought about another mental state is not linked to any particular implementation, and a number of connections have been made between consciousness and virtual machines that can run on any type of hardware. Some researchers have claimed that inner speech helps to constitute our sense of self and agency and may be important to consciousness as well. If this is the case, it might be possible to use the work on language acquisition in AI to develop conscious systems. Criticisms of Machine Consciousness The hard problem of consciousness
People working on consciousness commonly distinguish between the easy problem of explaining how we can discriminate, integrate information, report mental states, focus attention, etc., and the hard problem of explaining phenomenal experience. Although solving the ‘easy’ problem is far from easy, we do at least have some idea how it could be done, but it can be argued that we have no real idea about how to solve the hard problem. If we do not understand how human consciousness is produced, then it makes little sense to try to make robots phenomenally conscious. Objections to machine consciousness based on the hard problem may only be a caution about any current claims that we might want to make, because we may solve the hard problem in the future and discover that machines are capable of consciousness. Furthermore, it can be argued that asking questions about phenomenal states in machines and building models of consciousness are likely to improve our understanding of human consciousness and take us closer to a solution to the hard problem. It might also be possible to create conditions that allow consciousness to emerge in a system without understanding the causes of phenomenal states – for example, it has been suggested that consciousness could emerge in a detailed simulation of a human infant that develops by interacting with its environment. The hard problem of consciousness only becomes problematic for work on machine consciousness if it can never be solved, which would make it difficult to make strong claims about the consciousness of artificial systems. This is the position of Colin McGinn, who argues that consciousness depends on a natural property, P, that cannot be grasped by human beings. If McGinn is right, we will never know if machines can have property P and we will be unable to measure P in the systems that we have built.
The Chinese room One of the most influential attacks on the idea that AI could develop a mind was John Searle’s Chinese Room thought experiment in which a person in a room receives Chinese characters, processes them according to a set of rules, and passes out the correct result without understanding what the characters mean. This processing of characters could be used to create the external behavior associated with consciousness, to simulate the cognitive characteristics associated with consciousness, or to model a conscious architecture. However, Searle argues that the person processing characters in the room would not understand the objects represented by the Chinese characters or have intentional states about them, and so the Chinese Room would never become a real mind. Although consciousness as such barely featured in the original paper, it (or its absence) has since become a key feature of the Chinese Room scenario, which is now seen by many of its proponents as an attack on the idea that a computer executing a program could become conscious, rather than merely intelligent. One response to this argument is that the Chinese Room could be grounded in some kind of nonsymbolic lower level, such as images or sounds, which would give it the ability to understand the characters that it is manipulating and have intentional states directed toward the nonsymbolic content. This type of grounding could be achieved by building an embodied system that processed data from the real world, and neural models have been proposed as a way of grounding higher level symbolic representations in sensory inputs. A second objection to the Chinese Room argument is that both brains and computers are physical systems assembled from protons, neutrons, and flows of electrons, and Searle is happy to claim that consciousness is a causal outcome of the physical brain. Until the hard problem of consciousness has been solved, we will be unable to say whether the physical computer and the physical brain are different in a way that is relevant to consciousness, and since we have no idea about this at present, the Chinese Room argument does not offer any a priori reason why the arrangement of protons, neutrons, and electrons in a physical computer is less capable of consciousness than the arrangement of protons, neutrons, and electrons in a physical brain. Consciousness is not algorithmic Machine consciousness has been criticized by Roger Penrose, who claims that the processing of an algorithm is not enough to evoke phenomenal awareness because subtle and largely unknown physical principles are needed to perform the noncomputational actions that lie at the root of consciousness. If consciousness does something that ‘mere’ computation cannot, then it cannot be simulated by a computer and phenomenal states cannot be created in a machine. The most straightforward response
Artificial Intelligence and Consciousness
to Penrose is to reject his theory of consciousness, which has been heavily criticized by a number of people. However, even if Penrose’s theories about consciousness are accepted, it may still be possible to develop some kind of quantum computer that incorporates the type of physical action that is linked by Penrose to consciousness. The limitations of artificial intelligence
It has already been pointed out that AI research encountered deep problems when it attempted to implement intelligence using logical symbols manipulated by rules, and Hubert Dreyfus has argued that this attempt failed because human intelligence depends on skills, body, emotions, imagination, and other attributes that cannot be encoded into long lists of facts. While this is a good objection to using logic to try to develop systems that are as intelligent as humans in real world situations, there is no reason why machine consciousness could not be pursued in more limited ways independently of this objective. For example, some of the behaviors that require consciousness in humans could be created in a simple way, and imagination and emotion can be simulated without the expectation that they will work as effectively as human cognitive processes. It can also be argued that the work being carried out on imagination, emotions, and embodiment in machine consciousness addresses some of the areas that are lacking in traditional approaches to AI.
Research on Machine Consciousness Current Research Machines with the external behavior associated with consciousness
If consciousness is the result of natural selection, then the possession of consciousness must have changed organisms’ behavior in a way that gave them an evolutionary advantage and it should be possible to list the behaviors associated with consciousness. One problem with identifying a definitive list is that many waking behaviors can be carried out consciously and nonconsciously, which makes it difficult to decide whether the behavior is associated with consciousness or not. A commonly cited example is that we can drive home from work with our attention on other things, and patients suffering from epileptic automatism can carry out complex actions, such as diagnosing lung patients or cleaning guns, without any conscious awareness. While these examples show that humans can carry out a limited amount of complex behavior unconsciously, the stereotypical nature of this behavior suggests that more dynamic and interactive activities, such as interpersonal dialogue, can only be carried out consciously, and many new behaviors can only be learnt when consciousness is present.
41
Since there is not a clear specification of the behaviors associated with consciousness, relatively few people are explicitly working on this task, although there has been some recent work on a functional specification for a conscious machine that could eventually be implemented in computer code. Most of the research in this area is being carried out by people who are attempting to mimic human intelligence completely, either as part of work on AGI or in order to pass the Turing test (see later discussion). The work on AGI has been mostly logic-based, and to overcome the difficulties with this approach one company has been attempting to build a database with enough facts that would enable it to become generally intelligent, although it has had little success so far. There has also been some work on the development of systems with memory, reasoning, and learning that are being used to control virtual characters in online multiplayer environments. The dialogue systems that have been developed to pass the Turing test are often superficially convincing for short periods, but none of them have managed to pass the test yet. Machines with the cognitive characteristics associated with consciousness
A number of connections have been made between consciousness and cognitive characteristics, such as imagination and emotions, and it has been suggested that an internal model of the self plays an important role in consciousness. The modeling of the cognitive characteristics associated with consciousness has been a strong theme in machine consciousness and a number of different approaches have been explored. One approach to this area has been the simulation of the cognitive characteristics associated with consciousness using neural networks with tens of thousands of neurons. In one set of experiments a series of networks was used to model basic versions of emotion, imagination, volition, attention, and depiction; in other work, a neural network was developed that controlled the eye movements of a virtual humanoid robot and used models of imagination and emotion to decide whether it looked at a red or blue cube. There has also been some research that has used simple wheeled robots and small recurrent neural networks to simulate imagination and internal models. One set of experiments started by training a neural network to move a robot around a simple maze. The network’s predictions about the next sensory state were then connected up to its sensory input and the network was able to navigate ‘blind’ around the maze using its ‘imagination.’ A second group of researchers used a simple wheeled robot and a machine learning algorithm to develop a system that explored its environment and built up an internal model that could be graphically displayed. This work on internal modeling was taken further by the development of a system that used a
42
Artificial Intelligence and Consciousness
virtual robot as an internal model of a real humanoid robot. The virtual environment was updated with data from the real world and the system used this virtual model to imagine the consequences of different actions. When the virtual modeling suggested that an action would have a positive outcome, the behavior was executed by the real robot. Machines with an architecture that is claimed to be a cause or correlate of human consciousness
A number of theories have been put forward about the architectures that are associated with consciousness. Most of the machine consciousness work in this area has either modeled Baars’ global workspace or created brain-based models of the neural correlates of consciousness. The most well-known global workspace model is the Intelligent Distribution Agent (IDA) naval dispatching system developed by Stan Franklin that was created to assign sailors to new billets at the end of their tour of duty. This task involves natural language conversation, interaction with databases, adherence to Navy policy, and checks on job requirements, costs, and sailors’ job satisfaction. These functions were carried out using a large number of small programs that were specialized for different tasks and the whole system was organized using a global workspace architecture. There have also been a number of neural models based on global workspace theory. One neural network was developed with 40 000 neurons and a brain-inspired global workspace architecture that enabled a virtual robot to explore the consequences of potential actions and select the one that would result in a positive stimulus. Detailed neural models of the global workspace architecture have been used to model a number of psychological phenomena related to consciousness, such as the transitions between the awake state and sleep, anesthesia or coma. Within research on the neural correlates of consciousness, a brain-based model with 100 000 neurons was created to study neural computation, attention, and consciousness. In other work a simulated neural network was developed with areas that were based on the brain’s neuroanatomy. This network had connections between its motor and sensory areas, which some people have linked to consciousness, and it was used to control a virtual child whose ‘survival’ depended on its communications with a human operator. Phenomenally conscious machines
The previous three approaches to machine consciousness are all relatively uncontroversial, since they are modeling phenomena linked to consciousness without any claims about real phenomenal states. The fourth area of machine consciousness is more philosophically problematic, since it is concerned with machines that might have real phenomenal experiences – machines that are not just tools in consciousness research, but are actually conscious
themselves. This approach has some overlap with the other approaches, since in some cases it may be hypothesized that the reproduction of human behavior, cognitive states, or internal architecture will lead to real phenomenal experiences. On the other hand, phenomenal states might be achievable independently of other approaches to machine consciousness. For example, it might be possible to create a system based on biological neurons that was capable of phenomenal states, but lacked the architecture of human consciousness and any of its associated cognitive states or behaviors. The lack of consensus about consciousness has made most AI researchers cautious about attributing real phenomenal states to their systems. There have not been any strong behavioral grounds for attributing consciousness to these systems either, although it has been possible to make limited predictions about phenomenal states in artificial systems using particular theories of consciousness. Future Developments Progress in machine consciousness is likely to be slow and incremental, with increasingly complex systems being developed that are based on the cognitive characteristics and architectures associated with consciousness. These systems will gradually display more human-like behavior and they will be increasingly likely to possess phenomenal states. There has also been speculation that the expansion of the Internet or the apparently inexorable increase in computer power will lead to conscious systems. The number of computers linked together in the Internet (currently some hundreds of millions) and the bandwidth of connections are increasing every year, and so in principle the Internet can be regarded as an enormous and growing distributed computing system. Some people claim that, as the Internet grows further, it will achieve a size at which it will inevitably become conscious. There is no evidence from within AI suggesting that this should be the case, although it is worth noting that the information integration theory of consciousness claims that any system of active interconnected elements is potentially conscious. However, it will be difficult to tell which parts of the Internet are predicted to be conscious according to this theory because it is very difficult to calculate information integration in large networks, and there is only likely to be any potential for consciousness in the Internet as a whole when the data transfer rate between computers exceeds the data transfer rate within each computer. For the last 50 years, the density with which transistors can be packed onto an integrated circuit has doubled approximately every 2 years. This exponential growth rate, known as Moore’s law, is expected to continue for another decade or two, and this increase in packing density leads to proportionate increases in performance and decreases in cost. It has been claimed that by around 2019
Artificial Intelligence and Consciousness
a typical desktop computer will have the computational capacity of the human brain (although the way in which this has been calculated is certainly open to question), and will cost only $1000. If the key to producing an intelligent computer is sheer processing power, then whatever power is required will become available within a relatively short time frame. Assuming that AI software continues to improve (in some unspecified manner, but presumably able to take advantage of the computational resources), then some have claimed that it will be possible to build a machine that can pass the Turing test by 2029. For those happy to accept the Turing test as a test of consciousness as well as intelligence, this amounts to a prediction that an increase in the available computing power, combined with improved AI software, is all that is needed to produce a conscious machine. However credible the proponents of this argument may be as individuals, the argument itself is fatally weakened by the lack of specification of the principles on which the AI software will be built. A variant of this argument proposes that the available computing power could be used to run an exact simulation of a human brain, assuming that the relevant neuroscientific details will be available by then. Whether this will yield a conscious machine is a matter for computational neuroscience rather than AI. A third possibility that is often discussed in connection with this area is the extension or enhancement of consciousness with technology. In a trivial sense this happens whenever a person looks through a pair of binoculars or listens to amplified music, but advocates of this view believe that we will eventually be able to plug external devices into our brains and thus become directly conscious of the Internet, for example, without the mediation of our senses. While this would be more of a technological extension of our consciousness than a direct application of AI, it is likely that some form of AI would play a role in this, either to control what would be a very complex interface or more fancifully as an advanced AI that we would fuse with through a brain interface. However, while implanted electrodes are able to provide very basic vision and can be used to control a robot arm, more advanced extensions of our consciousness remain entirely within the realm of science fiction.
Testing for Consciousness in Artificial Systems Behavioral Tests Turing tests
The best-known behavioral test for consciousness is the Turing test, which was put forward by Alan Turing in 1950 as an answer to the question ‘Can machines think?’ Instead of defining what he meant by ‘machines’ and ‘think,’ he chose to limit the machines to digital computers and
43
operationalized thinking as the ability to answer questions in a particular context well enough that the interrogator could not reliably discriminate between the answers given by a computer and a human (via teleprinter) after 5 min of questioning. In the original paper consciousness was mentioned only in the context of the objection that thought in the brain is always driven and accompanied by feeling, and so the mere generation of text by a machine in the absence of feeling, however convincing it might otherwise seem, could not be taken as a sufficient indicator of thought. After pointing out that a refusal to accept the computer’s text at face value would constitute a kind of solipsism, a doctrine he thought unacceptable, Turing expressed the opinion that most people initially supporting the argument from consciousness could be persuaded to abandon it rather than embrace solipsism, and so they would probably be willing to accept the validity of the test. In accepting that consciousness cannot strictly be inferred from behavior, Turing does not seem to be claiming that the test can establish that a machine’s thinking is accompanied by feelings, but rather that the difficult issue of determining the involvement of feelings should be set aside in view of the (hypothetical) excellence of the linguistic outputs. More recently a number of people have suggested extensions to the standard test that involve processing audio and visual data or controlling a humanoid body in a humanlike way for an extended period of time. In order to pass any of the Turing tests, a machine would almost certainly have to have experience of the world, a capacity for imagination and an emotion system, since no sequence of preprogrammed responses is likely to be convincing over an extended period of time. Other behavioral measures The link between certain behaviors and consciousness was discussed earlier, and while this is only vaguely defined at present, it might eventually be possible to use these behaviors to identify consciousness in artificial systems. One potential problem with this approach is that we might not be prepared to attribute phenomenal states to all systems that can carry out the functions associated with consciousness, and so behavioral tests might have to be combined with other measures.
Internal Architectural Tests Another way of testing for consciousness in artificial systems is to inspect the internal architecture and states of the system and use a theory of consciousness to make predictions about its phenomenal states. This type of work is part of the nascent discipline of synthetic phenomenology, which is attempting to answer the question whether artificial systems are capable of conscious states and trying to describe these states if they occur. This type of work
44
Artificial Intelligence and Consciousness
overlaps with the emerging discipline of neurophenomenology, which makes predictions about human consciousness using objective measurements of a person’s brain obtained with electrodes or scanning technologies, such as functional magnetic resonance imaging (fMRI). At the current stage of research, architecture-based predictions about the phenomenology of a system are far from certain because there is no commonly agreed explanation of consciousness, and the best that can be done is to make predictions according to a particular theory. It is anticipated that work on the neural correlates of consciousness will lead to substantial progress in neurophenomenology and this may feed into the more controversial problem of making predictions about the consciousness of artificial systems.
Ethical and Legal Issues Many people are concerned that work on machine consciousness will eventually lead to the development of machines that take over the world. Some writers have suggested that this might be a gradual process in which we pass more and more responsibility to machines until we are unable to do without them, in the same way that we are increasingly unable to live without the Internet today. A number of people have responded to this scenario by pointing out that humans cause a great deal of death and destruction every day, and it is possible that conscious machines could run the world better and make humanity happier. Since our current machines fall far short of human intelligence and few people attribute consciousness to them, these science fiction predictions possibly tell us more about our present concerns than about a future that is likely to happen. It is also probable that our attitudes toward ourselves and machines will change substantially over the next century, as they have changed over the last one, and as machines become more human and humans become more like machines, the barriers will increasingly break down between them until the notion of a takeover by machines makes little sense. A second ethical dimension to work on machine consciousness is the question about how we should treat conscious machines. In order to build systems that are capable of consciousness we may have to carry out experiments that would cause conscious robots a considerable amount of confusion and pain. While we want machines that exhibit the behavior associated with consciousness and want to model human cognitive states and conscious architectures, we may want to prevent our machines from becoming phenomenally conscious if we want to avoid the controversy associated with animal experiments. To regulate this potential suffering, a number of people have suggested that conscious robots should be given rights.
A final aspect of the social and ethical issues surrounding machine consciousness is the legal status of conscious machines. When traditional software fails, responsibility is usually allocated to the people who developed it, but the case is much less clear with autonomous systems that learn from their environment. A conscious machine might malfunction because it has been maltreated, and not because it was badly designed, and so its behavior could be blamed on its carers or owners, rather than on its manufacturers. It might also be appropriate to hold conscious machines responsible for their own actions and punish them appropriately.
Conclusions AI has made significant progress over the last 50 years and developed from a purely logic-based approach into a range of successful technologies. Many of the problems encountered by AI over the course of its development have improved our understanding of consciousness and AI provides a good way of testing theories of consciousness. A substantial amount of research has been carried out on the construction of machines with the behavior, cognitive characteristics, and architecture associated with consciousness and this work is likely to make steady progress in the future, although this is unlikely to be in the ways predicted by some futurists. A number of people have raised objections to the possibility of machine consciousness, but none of these appear to be conclusive at the present time, and many theories openly embrace the possibility that consciousness could be realized in an artificial system. There has been some work on testing for consciousness in artificial systems and research on machine consciousness has raised a number of ethical and legal issues. See also: Cognitive Theories of Consciousness; An Integrated Information Theory of Consciousness.
Suggested Readings Aleksander I (2005) The World in My Mind, My Mind in the World: Key Mechanisms of Consciousness in People, Animals and Machines. Exeter: Imprint Academic. Chella A and Manzotti R (eds.) (2007) Artificial Consciousness. Exeter: Imprint Academic. Chrisley RJ, Clowes R, and Torrance S (eds.) (2007) Special Issue: Machine Consciousness: Embodiment and Imagination. Journal of Consciousness Studies 14(7). Dennett DC (1997) Consciousness in human and robot minds. In: Ito M, Miyashita Y, and Rolls ET (eds.) Cognition, Computation and Consciousness. Oxford: Oxford University Press. Gamez D (2008) Progress in machine consciousness. Consciousness and Cognition 17(3): 887–910. Haikonen PO (2003) The Cognitive Approach to Conscious Machines. Exeter: Imprint Academic. Holland O (ed.) (2003) Machine Consciousness. Exeter: Imprint Academic.
Artificial Intelligence and Consciousness Searle JR (1980) Minds, brains and programs. Behavioraland Brain Sciences 3: 417–457. Tononi G (2004) An information integration theory of consciousness. BMCNeurosci ence 5(42).
45
Turing AM (1950) Computing machinery and intelligence. Mind 59: 433–460.
Biographical Sketch
Professor Owen Holland graduated from Nottingham University in 1969 (BSc Hons. in psychology). After teaching experimental methods at Edinburgh University Psychology Department, he moved into commerce, and then into engineering. In 1988, he began his research on behavior-based robotics, for which he received a Department of Trade and Industry SMART award in 1990. In 1993, he moved to the University of the West of England, Bristol, (UWE) to help set up the Intelligent Autonomous Systems Engineering Laboratory. He has held visiting appointments at the University of Bielefeld (1993–94), the California Institute of Technology (1997, 2000–01), and the Ecole Polytechnique Federale de Lausanne (2007), and has also worked in private research laboratories in Cambridge and Brussels. In 2001, he moved to the University of Essex, where he is currently a professor of computer science. His main research interests include biologically inspired robotics, swarm intelligence, swarm robotics, the history of cybernetics, autonomous rotorcraft, and machine consciousness. He has published more than 100 research papers in these areas, and has been particularly closely involved in the development of the field of machine consciousness, having edited the first collection of papers on the topic, and having obtained the first major research grant in the area. With Professor Phil Husbands of the University of Sussex, he is currently working on a book about the Ratio Club, a postwar British cybernetic dining club of which Alan Turing was a member.
Dr. David Gamez completed his BA in natural sciences and philosophy at Trinity College, Cambridge, and went on to study for a PhD in continental philosophy at the University of Essex. This thesis applied methods taken from continental philosophy to contemporary problems in philosophy and science and he later developed and extended this work into a book titled What We Can Never Know (London & New York: Continuum, 2007). After leaving Essex he studied for an MSc in IT at Queen Mary, University of London, and then took up a research position on the EU Safeguard project, which developed an agent system that could protect electricity and telecommunications management networks against attacks, failures, and accidents. When the Safeguard project ended, he was offered a PhD position on Owen Holland’s CRONOS project to build a conscious robot (see www.cronosproject.net). During this PhD he developed a theoretical framework for machine consciousness and made predictions about the representational and phenomenal states of a spiking neural network. He passed this PhD in July 2008 and is now working as a software engineer for the Trinity Mirror Group.
Attention: Change Blindness and Inattentional Blindness R A Rensink, University of British Columbia, Vancouver, BC, Canada ã 2009 Elsevier Inc. All rights reserved.
Glossary Change blindness – The failure to visually experience changes that are easily seen once noticed. This failure therefore cannot be due to physical factors such as poor visibility; perceptual factors must be responsible. Focused attention is believed to be necessary to see change, with change blindness resulting if such attention is not allocated to the object at the moment it changes. Diffuse attention – A type of attention that is spread out over large areas of space. It is believed to be space-based rather than object-based. Focused attention – A type of attention restricted to small spatial extents. It is believed to act on small areas of space or on relatively small objects. Implicit perception – Perception that takes place in the absence of conscious awareness of the stimulus. It is generally believed to take place in the absence of any type of attention. Inattentional blindness – The failure to visually experience the appearance of an object or event that is easily seen once noticed. Attention (likely, diffuse attention) is thought to be necessary for such an experience. Inattentional blindness typically occurs when attention is diverted, such as when the observer engages in an attentionally demanding task elsewhere, and does not expect the appearance of the object or event.
is false – there are severe limits to what we can consciously experience in everyday life. Much of the evidence for this claim has come from two phenomena: change blindness (CB) and inattentional blindness (IB). CB refers to the failure of an observer to visually experience changes that are easily seen once noticed. This can happen even if the changes are large, constantly repeat, and the observer has been informed that they will occur. A related phenomenon is IB – the failure to visually experience an object or event when attention is directed elsewhere. For example, observers may fail to notice an unexpected object that enters their visual field, even if this object is large, appears for several seconds, and has important consequences for the selection of action. Both phenomena involve a striking failure to report an object or event that is easily seen once noticed. As such, both are highly counterintuitive, not only in the subjective sense that observers have difficulty believing they could fail so badly at seeing but also in the objective sense that these findings challenge many existing ideas about how we see. But as counterintuitive as these phenomena are, progress has been made in understanding them. Indeed, doing so has allowed us to better understand the limitations of human perception in everyday life and to gain new insights into how our visual systems create the picture of the world that we experience each moment our eyes are open.
Change Blindness Background
Introduction As observers, we generally have a strong impression of seeing everything in front of us at any moment. But compelling as it is, this impression
The ability to see change is extremely useful in coping with everyday life: we can monitor the movement of nearby automobiles (as drivers or pedestrians), notice sudden changes in the posture or location of people in front of us, and notice that the sky is quickly darkening. Given the importance 47
48
Attention: Change Blindness and Inattentional Blindness
of perceiving change and the fact that most humans can survive reasonably well in the world, it follows that our ability to perceive change must be such that few events in the world escape our notice. This agrees nicely with our impression that we perceive at each moment most, if not all, objects and events in front of us. Failures to see change have long been noticed, but they were usually taken to be temporary aberrations, with nothing useful to say about vision. This attitude began to change in the early twentieth century, when film editors discovered an interesting effect: when the audience moved their eyes across the entire screen (e.g., when the hero exited on the left side and the femme fatale entered on the right) almost any change made during this time (e.g., a blatant change of costume) would often go unnoticed. A similar blindness to change could be induced by making it during a loud, sudden noise (e.g., a gunshot), during which the audience would momentarily close their eyes. The scientific study of this effect began in the mid-1950s, with work on position changes in dot arrays and other simple stimuli. Here, a change in one of the items was typically made contingent on a temporal gap that lasted several seconds. A separate line of studies was also begun that investigated the perception of displacements made contingent on an eye movement (or saccade). In all cases, observers were found to be surprisingly poor at detecting changes made under such conditions. The next wave of studies, begun in the 1970s, was based on a more systematic examination of these effects. This work uncovered a general limit to the ability to detect gap-contingent changes under a wide variety of conditions; this eventually formed the basis for the proposal of a limited-capacity visual short-term memory (vSTM). Likewise, a general lack of ability was found for detection of saccadecontingent changes, which was traced to a limited transsaccadic memory. Both lines of research were eventually linked by the proposal that transsaccadic memory and vSTM were in fact the same system. A third wave began in the mid-1990s, extending the methodology and results of earlier work in several ways. To begin with, stimuli were often more complex and realistic: images of real-world scenes or dynamic events were used in place of
(a) One-shot paradigm
(b)
Flicker paradigm
Figure 1 Example of a technique to induce CB. Here
the change is made during a brief gap between the original and the modified stimulus. (a) One-shot paradigm. A single alternation of the stimuli is used, and the observer must respond to it. Performance is measured by accuracy of detection (or identification or localization). (b) Flicker paradigm. Stimuli are continually cycled until the observer responds. Performance is measured by the time taken until the change is detected (or identified or localized). Both types of measurement paradigm can also be applied to other techniques, such as changes made during saccades or splats.
simple figures. Next, changes were often repeated rather than occurring just once, allowing the use of time as well as accuracy to measure performance (Figure 1). Third, blindness was induced via several new kinds of manipulation, such as making changes during a film cut or during the appearance of sudden splats elsewhere in the image. Finally, all these effects – as well as the earlier ones – were accounted for by the proposal that attention is necessary to see change. This work therefore showed that this CB is a general and robust phenomenon that can last for several seconds, and can be connected to known perceptual mechanisms. As such, it supported the idea that change perception is an important part of perception, and that studying its failure can cast light on mechanisms central to our experience of the world. Conceptual Distinctions A clear understanding of CB – and its inverse, change perception – has been slow to emerge. This is partly due to the nature of change itself. Although this concept appears simple, attempts to formalize it over the years – from the earliest Greek thinkers to modern philosophers – have
Attention: Change Blindness and Inattentional Blindness
generally encountered difficulties. However, a number of important distinctions have become clear, which have helped our understanding of CB. Change versus motion The human perception of temporal variation is handled by two separate systems: a motiondetection system for variation in regards to a particular location and a change-detection system for variation in regards to structure. Temporal variations in the world activate both systems to some extent. For example, a moving automobile will activate motion detectors over the relevant part of the visual field, although these will not be able to determine whether there is an enduring spatiotemporal structure that remains constant. Conversely, a whirling dust devil will be seen as a coherent moving structure, even though there is nothing beyond the flow of dusty air in space. The key challenge in perceiving motion and change is to separate out the contributions of these two systems, so that temporal variation is assigned to the correct substrate. Change versus difference Another important distinction is that between change and difference. In both cases, reference is to a particular structure – an object or event of some kind. However, while change refers to transformation of a single structure, difference refers to the comparison of two or more separate structures. More precisely, change is based on the properties of the same structure at separate points in time, and can be perceived as a single dynamic visual event. For example, when a person is seen walking, their legs are seen at various positions over time. This kind of perception suggests – although does not prove – that the underlying representation has a spatiotemporal continuity of some kind. In contrast, difference is based on an atemporal comparison of structures that may or may not exist simultaneously, for example, comparing the height of two people. This kind of perception allows for the possible involvement of long-term memory elements that are only intermittently engaged. This distinction is important. Change detection and difference detection can be distinguished, at least conceptually; it is important to consider the type of experiment used when discussing either one. Spotting the difference between two side-by-side
49
stimuli may not engage the same perceptual mechanisms as detecting a change in successively presented stimuli. Failure to make this distinction can lead to erroneous conclusions being drawn about the mechanisms involved. Experimental Approaches and Results All studies to date have been based on much the same design: an initial stimulus is first presented (e.g., a picture), an altered version is then shown (e.g., some object in it is removed), and the ability of the observer to perceive the change is then measured (see Figure 1). Various aspects of perception have been explored in this manner based on particular selections among a relatively small set of design parameters. Type of task CB is the failure to perceive a change. Since there are several different aspects to perception, there can be – at least in principle – several different kinds (or levels) of blindness. These can be investigated by giving the observer the appropriate kind of task:
. Detection. This is the most basic and the most widely studied aspect, concerned with the simple noticing of a change. No properties of the change itself are necessarily perceived: the observer is simply asked to report whether it occurs. . Identification. This concerns reporting the properties of the change, that is, seeing what type it is (e.g., a change in color, a change in orientation). This can in principle be separated from detection: the observer could be asked to guess the type of change even if the change itself was not detected. Identification of change appears to be more difficult than detection, indicating that somewhat different mechanisms are involved. . Localization. This is concerned with reporting the location of the change. This can be decoupled from detection and from identification, at least in principle. Relatively little work to date has been done on this aspect of perception. Some results suggest that a separate memory for location may exist, although this has not yet been fully established.
50
Attention: Change Blindness and Inattentional Blindness
Type of response Another way to engage different mechanisms is to use different aspects of the response of an observer to a change in the external world. These effectively test different perceptual subsystems:
. Explicit percept. This is the approach used in most perceptual experiments, involving the conscious visual experience of the observer. A high degree of blindness can usually be induced. The proposal that attention is needed to see change is concerned with this type of experience. . Semiexplicit percept. Some observers can have a ‘gut feeling’ for several seconds that change is occurring, even though they do not yet see it (i.e., do not have a picture of it). The basis of this is controversial, but it may involve nonattentional or subattentional systems. . Implicit percept. Change that is not experienced consciously may still be perceived implicitly. If so, this must be measured by its effect on other processes. For example, some studies indicate that an unseen change can influence forced-choice guessing about its possible location. The existence of this form of perception is controversial, although evidence for it appears to be increasing. . Visuomotor response. This involves the response of a visually guided motor system to a change that is not consciously experienced. Systems used are almost always manual pointing or eye fixation. Both kinds of visuomotor response are faster than consciously mediated ones. They also appear to be more accurate, suggesting the existence of representations with highercapacity memory. Attentional manipulation A central tenet in change perception is that attention is needed to consciously experience change. In normal viewing, the local motion signals accompanying a change attract attention to its location, allowing the change to be seen immediately; this is why it helps to wave at a friend from across the room. If these local signals can be neutralized so that the automatic drawing of attention cannot help, a time-consuming attentional scan of the display will be needed. The observer will
consequently be blind to the change until attention is directed to the appropriate item. A variety of techniques have explored this: . Gap-contingent techniques. The change is made during a blank field or mask briefly displayed between the original and altered stimulus, which swamps the local motion signal (see Figure 1). Observers are very poor at detecting change if more than a few items are present; results suggest only 3–4 items can be seen to change at a time. . Saccade-contingent techniques. The change is made during a saccade of the eyes. Observers are generally poor at detecting change if more than a few items are present; again, a limit of 3–4 items is found. CB can also be induced for position change with even one item present, provided no global frame of reference exists. . Blink-contingent technique. The change is made during an eyeblink. Again, observers are generally poor at detecting such changes. Interestingly, blindness can be induced even if the observer is fixating the changing item. . Splat-contingent techniques. These make the change at the same moment as the appearance of brief distractors (or splats) elsewhere in the image. The blindness induced in this manner is relatively weak, but still exists, showing that it can be induced even when the change is completely visible. . Gradual-change techniques. Here, the transition between original and altered display is made slowly (e.g., over the course of several seconds). Observers generally have difficulty detecting such changes, even though no disruptions are used. The results from all these approaches are consistent with the proposal that attention is needed to see change. The finding that at most 3–4 items can be seen to change at a time is consistent with the capacity of attention obtained via other techniques such as attentional tracking. Perceptual set An important part of perception is the perceptual set of the observer, which strongly affects the mechanisms engaged for a given task. The issue of set is important for the question of which
Attention: Change Blindness and Inattentional Blindness
mechanisms are involved in everyday vision, and how these are related to performance as measured in the laboratory: . Intentional set. Observers are instructed to expect a change of some kind. They are assumed to devote all their resources to detecting the change, which provides a way to determine perceptual capacities. A further refinement is controlling expectation for particular types of change. Changes for presence and location are not affected by expectation of type, whereas changes for color are. . Incidental set. Observers are given some other task as their primary responsibility (e.g., count the number of sheep in an image); there is no mention of a possible change until after the task is over. The engagement of perceptual mechanisms is believed to be more representative of their use in everyday life. The degree of blindness found under these conditions is higher than under intentional conditions, indicating that relatively little is attended – or at least remembered – in many real-life tasks. Implications for Perceptual Mechanisms At one level, CB is an important phenomenon in its own right: among other things, it illustrates the extent to which we can potentially miss important changes in everyday life. Indeed, most people are unable to believe the extent to which they are unable to see change – in essence, they suffer from ‘CB blindness.’ However, CB itself can also be harnessed as a powerful tool to investigate the mechanisms by which we see. The exact conclusions obtained from such studies are still the subject of debate, but their general outlines are becoming clear. Visual attention and short-term memory All experiments on CB are consistent with the proposal that attention is needed to see change. Experiments on carefully controlled stimuli suggest that 3–4 items can be seen to change at a time, consistent with the limit on attention obtained via other techniques, such as attentional tracking. However, while a limit of some sort is involved, the nature of this limit is still somewhat unclear.
51
Contrary to subjective impression, change perception is not an elementary process. Instead, it involves – at least in principle – a sequence of several steps: enter the information into a memory store, consolidate it into a form usable by subsequent processes, hold onto this for at least a few hundred milliseconds, compare it to the current stimulus at the appropriate location, clear the memory store, and then shift to the next item. A limit on any of these steps would limit the entire process, making it difficult to determine the relevant step in a given situation. Most proposals for mechanisms have been couched in terms of either visual attention or vSTM. Both are similar in the results they cite and the mechanisms they propose. Part of this is caused by the extensive overlap that appears to exist between the mechanisms associated with attention and vSTM. Indeed, the difference may be largely one of terminology, caused by the vagueness in the definition of attention. As used in CB studies, attention is defined by the formation of representations coherent over space and time; such representations are not that different from those posited as the basis of vSTM. In any event, interesting new issues are emerging. One is whether the limiting factor applies to the construction of the coherent representation, or to its maintenance once it is formed, or perhaps both. Another issue is the nature of the elements that are attended and held in vSTM. Much evidence suggests that these are protoobjects that already have a considerable local binding of features; if so, the function of attention and vSTM would be to create a representation with extended spatial and temporal coherence. Results also suggest that the 3–4 items are not independent, but, rather, may have a higher level of interaction such that their contents are pooled into a single collection point (or nexus) that supports the perception of an individual object. Scene perception Given that only a few items may have a coherent representation at any time, our subjective impression of seeing everything that happens in front of us cannot be correct. But care must be taken in the particular inferences drawn. Since it deals only with dynamic quantities, CB cannot say anything
52
Attention: Change Blindness and Inattentional Blindness
about the static information that may or may not be accumulated. On the other hand, it does show the existence of severe limits on the extent to which changes are represented in those subsystems accessible to conscious perception. To account for the impression that we see all the changes that occur in front of us – not to mention that we can actually react to many of these – it has been proposed that scene perception is handled by a virtual representation. Here, coherent representations of objects – needed for change perception – are created on a ‘just in time’ basis, that is, formed whenever they are needed for a task and then dissolved afterward. Coordination of this process can be achieved via a sparse schematic representation of the scene – perhaps a dozen or so items, each with some properties – formed independently of the coherent representations. Guidance could be done on the basis of both high-level factors (e.g., schemas that enable testing of expected objects) and low-level factors (e.g., motion signals that draw attention to unexpected events). All results on CB are consistent with this proposal, and results from other areas (e.g., research on eye movements) appear consistent with this as well. The status of static scene information is still unclear. In principle, it could be accumulated to create a dense description that would match our subjective impression. However, such accumulation is unnecessary: a virtual representation could handle most if not all aspects of scene perception. Furthermore, no results to date have clearly shown storage of information beyond the relatively sparse information used for guidance and the contents of attention and vSTM. Some information about the prior state of a changed item appears to be stored in a longerterm memory not used for the perception of change; the amount of this is information not known. More generally, many scenes might be stored in long-term memories of various kinds, but the information density of each could still be quite low – perhaps a relatively limited amount of information from each of a dozen or so locations. Whether the dynamic element is total or partial, it nevertheless plays an important role in each individual’s perception of a given scene. CB has therefore been useful as a tool to investigate individual differences in perception: the faster a change is detected, the more important it is deemed to be.
Studies have shown an effect of training – including culture – on the encoding of objects and the importance attached to them, along with an influence of the particular task undertaken. There is also an emerging connection here with the design of interactive visual interfaces: these owe much of their success to the engagement of these dynamic mechanisms, and are therefore subject to many of the same considerations regarding operator and task.
Inattentional Blindness Background It has been known for thousands of years that people engaged in deep thought can fail to see something directly in front of their eyes. Such blindness can be easily induced when an observer intensely attends to some event, for example, waiting to see if an oncoming automobile will stop in time. Under such conditions, much of the visual field can effectively disappear from consciousness, even if it contains objects that are highly visible. It was long believed that such blindness might be due to the image of an unseen object falling onto the periphery of the eye, which is relatively poor at perceiving form. But decades of research have shown that location is unimportant – the key factor is attention. One of the earliest lines of research that encountered this IB was the study of head-up displays for aircraft pilots, which superimpose instructions and information over a view of the external world. Simulator studies in the early 1960s discovered that when pilots carried out difficult maneuvers requiring attention to outside cues, they did not see the unexpected appearance of instructions to alter course. Later studies showed the converse effect: if pilots focused on the displayed instructions, they often failed to see highly visible aircraft unexpectedly rolled out onto the runway in front of them. The first perceptual studies to investigate this effect took place in the 1970s. These used selective looking tasks, where observers were presented with two superimposed visual events (e.g., a sports game and the face of someone talking) and asked to attend to just one of these. Results showed that observers could easily report what happened in the attended event, but did not do well in the other:
Attention: Change Blindness and Inattentional Blindness
they often missed large unexpected events, such as the appearance of a woman carrying an umbrella. This happened even when the unseen stimulus was in the center of the visual field, showing that the cause of this blindness effect was not optical, but attentional. Although this effect was surprising, the use of superimposed stimuli caused it to be considered somewhat artificial, and interest in it never really developed. But in the 1990s a new wave of studies did manage to kindle interest. These differed from the earlier studies in several ways. First, they used opaque rather than superimposed (or transparent) stimuli, greatly reducing concerns about artificiality. Indeed, it was found that even large unexpected events – such as the appearance of a human in a gorilla suit – were still not noticed by most observers under these conditions. Second, in addition to videos of events, techniques were also developed based on brief-presented static images (Figure 2), which showed that IB could apply to both static and dynamic stimuli. Third, simple stimuli were often used in both the static and dynamic case, allowing more experimental control. And finally, there was greater examination of the
(a)
Non-critical trial
(b)
Critical trial
Figure 2 Example of a technique to induce IB. A static
test stimulus is presented while visual attention is focused on a primary task; here, the task is to report which of the two lines is longer. (a) Noncritical trial. The primary lines alone are presented for 200 ms, followed by a mask to prevent further processing. The observer must report which is the longer line. Two or three such trials are usually given to allow the observer to get into the appropriate mental state. (b) Critical trial. Along with the primary lines, an unexpected test stimulus is presented nearby. Blindness is measured by detection (or identification or localization) of the test stimulus. An additional round of trials is often presented, in which the test stimulus is now expected; this is believed to correspond to a condition of divided attention.
53
effects caused by the stimuli that were not seen consciously. Together, these developments showed that IB could indeed occur in everyday life, and that the mechanisms involved play a major role in everyday perception. Conceptual Distinctions Work on the theoretical and experimental aspects of IB have emphasized several conceptual distinctions. Some apply not only to IB but are also relevant to more general issues of awareness. Expression versus suppression Given that attention is needed to see something (i.e., consciously experience it), two general possibilities exist as to how it acts. First, attention might cause selected stimuli to be expressed consciously, for example, they enter awareness by being activated so as to exceed some threshold. In this view, attention does not just facilitate the conscious perception of a stimulus – it also enables it. On the other hand, attention might act to selectively suppress some of the stimuli, so that they vanish from the conscious mind. (In both cases, the status of unattended items is open: they might be weakly expressed in some way, or available implicitly.) It is worth noting that the two possibilities are not exclusive: it could be that both are used, with attention acting to express – or at least emphasize – some items while suppressing others. Restricted versus unrestricted effects A related issue is the extent of any expression or suppression. It was once believed that everything registered in the visual system was experienced consciously. However, given that implicit perception of various kinds has been found, an important issue is whether the selective effects underlying IB are restricted to conscious experience, or whether they spill over to other aspects of perception. For example, results show that attentional inhibition of a given location will increase the amount of blindness there. But does such inhibition attenuate the input strongly enough to also affect implicit processes at that location? Virtually all studies to date have assumed that selective effects are restricted to conscious experience, but this assumption has not been tested.
54
Attention: Change Blindness and Inattentional Blindness
Experimental Approaches and Results At the most general level, all studies of IB use the same basic approach. An observer is given a primary task that engages their visual attention, an unexpected test object (or event) then appears in the display, and the response of the observer to that appearance is then tested. Because expectation is an important factor, several trials without a test stimulus are generally presented first. Results are taken from the trial in which the test stimulus first appears (the critical trial). After this, the object is no longer entirely unexpected, and the blindness levels in subsequent trials are usually much lower, consistent with the idea that attention is now divided among the expected stimuli. Type of task As in the case of CB, several different aspects of perception can be distinguished, each of which can be tested by appropriate selection of task:
. Detection. This is the most basic aspect of perception, concerned with the simple noticing of something present; no properties of the stimulus itself need be involved. For isolated static items that are small, that is, have an extent of less than about 1 of visual angle, observers often fail to see anything at all in conditions that induce IB. For larger items, however, detection remains generally good. For dynamic events, informal observations suggest that the unattended event does not necessarily disappear entirely – observers often notice ‘something else’ going on, but cannot say what it is. . Identification. This is concerned with determining various properties of the stimulus. For example, observers can be asked to report the color or location of an item; in principle – although not often in practice – this can be done whether or not it was detected. Observers can usually report the color of a small item that is detected; identification of shape appears to be more difficult. Detection without identification has been reported, but it has been suggested that these are simply false positives. . Localization. Here, the task is to locate the test stimulus that appears in the critical trial. In principle, this can be tested independently of detection and identification, although this is
rarely done. Observers are able to report the location of a small item that is detected, showing that there may be some interaction between these two aspects of perception. Type of response Performance can also be studied in terms of the kind of percept experienced. Operationally, this can be done by testing different aspects of the observer’s response to the appearance of a stimulus. These different aspects effectively involve different perceptual subsystems:
. Explicit percept. This is the kind of percept used in most IB experiments. The observer is asked to respond to their conscious experience, that is, the picture they have of something. The usual definition of IB is in terms of the failure to have such an experience; the proposal that attention is needed to see is likewise concerned with this aspect. . Implicit percept. Items that are not experienced consciously may still be perceived implicitly. Given that the observer is blind to the stimulus, responses cannot be measured via direct subjective experience; they must instead be based on indirect effects. For example, relatively little blindness is found for emotionally laden words or pictures. This suggests that such words have been perceived implicitly, with attention drawn to them on the basis of their meaning. Similarly, a length illusion can occur in a set of lines perceived consciously, even if the background lines that induce the illusion are themselves not reported. . Motor response to primary task. The response to the appearance of a test stimulus can be measured by its effect on the speed or accuracy of the primary task. This is essentially the approach developed to study attention capture, which measures the effect of a new item on motor response times for the main task, regardless of whether the new item is seen. (An interesting variation would be to determine if such effects are conditional on the experiential state of the test stimulus.) . Visuomotor response. Here, the appearance of an item causes the eye to move toward it. This is another approach developed in the area of attention capture; again, this response is measured
Attention: Change Blindness and Inattentional Blindness
regardless of whether the item is seen. This might be useful for studying IB if measurement were made conditional on whether the test stimulus is consciously experienced.
Attentional manipulation An important factor in inducing IB is to present the test stimulus while the observer has their visual attention engaged on an unrelated task. This has been done in several ways:
. Superimposed stimuli. Two independent dynamic events (e.g., a basketball game and two pairs of hands playing a game) are presented simultaneously. In earlier studies, this was often done via half-silvered mirrors; more recent studies do this electronically. The result is perceived as a set of transparent or ghostly images. Observers can easily attend to one of these, but can often miss events in the other. . Interspersed stimuli. Here, two different sets of stimuli are presented; all are opaque and appear on the same display (see Figure 2). Tests on static stimuli generally use this method. Observers are asked to carry out a primary task on one of the sets (e.g., make a length judgment on a pair of lines); a test stimulus is presented near these lines a few trials later. Dynamic events have also been tested this way. A high degree of blindness can be found, even when the items in the two sets of stimuli are intermingled. . Dichoptic presentation. Here, two independent events are presented, each to a different eye; the observer is asked to pay attention to one of them. Events in the unattended eye often fail to be reported. In contrast with superimposed stimuli, where observers often have an impression of ‘something else’ going on, observers here fail to experience anything of the unattended set – it is simply not there. This may be related to the blindness experienced in binocular rivalry suppression. Perceptual set An important aspect of IB is the perceptual set of the observer. There are several ways this could influence performance:
55
. Control of selectivity. This is the extent to which selection – either expression or suppression – is invoked in the primary task. For example, observers can be required to attend only to the whiteshirted players in a game, while ignoring the black-shirted players. In such tasks, blindness appears to be due at least in part to observers suppressing the features of the ignored stimuli: the greater the similarity of the test item to the ignored items, the greater the blindness. In nonselective tasks, there is no need to screen out any stimuli, at least up to the critical trial. Blindness is still induced here, but it remains unknown whether this is due to a failure of expression or an invocation of suppression. . Control of capture. It has been suggested that high-level control may be exerted over the kind of stimuli that can capture attention, and thus be seen. Attention is usually drawn by the appearance of a new item or by the presence of a unique property. However, work on attention capture has shown that this can be overridden by a high-level attentional set. Results on IB are consistent with the proposal that an attentional set governs what is consciously seen, with distinctive stimuli experienced consciously only if they fit into the observer’s expectations. Implications for Perceptual Mechanisms The finding that observers can often fail to see highly noticeable objects and events touches on several issues concerning the way we cope with our world. For example, it suggests that we might not be aware of the extent to which we fail to see various aspects of our immediate environment, even if these are important and highly visible. (This might be termed IB blindness.) This has obvious implications for tasks such as driving, where the ability to accurately perceive objects and events is literally a matter of life and death, and where knowing about our limitations could well affect how careful we are. Meanwhile, work on IB can also tell us about the mechanisms involved in visual perception. In particular, it can provide a unique perspective on the mechanisms that underlie our conscious experience of the world.
56
Attention: Change Blindness and Inattentional Blindness
Visual attention All results to date are consistent with the proposal that attention is needed to see an object or event. For example, the degree of blindness has been found to increase with distance from the center of the attended location, in accord with space-based models of attention. In addition, the greater the attentional load of the primary task, the greater the blindness to the test stimulus, which is consistent with the proposal of a limited attentional capacity. Indeed, the inability of an observer to follow more than one coherent event supports the proposal that only one complex object or event can be attended at a time. Results also provide tentative support for the proposal that the high-level control of attention is achieved via an attentional set which determines the kinds of information that can capture attention, and perhaps also the kinds of information that can enter conscious awareness. This development potentially connects work on IB to work on attention capture. Another possible connection is with preattentive vision, usually studied by the rapid detection (or pop-out) of unique items in a display. Work here has pointed toward a considerable amount of processing achieved in the absence of attention. But this assumes that little or no attention is given to most items in a display at any moment. Results on IB suggest that this supposedly inattentional condition may be better viewed as a case of diffuse attention, with the inattentional condition characterized as one where no conscious perception exists. However, there are several findings that make these connections less than certain. For example, an observer in an IB experiment can usually detect the individual items in a group, although the grouped pattern itself cannot be identified. In addition, the pop-out of a unique item in a group occurs only if the group appears in the critical trial; if this group is shown earlier in noncritical trials (with the unique item the same as the others), pop-out no longer occurs. So what is the effect of the unique item here? If it is to draw attention to the group, why are the individual elements already seen in the earlier, noncritical trials? And why should pop-out occur in one condition, but not the other?
Scene perception Relatively little work has studied the aspects of natural scenes perceived under conditions of inattention. Studies using briefly presented scenes as test stimuli found that observers could usually report the scene gist (i.e., its overall meaning, such as being an office or a forest), along with several objects of indeterminate description. Some confabulation is also found. This is broadly similar to results on the ability of observers to rapidly perceive the gist of scenes from brief exposures of 100 ms or less, which is also likely done with little or no focused attention. Individual differences in the coding of scenes and events can be measured in terms of how blindness varies with the primary task. For example, experts in basketball could better detect the appearance of an unexpected object while they were attending to a basketball game. This suggests that they had encoded the scenes and events in a way that allowed them to divert some of their attention to occasionally monitor other items, without seriously affecting performance on the primary task. The model of scene perception often used to account for results on IB is the perceptual cycle. Here, sustained attention is believed to activate the conscious percept of a stimulus; once this has been done, the stimulus has entered the cycle, and helps select the appropriate schema to determine what information to admit next. In contrast, stimuli that do not become part of the predictive cycle may never be seen at all. Such stimuli – especially low-level signals – can guide the process, but do not enter awareness on their own. Similarly, information assembled at a preattentive level can also guide this process, although it too does not enter awareness automatically. This model has some similarities to the dynamic scene representation often used to account for CB, as well as the reentrant models used to account for conscious experience.
General Issues Change Blindness versus Inattentional Blindness CB and IB both involve a failure to perceive things that are easily seen once noticed, and both are
Attention: Change Blindness and Inattentional Blindness
believed to be due to a lack of attention. It is therefore likely they are related. But exactly how? First of all, the difference between CB and IB does not depend on the kind of input. Both can be found using dynamic images, and both can be found using static images. And the particular contents of the input do not matter for either. Instead, the critical difference between the two is the status of the information under consideration. IB is entirely concerned with first-order information – the simple presence of quantities. In contrast, CB involves second-order information – the transitions between these quantities. These can be separated: an alternating sequence of two images, say, could be experienced as a 50% presence of each image over time (same first-order distributions), but with different amounts of change (different second-order distributions) if the alternation rates are not the same. And just as CB can say little about first-order (static) information, IB can say little about second-order (changing) information. The two therefore refer to largely complementary aspects of the visual world. This distinction has consequences for the perceptual mechanisms involved. For example, the kind of attention required for each aspect of perception may be different – or at least, have different effects. The kind of attention involved in IB is space based: the degree of blindness increases with increasing distance from the center of attention. It is also easily diverted – hence the common use of a test stimulus that is completely unexpected. In contrast, the kind of attention needed to perceive change is object-based and is much less easily diverted (or at least slower), since telling the observer that a change will occur – and even giving them practice at perceiving it – does not affect performance to any great extent. Loosely speaking, IB might be identified with the absence of diffuse attention, and CB with the absence of focused attention. But a final determination of this must await a better understanding of attention itself. Blindness versus Amnesia An important issue in regards to the status of these induced failures is whether they are failures of perception or of memory. It might be, for instance, that an observer did experience something under
57
conditions that induced CB or IB, but then forgot it before they could make their report. If so, these effects would not be forms of blindness, but forms of amnesia. In the case of CB, the resolution of this issue is reasonably straightforward. Perception of change is often measured by asking the observer to respond to the change as soon as they see it; all that is needed to trigger this is a minimal conscious experience. When observers are asked to respond to a single change, only a few hundred milliseconds exist between its presentation and the initiation of the report (e.g., the pressing of a button). Thus, if the experience of change is forgotten in the absence of attention, it would be exceedingly brief and incapable of causing a response to be initiated. To all intents and purposes it would be as if it never existed, at least at the conscious level. (Note that if taken seriously, the possibly of such a fleeting perception would not be restricted to CB – it would apply to any failure of perception.) The situation for IB is more complex. In one sense, this issue was resolved by the finding that unattended – and thus unseen – items are indeed perceived, in that they can indirectly affect aspects of conscious experience. Entry into conscious experience itself, however, is not addressed by this. The observer is usually asked about a possible item only after the primary task has been completed, which allows several seconds to possibly forget it. And, the observer cannot be asked to prepare to respond to the test stimulus, since this sets up an expectation of the item, which severely diminishes the attentional diversion, and thus the degree of blindness. This issue has been grappled with in several ways. One is to present a highly surprising or meaningful item, and hope that the observer will spontaneously report it. However, this has not generally been successful: even if a human walks around in a gorilla suit or if an airplane is wheeled out onto the runway where the pilot is about to land, most observers still do not respond. It has been proposed that the observers do see the visual elements, but do not assign meaning to them. In essence, this phenomenon becomes one of inattentional agnosia. (This may explain the reports in some selective looking experiments, where observers see ‘something else’ going on, but do not know what it is.)
58
Attention: Change Blindness and Inattentional Blindness
Another perspective derives from studies of the neural systems involved. Patients with lesions to particular parts of the cortex can suffer conditions such as neglect and extinction, in which attention cannot be easily allocated to objects. Such patients, however, do not appear to experience forgetting – they simply do not report perceiving such stimuli, even when asked with the object in full view. In addition, functional imaging of the brains of normal observers shows that words are not consciously identified in the absence of attention, even if the observer is looking directly at them. These results make it highly likely that the failure is one of perception, and not memory. Visual Attention versus Visual Experience CB and IB can be regarded as two forms of the perceptual failure created by the diversion of attentional resources. They can be distinguished at the functional level by the type of information involved (second- or first-order information, respectively). They also appear to be distinguished by the type of attention involved (focused or diffuse) and the kinds of operations (e.g., comparison) associated with these. This division may correspond to the two modes sometimes proposed for conscious visual experience: an object mode associated with focused attention and a background mode operating as default. Beyond this, however, only partial and tentative conclusions can be drawn regarding the issue of how visual attention relates to conscious visual experience. In the case where attention of any kind is absent, there does not appear to be any conscious experience of stimuli (second-order quantities for focused attention; first-order quantities for diffuse). However, results still point to a considerable amount of processing being carried out. For example, work on IB indicates that unattended – and therefore unseen – items can influence the perception of attended items. Similarly, some models of CB posit low-level representations with a degree of detail and feature binding (proto-objects) that are formed in the absence of this kind of attention. It is worth pointing out that observers in IB experiments often report that they can detect something about the nonselected stimuli, even though they cannot always identify it. Importantly, this
kind of experience is found only in those experiments involving superimposed or interspersed stimuli; for dichoptically presented stimuli, there is a complete absence of perception of the nonselected event. This suggests that in the superimposed and interspersed conditions diffuse attention is given to nonselected events, with the main event given focused attention. If so, this would suggest that identification and localization may require more focused attention (or related resource), and that both diffuse and focused attention may be allocated simultaneously to different stimuli. This proposal would be consistent with work on CB. Focused attention is needed only for the perception of complex quantities such as change; background items not given focused attention might still be seen, but only in regards to detection and perhaps a limited form of identification based on relatively fragmented pieces of static items.
Conclusions Work on CB suggests that focused attention is needed for the conscious experience of change: without it, observers will be blind to even large changes, at least at the conscious level. Some ability to implicitly perceive change may exist; if so, this does not appear to require focused attention. Similarly, work on IB suggests that diffuse attention is needed to detect an unexpected object or event, that is, to see it as ‘something.’ Some results indicate that further (attentional) processing may be needed to more completely identify or locate it. There also appears to be some ability to pick up and process information about static items in the absence of any form of attention, even though such items are not experienced consciously. Beyond this, our current understanding is poor. More work is needed to expand our empirical knowledge of the basic phenomena. More work is also needed on the basic conceptual issues involved, in particular, on our understanding of terms such as attention and awareness. However, much exciting progress is being made on these fronts. And some of the most powerful sources of these new developments are the phenomena of CB and IB.
Attention: Change Blindness and Inattentional Blindness See also: Attention: Selective Attention and Consciousness; Neglect and Balint’s Syndrome; Perception: Unconscious Influences on Perceptual Interpretation.
Suggested Readings Brockmole JR and Henderson JM (2005) Object appearance, disappearance, and attention prioritization in real-world scenes. Psychonomic Bulletin & Review 12: 1061–1067. Irwin DE (1991) Information integration across saccadic eye movements. Cognitive Psychology 23: 420–456. Levin DT (2002) Change blindness blindness as visual metacognition. Journal of Consciousness Studies 9: 111–130. Luck SJ and Vogel EK (1997) The capacity of visual working memory for features and conjunctions. Nature 390: 279–280. Mack A and Rock I (1998) Inattentional Blindness. Cambridge, MA: MIT Press. Moore CM and Egeth H (1997) Perception without attention: Evidence of grouping under conditions of inattention. Journal of Experimental Psychology: Human Perception and Performance 23: 339–352.
59
Most SB, Scholl BJ, Clifford E, and Simons DJ (2005) What you see is what you set: Sustained inattentional blindness and the capture of awareness. Psychological Review 112: 217–242. Neisser U and Becklen R (1975) Selective looking: Attending to visually significant events. Cognitive Psychology 7: 480–494. Rensink RA (2002) Change detection. Annual Review of Psychology 53: 245–277. Rensink RA, O’Regan JK, and Clark JJ (1997) To see or not to see: The need for attention to perceive changes in scenes. Psychological Science 8: 368–373. Simons DJ (ed.) (2000) Change Blindness and Visual Memory. Hove, East Sussex, UK: Psychology Press. Simons DJ and Chabris CF (1999) Gorillas in our midst: Sustained inattentional blindness for dynamic events. Perception 28: 1059–1074. Simons DJ and Rensink RA (2005) Change blindness: Past, present, and future. Trends in Cognitive Sciences 9: 16–20. Wilken P (ed.) (1999) Mack & Rock: Inattentional blindness. Psyche 5–7. http://psyche.cs.monash.edu.au/symposia/ mack_rock/. Wolfe JM (1999) Inattentional amnesia. In: Coltheart V (ed.) Fleeting Memories, pp. 71–94. Cambridge, MA: MIT Press.
Biographical Sketch
Ronald A Rensink is an associate professor in the Departments of Computer Science and Psychology at the University of British Columbia (UBC). His interests include human vision (particularly visual attention), computer vision, and human–computer interaction. He has presented work at major conferences and in major journals on visual attention, scene perception, computer graphics, and consciousness. He received his PhD in computer science (in computer vision) from UBC in 1992, and was then a postdoctoral fellow for 2 years in the psychology department at Harvard University. This was followed by a position as a research scientist at Cambridge Basic Research, a laboratory sponsored by Nissan Motor Co., Ltd. He returned to UBC in 2000, and is currently a part of the UBC Cognitive Systems Program, an interdisciplinary program that combines computer science, linguistics, philosophy, and psychology.
Attention: Selective Attention and Consciousness P D L Howe, K K Evans, R Pedersini, T S Horowitz, J M Wolfe and M A Cohen, Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Balint’s syndrome – A neurological disorder caused by bilateral damage to the parietal cortex that results in the inability to perceive more than one object at a time. Binding problem – The difficulty of mentally conjoining features that belong to the same object. Blindsight – A neurological disorder, typically caused by damage to the primary visual cortex, that results in the patient being unaware of any stimuli located in a particular part of the visual field while still being able to detect them. Feature – An attribute of an object (e.g., its color). Illusory conjunction – An illusory combination of features from different objects. Receptive field – The region of the retinal image to which a particular neuron responds. Pop-out – In visual search, the situation where the speed or accuracy of target detection is independent of the number of distractors.
Introduction We are aware of the world around us, but not in a uniform fashion. We selectively attend to some stimuli and are consequently less aware of others. You are probably sitting, at present. If you selectively attend to the pressure of your posterior on the seat, you will become more aware of that sensation than you were a moment before. This article deals with the relationship between conscious awareness and selective attention. There are three possibilities. It could be that selective
attention has no effect on our conscious awareness. As the opening example makes clear, this is not a promising hypothesis and we will ignore it. At the alternative extreme, it has been proposed that conscious awareness is fully determined by selective attention – that we are conscious only of the current contents of attention. A more moderate position is that attention modulates awareness, but we have some awareness of unattended stimuli. This middle position reflects our view. At the outset, justifying this position is made difficult by the many uses of terms like ‘consciousness’ and ‘attention’ in common speech and technical writing. In this entry, we will restrict ourselves to the conscious awareness of visual stimuli, though the same questions arise in the other senses, between sensory domains and perhaps even when monitoring one’s own thoughts. We will often use the term ‘object’ – another term with a problematic definition. For example, consider an image of a face. One could consider the entire face to constitute a single object. Alternatively, one could consider the eyes, the nose, the mouth, and so on to be objects. Indeed, each of these objects could in turn be decomposed as you attend to, say, a pupil or a nostril. As there is no general agreement on what constitutes an object, we avoid the issue. Instead, we ask for the reader’s indulgence and use the term as a layman would – imprecisely. Colloquial speech tends to incorrectly treat ‘awareness’ (like ‘attention’) as a single entity. In fact, we can profitably distinguish between the type of awareness that accompanies attention and the type of awareness that seems to occur in the absence of attention. This is an old idea. In 1780, E´tienne Bonnot de Condillac asked his readers to imagine arriving at a chateau late at night. The next morning, you wake in a completely darkened room. Then the curtains are thrown open for just a moment on the scene out the window with its
61
62
Attention: Selective Attention and Consciousness
farms, hills, forest, and so on. Condillac argued that you would initially see something, perhaps just patches of color, throughout the scene, but you would be unable to identify what you were seeing until you had directed attention to different parts of the scene. Condillac’s patches of color are what we are calling awareness in the absence of attention. We contrast this level of awareness with that obtained from attending to one of those colored patches and consequently realizing that it represents, say, a meadow in the summertime. We will argue that in the absence of attention we can, at most, be aware of object attributes but not how they are related. For example, if an object is composed of a red vertical bar and a blue horizontal bar, then, in the absence of attention, we might be aware that there was a vertical bar and a horizontal bar and that there was red and blue. However, we would not know which bar was which color. To be able to relate (or ‘bind’) a bar’s color to a bar’s orientation requires that the bars be attended. To understand why this might be the case we need to consider the ‘binding problem.’
Feature Integration Theory, Object Recognition, and Awareness When we attend to an object, we usually feel that we are aware of multiple features of that object. For example, we might be aware of a round, red, revolving disk. That type of awareness requires that we ‘bind’ the roundness, redness, and motion to the same object. The neurons that analyze different attributes of an object are often located in different regions of the brain. Consequently, binding features together to form a coherent representation poses a problem. In a world filled with many objects, often in close proximity, how do we know that the red computed in this part of the brain goes with the motion analyzed in this other part? This issue is known as the binding problem. One proposed solution is that it is the act of attending to an object that allows different features of the same object to be conjoined and features from other objects to be excluded. Indeed, this may be the main function of selective attention. In principle, our brains could have been constructed in such a way that we would not suffer
from the binding problem. For example, the optic tectum of the common toad (Bufo bufo) contains a class of ‘fly-detector’ neurons that signal the location of small, moving black dots. The method effectively avoids the binding problem because the toad can detect the fly directly without having to first measure the fly’s individual attributes such as its motion, color, and size. Unfortunately, this method allows for the detection of only a small number of different types of objects, as it needs a dedicated group of neurons for each type of object that it is to detect. Human visual systems (indeed mammalian visual systems, in general) have a flexible ability to represent arbitrary combinations of attributes like color, size, orientation, and so forth. In order to understand the relationship of attributes, these visual systems have had to solve the binding problem. Our understanding of the structure and function of the visual system (Figure 1) is obtained from multiple sources including neuroimaging techniques like functional magnetic resonance imaging (fMRI) in human observers and more invasive neuroanatomical and neurophysiological methods performed mainly in animal models such as the cat and the monkey. Visual information flows from the retina to the lateral geniculate nucleus (LGN) of the thalamus. The LGN in turn relays that information to the primary visual cortex (V1) located on the rear surface of the brain, mostly inside the calcarine fissure. From here the pathway divides, with the dorsal and ventral streams being particularly important IT
MST
V4
MT V2 V1 LGN Eyes
Figure 1 A schematic representation of the ventral
(left) and dorsal (right) visual pathways. Also known as the ‘what’ and ‘where’ pathways. Abbreviations are explained in the text.
Attention: Selective Attention and Consciousness
subdivisions. The dorsal stream includes visual areas V1 and V2, the middle temporal (MT) area, and the medial superior temporal (MST) area. The ventral stream also includes areas V1 and V2 and then proceeds to area V4 and to areas in the inferior temporal (IT) cortex. As one progresses along either stream, the neural activity become less purely stimulus driven, more readily modulated by changes in the attentional state, and increasingly likely to mirror the reported conscious percept. The dorsal stream is often referred to as the ‘where’ pathway as it is particularly sensitive to spatial information. For example, the MT area is especially sensitive to the motion of an object. To a large extent, the impression of the object’s motion is closely related to the activity in this area. Electrical microstimulation of neurons in the MT in macaque monkeys influences their judgment of motion. Damage to the MT can cause akinetopsia, an inability to perceive motion. Sufferers of this condition can report that an object was in one position and is now in another. However, they have no conscious perception of the movement of the object. Conversely, for those that do not suffer from akinetopsia, it is possible to have a conscious percept of motion even when the visual stimulus does not move. For example, if one stares fixedly at a coherent moving pattern, such as a waterfall, and then fixates on a stationary object, the stationary object will appear to move (a motion aftereffect) and MT will be activated. If transcranial magnetic stimulation (TMS) is used to prevent the MT from activating, then the motion aftereffect is not seen. The ventral stream is often referred to as a ‘what’ pathway, as it is particularly sensitive to the identity of the object. For instance, activity in the IT closely reflects the subject’s impression of the object’s shape. This was elegantly demonstrated by David Sheinberg and Nikos Logothetis in a series of neurophysiological recordings in the macaque IT. First, they would isolate a neuron and find an image to which it responded strongly and one that did not excite it. Then, they would present one of these images to one of the monkey’s eyes while simultaneously presenting the other image to the other eye. The monkey had been trained to pull a lever to indicate which image it saw. As with humans, the percept reported by the monkey alternated between the two images, even though the
63
images themselves were constant, a phenomenon known as ‘binocular rivalry.’ They found that almost every IT cell responded only when the monkey reported seeing the image that had previously been shown to excite that cell. Crucially, these cells did not respond when the monkey reported seeing the cell’s nonpreferred image even though the preferred stimulus was still present on the retina of the other eye. The activity of these cells reflected conscious perception, as opposed to the unchanging retinal image. It should be stressed that the activity in the MT and IT does not cause the perceptual awareness of motion and shape, respectively. Indeed, when monkeys are rendered unconscious by an anesthetic, MT and IT continue to be active. Instead, it seems that when a monkey is conscious of an object, much of its awareness of motion and shape is reflected by the activity in these areas. For present purposes, the important observation is that when you see a moving object, its shape and motion are typically bound into a single coherent percept. In this physiological framework, the binding problem is the problem of understanding how motion information from the MT, shape information from the IT, and various other bits of information from other visual areas come to be unified in a bound percept. Feature Integration Theory Anne Treisman’s feature integration theory (FIT), first proposed in 1980, holds that attention is critical to the formation of bound representations of objects and, by extension, it proposes that attention is critical to our conscious experience of those bound representations. In FIT, following the understanding of the visual neurophysiology given above, the visual system first decomposes the visual scene into its composite features, arrayed in a set of ‘feature maps.’ The preattentive description of a scene or object comprises a list of such features. The term ‘preattentive’ has been controversial, but it can be operationally defined here as the representation of a stimulus before selective attention is directed to that stimulus. In FIT, the approximate position of each feature is recorded on its preattentive feature map. For example, if the visual scene contains two red
64
Attention: Selective Attention and Consciousness
objects, the feature map corresponding to redness would be activated at two points roughly corresponding to the locations of the red objects. If each feature were associated with a precise region in space, this might solve the binding problem. Features that correspond to the same region in space could be automatically conjoined, thus guaranteeing veridical perception. Unfortunately, the location of many features is measured in an imprecise fashion. For example, the smallest receptive fields in IT, the region whose activity correlates well with shape perception, have a spatial extent of a few degrees of visual angle. Within this region, the cell will respond to an object in an approximately translation invariant manner. Thus, a neuron in the IT cannot signal the location of a particular shape with a precision of better than a few degrees, while the perception of coherent objects requires a much finer resolution. Because of the poor resolution of these feature maps, if two objects are close together, then there is the potential that the features from one object may become conjoined with the features of the other object thus creating a percept of an object that did not in fact exist. For example, if the visual scene contains a red vertical bar and a blue horizontal bar then one might see a blue vertical bar and a red horizontal bar. Such inappropriate combinations of features are known as illusory conjunctions. FIT suggests that attention hinders the formation of illusory conjunctions. Supporting this assertion is a series of classic experiments by Treisman and her colleagues showing that, if attention is occupied elsewhere, illusory conjunctions are, in fact, reported. In one version of the experiment, observers viewed a display of five characters aligned horizontally. The outer two characters were always digits and the inner three characters were always letters. While the digits were always black, the letters were colored. The observer’s primary task was to name the digits. After doing that, the observer reported the letters and their associated colors. When the display was presented sufficiently rapidly, observers would often report seeing an incorrect conjunction of a color and a letter. For example, if the display contained a red X and a green T, they might report seeing a red T. Crucially, these illusory conjunctions occurred at a much higher rate than could be attributed to
the observer simply misperceiving a given feature. Generally, the observer correctly perceived the features present in the display. It was the conjoining of features that proved to be problematic. When asked to report how confident they were that they had actually seen an object, observers were just as confident when they reported seeing an illusory conjunction as they were when they correctly reported the features of an object. Indeed, although all observers were told that the digits would always be black (and in fact always were) about half the observers spontaneously reported that the digits sometimes appeared to be colored, sometimes even going as far as to argue with the experimenter about the issue! This raises an interesting problem in the study of attention and awareness. In tasks of this sort, one can only ask about what was seen, after the fact. If one asks about the current status of a visible object, the observer will attend to it in order to answer the question and will be unable to give an accurate report of the unattended state. Nevertheless, the phenomenology of illusory conjunctions does show that, within a fraction of a second of the disappearance of a display, observers can be quite convinced that they have seen something that was not, in fact, present. Subsequent studies have shown that illusory conjunctions can be perceived even when the subject attends to the objects, especially if the objects are perceptually grouped. Clearly, attention does not always succeed in solving the binding problem. There is neuropsychological evidence, from studies of patients with Balint’s syndrome, which supports the idea that attention can inhibit the formation of illusory conjunctions. This syndrome occurs when both the left and right parietal lobes are damaged. As these areas help govern the deployment of attention, such patients have great difficulty in directing their attention to a given object, resulting in the inability to perceive more than one object at a time. As would be expected, they are also prone to suffer from illusory conjunctions, experiencing them even when the image is displayed for several seconds. Neurophysiological support also comes from work by Robert Desimone and colleagues. They performed a series of extracellular studies in area V4 of the macaque monkey that have shown that attention can help solve the binding problem.
Attention: Selective Attention and Consciousness
First, they would find a stimulus that, when presented on its own, would elicit a strong response from the neuron in question (the preferred stimulus), and another that would elicit only a weak response (the nonpreferred stimulus). They would then present both stimuli simultaneously so that both were within the neuron’s receptive field. In the absence of attention, the cell would simultaneously respond to both stimuli, with its response (spike rate) lying between that generated by each stimulus when presented on its own. In other words, the response reflected contributions from both stimuli, meaning that the cell could not distinguish between the two. However, when the monkey attended to one of the stimuli, the situation changed and the cell responded primarily to the attended stimulus. Specifically, when the monkey attended to the preferred stimulus, the cell would respond strongly, but when the nonpreferred stimulus was attended, only a weak response was elicited. In this case, attention is able to solve the binding problem, at least at the neuronal level, by shrinking the receptive field of the cell to include just the selected item, thereby removing the influence of the unattended item. This constriction of the receptive field does not explain how signals about one feature analyzed in one cortical area can be bound to signals about another feature from another area. Other mechanisms have been suggested to account for this aspect of binding. Several of these are based on the idea that neurons in different cortical areas that respond to the same object synchronize their activity, so that they create action potentials at the same time. Consequently, a third brain area could determine whether two neurons in two different parts of the brain are responding to different features of the same object by being sensitive to this synchrony. As attention is known to increase neural synchrony, theories based on synchrony are consistent with the notion that attention is needed to solve the binding problem.
65
that might be available to be bound. Various tests have been proposed, most of them based on the premise that the attributes on this list can be analyzed in the absence of attention. For example, if an item is the only item in a display that has a particular feature, that item will tend to ‘pop-out’ of the display, summoning attention (as long as the other items are not too similar to that target item nor too different from each other). The set of items that pop-out in this manner is one definition of features. Figure 2(a) shows a case where the target has a unique feature. It is the only red object in the scene. Consequently, it pops-out and can be located very quickly, independent of the number of other items. Conversely, in Figure 2(b), the target and distractors share the same features. The target is a rectangle that is green on the left but red on the right, whereas the distractors are red on the left and green on the right. In this case, finding the target is a slow process. Texture segmentation is another test. Consider two regions of a display, one with a putative feature and the other without. If a border between those regions can be effortlessly detected, one could declare that there is a feature difference that permits the segmentation. Unfortunately, these and other methods for identifying features agree imperfectly. It is quite clear that some attributes, like color, motion, and orientation pass all the tests. Other attributes (e.g., various aspects of form) are more problematic. Feedforward Models of Object Recognition Even if attention is needed for binding, it is not necessary – and probably incorrect – to hold that
Features While it is easy to say that the visual system decomposes a visual object into its constituent features, it is harder to be precise about what this statement might mean. In particular, there is imperfect agreement about the list of features
(a)
(b)
Figure 2 Two visual search experiments. (a) The
target (the red rectangle) is easy to find. (b) The target (the green-on-the-left-red-on-the-right rectangle) is hard to find.
66
Attention: Selective Attention and Consciousness
attention is always needed for object recognition. A class of feedforward models of object recognition shows how some quite sophisticated object recognition could occur without explicitly invoking selective attention. As we will see later, this fits well with evidence that humans have some ability to detect objects in one part of the field even when their attention is occupied elsewhere. For example, Maximillian Riesenhuber and Tomaso Poggio developed a theory of object recognition based on an idealization of the hierarchy of the monkey visual system. As one progresses through the monkey visual system, cells become selective for increasingly complex visual stimuli. For example, in the LGN some cells have an on-center offsurround receptive field organization. The stimulus that most excites these cells is a spot of light on a black surround. On entering the primary visual cortex, we find cells whose optimum stimulus is more complex, perhaps a bar of a particular orientation and length. In V2, there are cells whose optimum stimulus is two bars in a particular configuration. In V4, some cells are most excited by a collection of bars joined together in a particular manner. Riesenhuber and Poggio were able to build a feedforward model that was able to explain how these selectivities were generated. Crucially, this was achieved without invoking any feedback mechanisms. Since attention must be mediated by feedback, they argued that this showed that at least some recognition can occur in the absence of attention. The original model was applied to shapes that resembled bent paper clips. In subsequent work, they and others have developed models that can recognize objects like cars and faces. Does this mean that attention is unnecessary? Probably not, since the models tend to fail when there are many objects in the display. These models do show that it is possible, in theory, to have some degree of recognition without attention. This, in turn, makes it plausible that one might have some awareness of an object, even if that object is not the target of selective attention. Reverse Hierarchy Theory Reverse hierarchy theory (RHT), proposed by Shaul Hochstein and Merav Ahissar, is an example of a model combining feedforward and feedback
components. The feedforward component is similar to the Riesenhuber and Poggio model. It explains how the hierarchy of the visual system allows for visual scenes to be processed to some degree in a feedforward manner in the absence of attention. It is the feedback component that allows for more detailed perception to occur (hence reverse hierarchy). For example, it suggests that to appreciate fine differences in orientation, the brain would deploy feedback from a high-level representation of the stimulus, in order to pay attention to the detailed orientation information held in cells in the early visual cortex. One interesting aspect of this proposal is that it suggests that the high-level information (e.g., animal or face) might reach awareness before information about low-level features. To summarize the argument to this point; precise binding of features to objects is a problem that the visual system seems to solve by the use of selective attention. It follows that attention is required for awareness of those bindings and for awareness of object identities that rely on those bindings. However, the example of Condillac’s chateau indicates that there will be awareness of something in regions not yet visited by selective attention. Moreover, feedforward models show that, at least in theory, the unattended awareness of something need not be limited to raw local features. Some quite sophisticated analysis and awareness might be possible away from the current focus of attention. It is to that awareness without attention that we turn to in the next section.
The Relationship between Attention and Awareness In this section, we wish to distinguish between the hypothesis that some awareness occurs outside of the current focus of attention and the hypothesis that we are aware only of the current contents of attention. Recall the phenomenon, described above, of illusory conjunctions in which observers correctly report the colors and letters in a display but fail to correctly report which letter goes with which color. It could be that the experience of unbound colors and letters represents awareness without attention. However, there is a contrary
Attention: Selective Attention and Consciousness
point of view. Perhaps the imperfect awareness of the letters arises from imperfect attention to the letters. There is no guarantee that naming the digits in an illusory conjunction experiment withdraws all attention from the letters. Perhaps if all attention had been really withdrawn from the letters then the observers would not have been able to report any features of the letters at all. That would be the prediction if awareness cannot occur in the complete absence of attention. When Awareness Requires Attention Although we will argue against this extreme viewpoint, we will discuss a set of phenomena that have been used to argue for this strong link between attention and awareness: inattentional blindness, change blindness, and the attentional blink. These topics are more extensively discussed in other articles of this encyclopedia. Inattentional blindness was first described by Arien Mack and Irv Rock. They had observers performing an attentionally demanding perceptual task (e.g., which of the two lines is longer?). On one critical trial, the briefly presented display contained an unexpected item. Observers were frequently unable to report that it had been presented. Any awareness of that item left no trace that could be reported after it was gone. Perhaps, attention to the primary task, prevented irrelevant items from ever rising to conscious awareness. As an experimental tool, one problem with this task is that it produces only one trial per observer. Once you ask about the unexpected item on one trial, other unexpected items on other trials tend to be successfully reported. Change blindness is a more resilient phenomenon. While the phenomenon was initially discovered in the late 1950s and early 1960s, a major renaissance on the topic emerged in the mid1990s. Dan Simons and Dan Levin, as well as Ronald Rensink and Kevin O’Regan and their colleagues presented observers with complex natural scenes (e.g., a photo of an airplane on the tarmac) and measured the ability to detect fairly large changes to these scenes (e.g., the plane’s engine disappearing and reappearing). Critically, the visual transient generated by the change was masked by an eye movement, a brief blank interval,
67
or some other visual transient. Observers thus had to actually detect the change in the image, rather than just the transient caused by the change. Change detection under these circumstances turns out to be very difficult. Observers could fail to notice changes even though they spent many seconds examining both versions of the display. If an object was attended during the transition between two frames, the change could be noted. Otherwise observers were unable to report it. One interpretation of these data would be that the observer was only truly aware of the currently attended object, while the apparent awareness of the rest of the display was, in some sense, an illusion. The attentional blink, originally described by Donald Broadbent, and later characterized by Jane Raymond, Kim Shapiro, and Karen Arnell, is a quite different phenomenon that might point to a similar conclusion. In a typical attentional blink experiment, observers monitor a stream of images, letters, for example, appearing at fixation, at a rate of one every 100 ms. The observers are looking for particular targets, say ‘E’ and ‘X.’ At this rate of presentation, an observer can easily report a single target letter appearing anywhere in the stream. However, if there are two targets, the second one is much more likely to be missed if it appears 200–500 ms after the first. This is not simply a matter of perceptual masking from the first target to the second. Given the same stream of letters (e.g., ‘J W E B P X L’), the target ‘X’ will be easily detected if the observer does not have to report the ‘E,’ but will likely be missed if she does. Something about the attention to the first target causes an ‘attentional blink’ that makes it harder or impossible to report the second. Interestingly, ‘blinked’ items can be shown to have effects on the observer. A ‘blinked’ word can produce semantic priming effects indicating that it has been read. Perhaps the type of attention that is tied up by the first target in an attentional blink display is the type of attention that permits awareness of an object. Phenomena such as inattentional blindness, change blindness, and the attentional blink provide some evidence that in the absence (or, at least, near-absence) of attention, the observer may be unable to recognize or even see an object. In its strongest form, this argument proposes a tight
68
Attention: Selective Attention and Consciousness
linkage between selective attention and visual awareness. The fact that in some situations an observer is unable to report an unattended object does not prove that she is never able to do so. In the following, we describe phenomena that indicate that the proposed strong linkage of attention and awareness is too strong. The linkage of attention, binding, and awareness is challenged by studies that seem to show that there can be some degree of recognition and awareness of objects in the near-absence of selective attention. When Awareness Does Not Require Attention In a series of experiments done by Fei-Fei Li and her collaborators, observers performed an attentionally demanding visual search task in one location while concurrently monitoring another portion of the visual field for some class of targets (e.g., animal or vehicle). The goal was to tie up selective attention with the search task and to determine what, if anything, could be detected elsewhere at the same time. Some tasks (e.g., determining if a red square was to the right of a conjoining green square or vice versa) are profoundly disrupted when attention is thus engaged. Interestingly, however, detection of the presence of animals or vehicles in a briefly presented scene is no worse when selective attention is occupied than when it is not. Note that detecting that a scene contained an animal is not the same as determining exactly what that animal was. Observers in Fei-Fei Li’s experiments were not necessarily sure what type of animal they had detected or where it was in the display. Moreover, when Karla Evans and Anne Treisman asked the observers to find animals in a rapid sequence of scenes, they found that the observers were significantly impaired if the stream also contained humans. In the near-absence of attention, some image statistics seem to permit awareness of the presence of an animal (human or other) or a vehicle, but it would be going too far to argue that these data make the proposed feature-binding role of attention unnecessary. Awareness of this specific animal or vehicle, which would presumably require feature-binding, appears to require attention.
A number of other phenomena also challenge the attention-binding-awareness linkage. For instance, Mary Potter and others have shown that highorder representations (i.e., gist) can be accessed very rapidly from natural scenes presented at rates of up to 10 per second, far too little time for selective attention to be directed to more than a small handful of items in that scene. ‘Gist’ in this case refers to a broad categorical label for the scene: beach, kitchen, and so on. Does this challenge the relationship of attention to binding to object recognition? It certainly would if recognizing the gist of a scene involved promiscuous binding of multiple objects without attention. Alternatively, these results might be demonstrating that performance of this scenecategorizing task does not require binding. If information in the unbound feature statistics could support performance of the task, then it would not be necessary to assume binding. Aude Oliva and Antonio Torralba have shown that this can be done, in principle, for scenes. They have created filters whose output can be used to put a label on a scene (indoor, outdoor urban, beach, etc.) based on the raw image statistics of the scene; without the need to parse the image into objects, regions, and so forth. This ‘unbound’ analysis is not adequate to identify a specific scene, for example, Crane’s Beach in Ipswich, MA, but it could provide the gist of a scene in the near absence of attention. As with the detection of animals with selective attention occupied elsewhere, findings of this sort suggest that awareness of a visual stimulus is not a unitary, all-or-none experience. Awareness of an unattended scene may be different than awareness of a well-attended scene, but the scene is seen in both cases. This distinction may be reflected in neurophysiological findings showing that categorization and perception are mediated by different cortical areas. Specifically, those neurons that can categorize a target are not necessarily the same neurons that can, say, locate it in the image. Neurons in the visual cortex are able to signal a target’s location, but are generally insensitive to whether an object is a target or not (i.e., they cannot categorize it). However, David Freedman and colleagues have elegantly demonstrated that neurons in the prefrontal cortex can be highly sensitive to the categorical status. They used a photographic morphing
Attention: Selective Attention and Consciousness
technique to create a picture of an animal that represented a combination of a cat and a dog. By varying the morphing parameters, they could vary the similarity between this computer-generated animal and prototypical cat and dog images. They presented a series of these computer-generated pictures to monkeys that had previously been trained to indicate whether each picture more closely resembled a cat or a dog. They found cells in the lateral prefrontal cortex that encoded the monkey’s categorization. Such cells responded similarly to images that belonged to the same category, even when the images appeared very different. Conversely, the cells responded very differently to images that appeared very similar, but which belonged to different categories. These cells therefore responded to the categorization of the images, as opposed to the visual image itself. We have seen that a strict linkage of attention, binding, object recognition, and awareness leads us to a theory that does not have room for the full range of phenomena. Still, it seems likely that attention is required for the recognition of specific objects and that, as Condillac argued, this act of attention changes the state of our visual awareness. Not being privy to more recent developments, Condillac does not tell us what he thinks we would see if we were performing an attentionally demanding task at fixation when the curtains were thrown wide, revealing the scene outside the chateau for the first time. However, it seems likely that his answer would have been much the same. You would have some impression of the outside world, but you would not understand what you were seeing until you had attended to the scene.
69
As he became familiar with the current literature, Condillac might agree that some information about the gist of the scene might be available in that first moment, but his description of the initial state of awareness, modified by subsequent attention, would remain essentially unchanged. When Attention Does Not Imply Awareness It is hard to gain any introspective access to this with real scenes because we are too good at analyzing them. However, the very unreal scene of Figure 3 may serve the purpose. Note that when you first look at this scene, you are aware of the patches of color and orientation across the entire stimulus. More than that, you are aware of some structure in the scene. There are plusses everywhere with a scattering of items with more than four line terminations. There are blue and yellow objects in the upper left and red-green elsewhere. However, if you are asked to detect red vertical components, you will need to direct your attention to specific items over a period of time and, having deployed your attention, your awareness of the stimulus will change. Now you will find that the red and green plusses are not all the same. There is a region of red verticals in the upper right and an isolated example at the bottom center. You have awareness with and without selective attention. What about attention without awareness? Can you select and bind an object without being aware of it? Returning to the figure, imagine you are asked to locate the five-pointed item. It is entirely possible that you had already attended to
Figure 3 A display that gives some idea of one’s level of visual awareness. Please see the text for details.
70
Attention: Selective Attention and Consciousness
that object during a search for red vertical without becoming aware of it. Thus, it seems possible that you can attend to an object without ever becoming aware of it. The point is tricky since one could argue that you were aware of the five-pointed item when you putatively attended to it, but you forgot about it prior to being queried about it. It is often difficult to distinguish between having been unaware and being amnesic. Studies of blindsight patients provide converging evidence for the hypothesis that attention is not always sufficient for visual awareness. In blindsight, damage to the primary visual cortex results in a condition where patients report being unaware of part of the visual field (the unaware area). Interestingly, when asked to guess what stimuli are located in the unaware area, patients can perform at above chance levels. Furthermore, when spatially cued to the unaware area, they exhibit speeded discrimination of targets that subsequently appear in that area, demonstrating that they can attend to stimuli that they cannot see. Evidently, attention does not necessarily result in awareness. Postattentive Awareness Consider the red-vertical plus at the bottom center in Figure 3. We can posit that you had some representation of it before it was selected. That representation can be called ‘preattentive.’ You also had a representation of the plus while it was attended. Call that an attended representation. What is the representation of that plus when you then move your attention to the blue-vertical star, up and to the right? This can be called the ‘postattentive’ representation. A series of experiments show that changes to the plus (e.g., changing it to red-horizontal, green-vertical) once attention has been shifted elsewhere, will go unnoticed, until attention is directed back to the item (providing that the transients produced by such changes are masked by, say, a blink, a saccade, or a visual transient). Observers show no more awareness of the current binding of features in a postattentive object than in a preattentive object. At the same time, you – the observer – are aware that there was a red-vertical plus at that location so, in that sense, your postattentive awareness of that particular plus is different than your preattentive awareness.
Of course, you would also be aware of the plus, in that sense, if the lights went out and you relied on memory. The already difficult topic of visual awareness becomes more difficult once we admit a role for memory. Your awareness of a familiar face is different from your awareness of a new face. That difference is tied to memory and it is hard to know whether one should consider this to be part of the definition of visual awareness. Whatever one concludes from this question, the postattentive vision research suggests that selective attention affects postattentive awareness of an object through memory and not through some sort of persistent binding that continues once selective attention is disengaged from an object. Awareness of Awareness The account outlined above suggests that we think we are aware of more than we actually are. We greatly overestimate our own awareness. Our naı¨ve impression of our visual awareness is that we are aware of a large visual scene at a high resolution. Yet this is not so. At any moment, the only part of the visual scene we can see in high resolution is the small area around the current point of fixation. A particularly striking demonstration of this (beloved by people who sell eye trackers!) is to use an eye tracker to monitor the observer’s point of fixation. The observer’s task is to read some text presented on a computer monitor, similar to that shown in Figure 4(a). At some point, the eye tracker salesperson pushes a button and, during the observer’s next saccade, the letters in the display become jumbled except for those in the words near the point of fixation (Figure 4(b)), which, in this figure, are assumed to be in the top-left corner. Every time the observer saccades to a different point, the letters in the words near that point become unjumbled, while the letters in all other words either become or remain jumbled. Provided all changes to the display occur during a saccade, the observer is unaware of the scrambling. She reports that she is simply reading a normal text. Similarly, if the image away from fixation is appropriately blurred, an observer would be unaware of this degradation and will have the impression of looking at the usual, apparently well-focused scene.
Attention: Selective Attention and Consciousness
Some random text for the observer to read
some random xtet ofr het ujsecbt ot edra
Some random text for the observer to read
some ranmod xtte rfo hte jsbcute ot raed
Some random text for the observer to read
oems narmdo xett rof eht bjsteuc ot ader
(a)
71
(b)
Figure 4 The observer starts to read the text shown in (a). During an eye movement, all the text that is not near
the observer’s point of fixation (assumed to be at the top-left corner) becomes jumbled (b). The observer does not notice the change and so cannot differentiate between (a) and (b), thereby demonstrating how limited visual awareness really is.
The observer in these illustrations is aware of something, but it turns out not to be a ‘true’ assessment of the contents of the current visual representation. Beyond a simple contribution of memory to awareness, this indicates a contribution of theory to awareness. In the reading example, the observer is not aware of the scrambling occurring away from the fixation. Wherever she fixates on the page, the letters form themselves into English words. It is a reasonable theory that the page consists of readable English words and our observer’s awareness incorporates that theory. Her visual awareness is affected by what she thinks she knows. Returning to the role of attention in awareness, we see that selective attention alters not only the awareness of the attended object, but potentially, the awareness of other unattended objects, potentially divorcing that awareness from the actual perceptual facts. Awareness of Attention If one were inclined to propose a tight linkage of attention and awareness, one might propose not only that observers are only aware of the objects of attention, but that observers are aware of all of the deployments and, thus, all the objects of attention. However, in visual search experiments, it is estimated that observers can attend to 20–50 items per second. This rapid selection seems to occur without a clear awareness of which items in a display have or have not been selected. At least observers do not use any such awareness to guide their search. For example, Todd Horowitz and Jeremy Wolfe conducted a visual search experiment in
which all the items in the display were randomly relocated every 111 ms. This made it impossible for observers to keep track of which items they had already attended to. Surprisingly, the search was no less efficient in this case than in the control condition where the items were static. This showed that, at least in some cases, visual search had no memory. Observers acted as if they were unaware of what they had attended to.
Conclusions In this article, we have considered the relationship of conscious awareness and attention from the perspective of vision. Following Condillac, we found it helpful to differentiate between the awareness that results from attending to an object or group of features and that which occurs in the absence of attention. Condillac is more famous for his statue than for his chateau. He asked his readers to imagine the mental life of a statue with no senses. In his honor, we can imagine a statue with senses, but without attention or, perhaps better, with attention disabled. In the absence of attention, the evidence indicates that our statue would retain some visual awareness, but would be unable to form any percepts that would require the binding of two of more features. This level of awareness might allow our statue to classify scenes (beach, mountains, etc.) or declare that they did or did not contain an animal, but would not allow it to determine specifics such as which animal occurred in a given scene. If we now endow this statue with selective attention, it can then solve the binding
72
Attention: Selective Attention and Consciousness
problem, and have a more complete awareness. Specific objects can be selected, perceived, and identified. If we now allow the statue to have a memory, the statue will know that a given object was at a particular location and, in the absence of contradicting information, the statue is likely to assume that the object continues in that location. If we add a theory-building capability, the statue can generalize from the fact that all selectively attended samples are seen in sharp focus to the assumption that all objects really are in sharp focus and then use this assumption to modify visual awareness accordingly. The statue now has an approximation of human visual awareness. See also: Attention: Change Blindness and Inattentional Blindness; Mind Wandering and Other Lapses; Neglect and Balint’s Syndrome; Neuroscience of Volition and Action.
Suggested Readings Evans K and Treisman A (2005) Perception of objects in natural scenes: Is it really attention free. Journal of Experimental Psychology: Human Perception and Performance 31: 1476–1492.
Freedman DJ, Riesenhuber M, Poggio T, and Miller EK (2001) Categorical representation of visual stimuli in the primate prefrontal cortex. Science 291: 312–316. Freedman DJ, Riesenhuber M, Poggio T, and Miller EK (2003) A comparison of primate prefrontal and inferior temporal cortices during visual categorization. Journal of Neuroscience 23: 5235–5246. Hochstein S and Ahissar M (2002) View from the top: Hierachies and reverse hierarchies in the visual system. Neuron 36: 791–804. Horowitz TS and Wolfe JM (1998) Visual search has no memory. Nature 394: 575–577. Kentridge RW, Heywood CA, and Weiskrantz L (2004) Spatial attention speeds discrimination without awareness in blindsight. Neuropsychologia 42: 831–835. Li FF, VanRullen R, Koch C, and Perona P (2002) Rapid natural scene categorization in the near absence of attention. Proceedings of the National Academy of Sciences USA 99: 9596–9601. Moran J and Desimone R (1985) Selective attention gates visual processing in the extrastriate cortex. Science 229: 782–784. Potter MC and Faulconer BA (1975) Time to understand pictures and words. Nature 253: 437–438. Riesenhuber M and Poggio T (1999) Are cortical models really bound by the ‘binding problem. Neuron 24: 87–93. Sheinberg DL and Logothetis NK (1997) The role of temporal cortical areas in perceptual organization. Proceedings of the National Academy of Sciences 94: 3408–3413. Treisman AM and Gelade G (1980) A feature-integration theory of attention. Cognitive Psychology 12: 97–136. Treisman A and Schmidt H (1982) Illusory conjunctions in the perception of objects. Cognitive Psychology 14: 107–141.
Biographical Sketch
Piers Howe graduated as an exhibitioner from Oxford University in 1998, with a masters in physics. Winning a Presidential University Graduate Fellowship, he obtained his PhD from Boston University in 2003, under the guidance of Stephen Grossberg. His PhD thesis was titled ‘Cortical mechanisms of depth and lightness perception: Neural models and psychophysical experiments.’ He then worked as a Helen Hay Whitney postdoctoral fellow with Margaret Livingstone at Harvard Medical School before moving on to Brigham and Women’s Hospital to work as a research fellow with Todd Horowitz and Jeremy Wolfe. His research has involved a variety of techniques including computational modeling, macaque neurophysiology, human fMRI, and human behavioral experiments. He has published articles on lightness perception, motion perception, depth perception, and
Attention: Selective Attention and Consciousness
73
visibility. More recently his focus has shifted to visual attention and to devising computational techniques for determining brain connectivity from fMRI data. He is a member of the Harvard fMRI Center for Neurodegeneration and Repair. He has taught a psychology course at the University of Massachusetts, Boston, MA and a cognitive modeling course at Boston University. He lives in Boston, MA.
Karla Evans graduated summa cum laude from Trieste University, Italy, in 1998, with a degree in psychology and went on to obtain her PhD in 2007, from Princeton University, studying with Anne Treisman. Her PhD thesis was entitled ‘Crossmodal interactions of corresponding auditory and visual features.’ Evans then became a postdoctoral associate at the McGovern Institute of Brain Research at MIT. During that period, she published papers on natural scene perception, crossmodal correspondences, and affective person knowledge and face perception. She is currently a research fellow working with Jeremy Wolfe and Todd Horowitz. Karla Evans was a recipient of a science/engineering fellowship, Princeton University and the Summer Institute Fellowship in Cognitive Neuroscience at Dartmouth College. She lives in Somerville, MA.
Riccardo Pedersini graduated cum laude from the University of Padova in 2001, with a degree in experimental psychology. He continued in Padova and Amsterdam obtaining his European masters in mathematical psychology in 2003, and his PhD in perception and psychophysics in 2006, from the University of Padova. His PhD was obtained under the guidance of Prof. Elisabetta Xausa and Prof. Marc Le Menestrel. He visited the University Pompeu Fabra in Barcelona and Sussex University in Brighton, and he taught several graduate courses at the faculty of medicine in Padova. He was a two-time regional winner of the Mathematics Olympiad. In addition to two Marie Curie fellowships, he won scholarships from both the University of Padova and the European Community before moving to the United States, to work as a research fellow with Todd Horowitz and Jeremy Wolfe. His research interests include decision theory, game theory, nonlinear dynamical systems, and perceptual decisions.
74
Attention: Selective Attention and Consciousness
Todd Horowitz obtained his BS in psychology from Michigan State University in 1990. He then went on to study under Anne Treisman at the University of California, Berkeley, receiving his PhD in cognitive psychology in 1995, for his thesis titled ‘Spatial attention: Inhibition of distractor locations.’ In 1995, he came to Brigham and Women’s Hospital and Harvard Medical School for postdoctoral work on selective attention and circadian rhythms with Charles Czeisler and Jeremy Wolfe. In 2000, he joined the faculty of Brigham and Women’s and Harvard Medical School, where he is currently an assistant professor of ophthalmology. His work is currently funded by the National Institutes of Health. He has published over 30 papers on visual attention, visual search, multiple object tracking, and circadian rhythms in journals such as Nature, Psychological Science, Neuropsychologia, American Journal of Physiology, Psychonomic Bulletin and Review, and Journal of Experimental Psychology. He is currently a consulting editor for Perception & Psychophysics, and a member of the Association for Psychological Science, the Psychonomic Society, and the Vision Sciences Society. He lives in Medford, MA.
Jeremy Wolfe graduated summa cum laude from Princeton in 1977 with a degree in psychology and went on to obtain his PhD in 1981 from MIT, studying with Richard Held. His PhD thesis was entitled ‘On binocular single vision.’ Wolfe remained at MIT until 1991. During that period, he published papers on binocular rivalry, visual aftereffects, and accommodation. In the late 1980s, the focus of the lab shifted to visual attention. Since that time, he has published numerous articles on visual search and visual attention. In 1991, Wolfe moved to Brigham and Women’s Hospital and Harvard Medical School where he is a professor of ophthalmology. The lab is funded by the US National Institutes of Health, Air Force, and Department of Homeland Security. Wolfe teaches psychology courses at MIT and Harvard. Jeremy Wolfe is a past-president of the Eastern Psychological Association. He won the Baker Memorial Prize for teaching at MIT in 1989. He is a fellow of the AAAS, the American Psychological Association (Div. 3 and 6), the American Psychological Society, and a member of the Society for Experimental Psychologists. He lives in Newton, MA.
Attention: Selective Attention and Consciousness
75
Michael Cohen graduated from Tufts University in 2007, with a BA in philosophy and his minor in cognitive science. While at Tufts, he worked extensively with Daniel C Dennett on the problem of consciousness and its relation to visual attention. He previously worked in the Tufts NeuroCognition Lab under Philip Holcomb, where he used ERPs to investigate how the brain processes conceptual anaphors. He is presently a research technician in the visual attention lab at the Brigham and Women’s Hospital. His current research focuses on multiple object tracking, attentional capture, and visual short-term memory.
Autobiographical Memory and Consciousness M A Conway, University of Leeds, Leeds, England ã 2009 Elsevier Inc. All rights reserved.
Introduction The term ‘autobiographical memory’ refers to longterm memory for personal experiences and personal knowledge of an individual’s life. A specific autobiographical memory usually consists of at least one detailed memory of a personal experience (an episodic memory) and various associated items of knowledge. Consider a typical example of this mix of episodic memory and autobiographical knowledge, taken from a web survey of self-defining memories: I was about to go away for a year to Chile to work as an English teacher when I discovered that my dad had to have a major heart operation. It was a dream of mine to go, but that changed when my dad went into hospital. I went to visit him to tell him that I wouldn’t go. The hospital was really hot, and smelt of years of disinfectant. Dad was in a room by himself, looking very gaunt and worried. We talked about anything but my trip until I was due to go, but before I could say anything he got up and hugged me, saying that he would be more upset if I didn’t go. We both started to cry, I couldn’t speak. Dad said it would help him to know that I was fulfilling a dream. I couldn’t believe how much he loved me. He walked me to the lift and had to wrench himself away. I said ‘‘Thank you Dad, I love you.’’ and the doors closed.
This memory description has all the major features of a specific autobiographical memory: sensory– perceptual episodic details, affect, information about goals, personal interactions, and more conceptual autobiographical knowledge that contextualizethe episodic details in the rememberer’s life. Here we will explore how these different types of knowledge give rise to different types of conscious feelings of memory, how these may function in everyday life, and what occurs when they malfunction.
Cognitive Feelings: Remembering, Familiarity, and Knowing When a specific autobiographical memory comes to mind then a rememberer has ‘recollective
experience.’ That is, they experience remembering consciously and have what has been termed ‘autonoetic consciousness.’ Typically images enter conscious awareness, often visual in nature, attention turns inward, other highly specific knowledge may also feature too, and there is a strong sense of the self in the past. Additionally there is a distinct ‘feeling of remembering’; a feeling that what is in consciousness is a memory. Such feelings are part of a class of mental experience that have been termed ‘cognitive feelings.’ Cognitive feelings let us experience our mental states and without them we would have to, perhaps consciously and laboriously, infer what state we were in at any given time. Thus, the feeling of remembering, triggered by mental content such as visual images of past experiences, lets us know automatically and experientially that we are remembering – no further inferential reasoning is required to determine the state. The conscious feeling of remembering may be important too in convincing a person that they are indeed remembering and then to act on that. To give a trivial but not uncommon example, if a person remembers that they locked the door when they left the house, may be a visual image comes to mind and perhaps other details, and they have a feeling of remembering then they almost certainly would not go back to check. If, however, the feeling of remembering was not triggered by the knowledge brought into consciousness then determining whether what is in consciousness is in fact a memory and not, for instance, a generic visual image or set of images is more difficult and in some cases perhaps not possible at all. In which case, the probability of repeating one’s actions is increased. The feeling of remembering and autonoetic consciousness may then have quite powerful, direct, and important effects on behavior. The conscious experience of remembering can be contrasted with other states of memory awareness. For example, feelings of familiarity and knowing occur when autobiographical knowledge is brought to conscious awareness without associated
77
78
Autobiographical Memory and Consciousness
episodic memories. Remembering a school attended, the name of a friend, a work project, a holiday, a repeated event, etc., without remembering any single specific event is associated with familiarity and knowing but not with recollective experience. This type of remembering features what has been termed ‘noetic consciousness.’ Noetic consciousness does not feature specific representations of the self in the past such as those that are represented in episodic memories. Noetic consciousness cannot then trigger the feeling of remembering, although as will we shortly see this may occur in malfunctions of memory. Cognitive feelings characteristic of noetic states appear to be related partly to orienting processes and partly to metamemory functions. A person’s face for example might trigger a feeling of familiarity and lead to a metamemory inference that that person has been recently encountered. This may prime the memory system for detail recollection if current tasks come to require that, on the other hand, it may not lead to full recollection so reduce the attentional costs that would be incurred by constructing a full and detailed memory in consciousness. This ‘feeling of familiarity of recent occurrence’ may then serve to optimize cognitive performance and keep processing online and taskorientated while still supporting some recognition of past occurrence. In a similar way a more pervasive feeling of familiarity triggered by features of an individual’s habitual environment may facilitate orientation to that environment and fluent processing of it. In other words this ‘feeling of familiarity of the habitual’ may reduce attentional and processing demands on cognition and consciousness by signaling what does not need close attention and more detailed processing. It is particularly interesting that many brain-injured patients, who are disoriented in time and space, often claim that everything is unfamiliar, as though the feeling of familiarity triggered by the habitual had in some way been dysfacilitated. There may well be further shades of cognitive feelings here related to familiarity in the temporal dimension. For example, a conscious feeling triggered by the beginnings and endings of events, the feeling that events are proceeding fluently, feelings that one’s autobiographical memory is continuous, anticipatory feelings of
imagined future events (which share activation of many of the brain areas that are activated when remembering), are all memory-related cognitive feelings that await further investigation. The feeling of knowing may be more associated with metamemory functions than with orientation and again there seem to be two functions of this feeling: one is to let us feel what it is we know and other is to let us feel what we might know. For example, during the process of learning at universities, students initially rely on memories of knowledge presented in lecture, encountered in books, in conversation with teachers, or even in conversation with each other. Later, however, as they acquire knowledge they no longer depend on memories of times when knowledge was encountered and instead come to ‘just know.’ That is they just know, if they are psychology students, that the behaviorists rejected introspectionism, Freud discovered the unconscious, and in a with-subjects design the same participants take part in all the experimental conditions, and so on. Verification of these ‘facts’ is accompanied by a ‘feeling of knowing’ – a feeling that lets the individual know what they know experientially and without having to engage in further extended processing. People just know that Paris is the capital of France, that last year’s holiday was on a Greek island, that they once lived in London, and so on. Conversely there is knowledge in long-term memory that is available but not currently accessible to consciousness, usually because an appropriate cue cannot be located that would activate the knowledge to a level where it could enter consciousness and be experienced as known. Nonetheless extensive evidence shows that people can discriminate on the basis of ‘a feeling of knowing’ between items that they would be able to recognize and even remember with an effective cue from those they would not be able to recognize or remember even with a cue. Presumably this is because the knowledge in long-term memory is sufficiently activated to trigger a feeling of knowing, but not support incorporation of that item in a conscious representation. Were such an available item to become consciously accessible, then there is nothing to preclude it triggering the ‘just know’ feeling state. The relation between the feeling of knowing for temporarily inaccessible items and the feelings these same items trigger when they eventually do enter consciousness is not known.
Autobiographical Memory and Consciousness
Memory Feelings in Everyday Life One of the key features of conscious feeling states is that we act on them. If we remember the items we intended to buy at the supermarket we buy them, if we remember where we parked the car we go to that location, if we remember the emails we had to send, calls to make, letters to write, we do not repeat these. Similarly if we know we are meeting a colleague for lunch, if we know we completed a work project, we act on these conscious states of remembering. But what if, instead, we had not completed the project but only imagined it? Or perhaps a person might imagine going round the supermarket locating items on the shelves, in the fridges and freezers, in anticipation of actually making the shopping trip. How do imaging, remembering, and actually doing differ from one and other? One of the problems of consciousness is a reflexive one and that is to know what state (of consciousness) one is currently in, in order to act on it appropriately and adaptively. In the case of memory the feelings of remembering, familiarity, and knowing allow us to experience our conscious states ‘as’ particular states that can be acted on in ways that are adaptive for the state. However, these memory feelings can also be triggered by conscious representation of the future too. Indeed, imagining the future may often feature access of knowledge and memories in long-term memory and when this happens corresponding memory feelings will be triggered. Our model of the future, particularly the near future, is based on the past, especially the near past. This probably has adaptive value because much of the time (but certainly not always) the near future is closely related to the near past. Indeed, it has recently been found that the brain regions that become active during remembering and during imaging plausible false events are highly similar. Thus, it may be more correct to talk of a ‘remembering-imaging system’ than of a memory system alone. Within the remembering-imaging system, conscious states relating to the near past, present, and near future are colored by memory feelings that fluently arise in response to the contents of consciousness and allow us to feel continuity and familiarity in time – in short, what James so memorably called the ‘stream of consciousness.’
79
However, our memories for the immediate past, what happened earlier today, yesterday, and may be 1 or 2 days further in the past are highly detailed and often very specific and frequently trigger the feeling of remembering. It seems likely that our prospective memory for the future, our imagings of future events, and our goals and plans may also have a similar specificity for the near future, perhaps just a few days, and gradually become more generic and abstract as the future becomes more distant. Memory for the past and the future and conscious feelings that are triggered may operate in a remembered-imagined window of time. As this window moves, linearly, through time, the past and future components are constantly changing but we experience this as fluent and continuous because the patterns of memory feelings bind together moments of consciousness. When these patterns of conscious memory feelings are interrupted, conscious experience becomes discontinuous and fluency is lost. Under such circumstances making sense of the world is compromised and effortful.
Malfunctions of Memory Feelings One of the most severe cases of amnesia described in the literature is that of the musician Clive Wearing. He suffered brain damage due to a viral infection and this resulted in two types of amnesia. Retrograde amnesia refers to the loss of memories that were formed before a brain injury. Often this is temporally graded in the sense that the amnesia reaches back in time to a point where memories can be recovered again. For instance, in many closed head injuries (essentially a bang on the head) the amnesic period may only be for a few hours, days, or possibly weeks. In more severe cases it may cover years and even decades often stretching back to the early adulthood and teenage years. Wearing’s retrograde amnesia was extremely dense and cover all his prebrain damage life preserving only a few pieces of autobiographical knowledge, for example, he knew he was a musician, had a wife, had children, and a few more factual aspects of his life. This patient, very unfortunately, also suffered a second type of amnesia, anterograde amnesia (the more common type) and this was an inability, following his brain damage, to
80
Autobiographical Memory and Consciousness
encode new information into long-term memory. His short-term memory was, however, intact and so he had a window of consciousness that persisted for about a 2 min period, but which could not be updated. One of his disturbing behaviors was that he would write in a note book, many times a day, ‘I have just woken from being dead,’ filling many notebooks over the first period of his illness. Without the dynamic change remembering-imaging window of conscious driving memory feelings, he was wholly unable to operate on the world. Despite the personal tragedy of this case it is interesting to note that he was aware that consciousness was changing over time and must have had a feeling of change every few minutes. Clive Wearing is an unusual and extreme case. Other less severe types of amnesia may have other consequences. Patients have been described who have dense anterograde amnesia and cannot encode new experience at all or only to a limited extent but who have less dense temporally graded retrograde amnesia. One patient, for example, a man in his 60s, could recall reasonably well his early 20s, adolescence and childhood. In fact, in his late teens he had been on naval duty during World War II. It is clear from his descriptions of his memories that he had recollective experience and the feeling of remembering when he brought these to mind. The unfortunate consequence was that he acted on his memory feelings and came to behave like a 19-year old and have beliefs about the world that explained it as though he was still a naval rating, that is, that television constantly showed science fiction programs. One of the functions of conscious memory feelings such as recollection is to provide a basis for comprehension of the world and for adaptive actions. What one remembers of the past and the feelings which that triggers, drive our engagement with the world and in this case, as well as several other similar ones, demonstrate how powerful memory feelings can be in this respect. Understandings of the world that are erroneous but generated in good faith on the basis of mental representations are sometimes referred to as ‘confabulations’ or as one researcher insightfully put it ‘honest lies.’ What is most interesting about plausible confabulations of the past is that they have often
been found to be accompanied by recollective experience. Patients have in consciousness images of past experience and knowledge of their lives but configured in ways that are incorrect. One patient for example recounted his memory of the death of his brother in a car crash when he was much younger – in fact the brother was visiting him in hospital the same day. Such confabulations are not infrequent in certain types of brain-injured patients in whom false memories are experienced as true memories. It should be noted here too that such memories are not uncommon in the everyday life of healthy adults and, moreover, it has been found to be relatively simple to induce false recollections in laboratory settings. Inducing false memories of events dating to childhood; false memories that come to be recollectively experienced is well documented as is inducing false memory details in adults, details that again are recollectively experienced. One explanation of these experimentally induced false memories is that they arise from errors in ‘source monitoring.’ The individual, for whatever reason, fails to evaluate the source of the memory information in consciousness and attributes it to memory rather than some other source, for example, imagination. How this then gives rise to the conscious experience of remembering is not known. One possibility is that information in consciousness that is evaluated, by nonconscious processes, to be of memory content, perhaps for example it is found to be highly associated with other memory content, then triggers the experience of remembering. Whatever the case, it is clear that normal, intact, memory is also prone to at least the occasional honest lie. Consider the reverse situation and the conscious state that might occur if a person had recollective experience but no specific autobiographical memory entered conscious awareness. This powerful experience of remembering but with no memory in mind is the state that has been termed de´ja` vu. But de´ja` vu means ‘having seen before’ and as such approximates more to familiarity and knowing than to recollective experience. Researchers have suggested that the term de´ja` vu vecu, which is the experience of ‘having lived the present moment before,’
Autobiographical Memory and Consciousness
better captures the experience of recollection without a memory in conscious awareness. Some braindamaged patients suffer from the relatively frequent experience of de´ja` vu vecu and this may be because the brain circuits that mediate recollective experience are no longer constrained to activate only when a specific autobiographical memory is in conscious awareness. One proposal is that there is a brain circuit which when activated mediates the conscious feeling of remembering. The circuit is activated when specific memory content enters consciousness. However, specific memory content is most probably being automatically activated all the time by cues encountered in the environment and generated in the brain. One view is that control process gate this information entering consciousness and suppress the memory feeling circuit. When control processes are damaged such that they can no longer suppress the memory feeling circuit, but can still gate what knowledge enters consciousness then it is possible to have frequent experiences of recollection when no memory content is in consciousness. The consequences of this for patients who suffer this malfunction of consciousness are severe and far-reaching. For these patients the present moment is consciously felt as a memory. Because of this they act on their de´ja` vu vecu feelings because for them they are simply remembering. Thus, they stop reading newspapers (which only report old news), they cease watching television because it all repeats, they stop answering letters, sending letters and cards, fail to keep appointments and by their disturbing behavior induce anxiety, anger, and incomprehension in their carers (usually a life partner). A conscious feeling of remembering can then powerfully drive a wide range of behaviors. More generally, the experience of recollection without a specific memory in mind may occur in anyone if an activated memory triggers recollection but the memory itself does not enter conscious awareness. The fact that this is a relatively rare, but by no means unknown, malfunction of memory and consciousness suggests that the function of recollective experience, to allow us to know in an experiential way that we are remembering, is a particularly important function of this type of memory consciousness.
81
Conclusions The conscious experience of memory lets us feel what states we are in, allows us to identify distinct states, and supports the flow of experience through time in a window of remembering and imagining, and when it malfunctions disorientation, confusion, and confabulation arise. Furthermore, and most importantly, conscious feelings of memory drive behavior and when they are triggered inappropriately they can give rise to dysfunctional patterns of behavior. A further possible function of the conscious experience of remembering might also be to stimulate the adoption of plans and goals. Recollectively experiencing events which featured goal processing, knowing the outcomes of previous actions, maintaining the sense of familiarity, might drive the creation of new plans extend into the remembered-imagined future. It is particularly interesting that one deficit in amnesia appears to be the loss of the ability to imagine the future and the loss of goals extending into that future. Without a dynamic constantly changing and renewing window of remembered-imagined consciousness it seems that a future may not be possible. See also: Memory: Errors, Constructive Processes, and Conscious Retrieval; Self: Personal Identity; Self: The Unity of Self, Self-Consistency.
Suggested Readings Addis DR, Wong AT, and Schacter DL (2007) Remembering the past and imagining the future: Common and distinct neural substrates during event construction and elaboration. Neuropsychologia 45: 1363–1377. Conway MA (2005) Memory and the self. Journal of Memory and Language 53(4): 594–628. Gardiner JM (2008) Remembering and knowing. In: Roediger HL, III (ed.) Cognitive Psychology of Memory, Vol. 2 of Learning and Memory: A Comprehensive Reference, Byrne J (ed.), 4 vols, pp. 169–198. Oxford: Elsevier. James W (1950) Principles of Psychology. New York: Dover. Johnson ML, Hashtroudi S, and Lindsay DS (1993) Source monitoring. Psychological Bulletin 114: 3–28. Moulin JAC, Conway MA, Thompson R, James N, and Jones RW (2004) Disordered memory awareness: Recollective confabulation in two cases of persistent de´ja` vecu. Neuropsychologia 43: 1362–1378.
82
Autobiographical Memory and Consciousness
Roediger HL and McDermott KB (1995) Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition 21: 803–814. Sacks O (1985) The Man Who Mistook His Wife for a Hat. London: Duckworth. Schacter DL and Addis DR (2007) The cognitive neuroscience of constructive memory: Remembering the past and imagining the future. Philosophical Transactions of the Royal Society of London Series B 362: 773–786. Tulving E (1983) Elements of Episodic Memory. New York: Oxford University Press. Tulving E (1985) Memory and consciousness. Canadian Psychology 26: 1–12.
Tulving E (2002a) Episodic memory: From mind to brain. Annual Review of Psychology 53: 1–25. Tulving E (2002b) Chronesthesia: Awareness of subjective time. In: Stuss DT and Knight RC (eds.) Principles of Frontal Lobe Functions, pp. 311–325. New York: Oxford University Press. Tulving E and Pearlstone Z (1966) Availability versus accessibility of information in memory for words. Journal of Verbal Learning and Verbal Behaviour 5: 381–391. Wheeler MA, Stuss DT, and Tulving E (1997) Towards a theory of episodic memory: The frontal lobes and autonoetic consciousness. Psychological Bulletin 121: 351–354.
Automaticity and Consciousness W Schneider, University of Pittsburgh, Pittsburgh, PA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Automatic attention response – A specific type of automatic process that does priority assessment of stimuli based on the priority for future processing and places information in working memory with a priority signal can interrupt ongoing controlled processing and facilitates the processing of the high-priority stimulus. Automatic process – Cognitive process that develops for consistently executed component behaviors that are performed rapidly, with minimal effort or attention involving a long-term associative connections and require an appreciable amount of consistent training to develop fully. The processing is domain specific, occurring in multiple perceptual/response area of the brain. Cognitive control network – A network of brain areas that provide endogenous control of attentional operations performing controlled processing operations including domain general cortical areas. The brain areas include anterior cingulate cortex/ presupplementary motor area (ACC/pSMA), dorsolateral prefrontal cortex (DLPFC), inferior frontal junction (IFJ), anterior insular cortex (AIC), dorsal premotor cortex (dPMC), and posterior parietal cortex (PPC). Controlled process – Cognitive process that involves attention and top-down control that performs novel or varied tasks that is characterized as being slow, serial, effortful, easy to set up and alter, laying down a strong memory trace, and nonrobust (e.g., sensitive to stress, alcohol, sleep deprivation). It occurs in the domain general cognitive control network of the brain. Spatial neglect – A neurological condition after damage to the parietal cortex that
produces a deficit in attention to and awareness of one side of space. Visual search – A task in which an observer is given a memory set of items to search for, for example, letters, words, and pictures and search of display of potentially multiple items, and responding if there is a match (e.g., push a button if you see the letters ‘c’ or ‘b’ in the string of ‘detbn’).
Introduction The concepts of automaticity and consciousness have been discussed together extensively for centuries and this continues in the current literature. The concept of automaticity was a major focus in Williams James’s Principles of Psycholog y (1890), which contrasted habit and ideational/will processing. In modern times, automatic processing has been an important issue in attention, skill acquisition, social perception, and cognitive neuroscience. Most human behaviors involve automatic processing of consistent components of the task. For example, the word encoding of this text is an automatic process. When you look at a familiar word, the meaning of the word is activated. This takes little effort or awareness of the component processes of perception. In contrast, if we alter the words such that the automatic processes fail, reading becomes slow and effortful and you become aware of the processing. Interpreting the text string where the letter order is reversed (e.g., what is ‘‘ssensuoicsnoc’’) is hard and error prone, and disrupts the normal reading process severely. Note, however, that you were likely more aware of the rearranged version of ‘consciousness’ read in reverse order than you were the other times that word occurs on this page. The type of processing done on the reversed text is typically referred to as controlled processing, 83
84
Automaticity and Consciousness
which is characteristic of novel or early trials of a task. The distinction between automatic and controlled processing relates to dual process theories of human cognition. These two processing modes differ in the types of processing they perform and the brain mechanisms involved. Multiple theoretical positions align controlled and automatic processing with the concepts of conscious and nonconscious processing. Although there is a high overlap of concepts where most control processes are conscious and automatic processes are not, there are well-known exceptions that indicate consciousness has a more nuanced relationship. Consciousness is related to a subset of control process events and some automatic processes can place information into consciousness.
Automaticity and Dual Processing Theory The concept of automaticity relates to dual process theories that assume that most human behavior results from the interplay of two types of processing referred to as automatic and controlled processing. The theoretical and empirical work of Walter Schneider and Richard Shiffrin provided modern day definitions and quantitative models of these cognitive processing modes. ‘Automaticity’ or ‘automatic processing,’ refers to the cognitive process that develops for consistently executed component behaviors that are performed rapidly with minimal effort or attention, involve long-term associative connections, and require an appreciable amount of consistent training to develop fully. Automatic processing is typically characterized as being fast, parallel, requiring low effort, minimal memory demand, and is a robust type of processing. Practiced consistent behaviors involve automatic processing in tasks such as driving, brushing teeth, word encoding, and social perception. In contrast to automatic processes, ‘controlled processing’ occurs typically in novel or varied situations (e.g., searching for a random subset of letters in a display). Controlled processing is characterized as being slow, serial, effortful, easy to set up and alter, laying down a strong memory trace, and nonrobust (e.g., sensitive to stress, alcohol, sleep deprivation). A novel task is initially executed
by controlled processing. Controlled processing involves attentive operations that can shift attention, compare memory states, execute sequential rules, induce learning, and initiate responses. Empirical results show automatic and controlled processing differ in six characteristics: (1) extended consistent training is required for automatic processing while controlled processing requires only a few trials to develop; (2) automatic processing is fast and parallel, while control processing is slow and serial; (3) automatic processing requires low effort whereas controlled processing is high effort; (4) automatic processing is robust to stressors while controlled processing is not, (5) automatic processing is hard to alter whereas controlled processing is easy; and (6) learning is dependent on the amount and nature of controlled processing where as little is learned in pure automatic processing. These differences between controlled and automatic processing are quite marked and often interpreted as indicating qualitatively different forms of processing. For example, with training humans can perform surprisingly fast parallel automatic tasks (e.g., looking for words, icons, numbers colors at the same time and processing 96 stimuli per second at high accuracy). In an example of semantic category search (e.g., find words that are members of the categories, press a button if the word is from the categories ‘vehicles, land formations, units of time, and fruits’ then be presented words such as ‘hammer, foot, bank, peach’). If this task is novel or the stimuli have a varied mapping (i.e., which stimuli are targets or distracters varies over trials) controlled processing is active. Processing requires about 200 ms for each comparison, so with 16 comparisons the reaction time would be 3.2 s. The mental effort would be slow and serial as each category would be checked against each word. In contrast, if you searched for the same four categories consistently for many trials, automatic processing would develop. Automatic processing would be fast (1 ms/comparison, 100) and parallel in memory for small displays (two words). An important trait of automatic search is the search appears to be a parallel search for example, searching for a match of many possible previously detected stimuli can be as fast as searching for any specific stimulus.
Automaticity and Consciousness
The terminology of automatic/controlled processing emphasizes the difference in control of the processing. Controlled processes are easy to set up and change (e.g., respond to red stimuli on one trial, and on the next, respond to green stimuli). However, if you perform a given operation many times, it becomes easier to perform and may become automatic where it is difficult to inhibit the processing of the task. For example, consider the Stroop color naming task, where participants name the ink color of words which are incompatible with the color meaning of the text (e.g., the text ‘blue’ printed in red ink and the response is to answer ‘red’). This slows reaction time dramatically (i.e., 70% slower response time), relative to naming the colors of squares without the incompatible words. This effect is a result of developing automatic word processing, which evokes the color response even when it is not intended. Altering this automatic process requires effort and takes a week of training to overcome. Action slips also show and lack of control in automatic behaviors. Action slips are common errors such as getting out of an elevator on the wrong floor because the doors opened up a floor early or continuing on your standard driving route to work when you intended to go somewhere else illustrate that an automatic process can determine behavior with minimal consciousness. Addictions such as cigarette smoking and pathological gambling illustrate the difficulty of altering well practiced consistent behaviors even if they are very detrimental. These two forms of processing provide complementary benefits. Tasks can be learned quickly via controlled processing but are slow, serial, and effortful, and nonrobust. After much consistent practice in the task, it can be performed much faster with lower effort by automatic processing. In martial arts contests, the master’s speed is far beyond that of the novice and in a fight the master would have great advantage relative to the novice. However, the novice’s ability to hear instructions from the master allows the student to execute rudimentary forms of the moves in few trials. The student is consciously aware of the specific steps of the moves and often rehearses them before execution. With instruction, students learn good moves, much more rapidly than without instruction, and
85
conscious control of their actions. The presence of control processing provides early skill and produces the memory representations that with practice allow fast execution of effective moves. The two forms of processing dynamically interact resulting in cognitive processing that is a result of exogenous and endogenous processing. Stimuli with certain features (e.g., loud noises, moving/ looming visual objects, pain) and learned responses (your name, cry of your child), produce an ‘automatic attention response’ that activates the related representation and focuses attention on the stimulus. Note this can occur without any intention (e.g., missing a stair). The classic ‘cocktail party effect’ is where you hear your name in a presumably unattended conversation, which causes you to shift your attention to that conversation, and you become consciously aware people are talking about you. Control processing can also alter automatic processing. In a fraction of a second you can alter how you walk. This is remarkable in that you can alter a skill you have perhaps practiced for decades and dramatically alter your motor movements (e.g., try to walk but do not let the outsides of your feet touch the floor). Your walking will likely be much slower, demand much more attention, and you will be far more consciously aware of the act of walking.
Consciousness in Dual Processing Theory The concepts of controlled and automatic processing are closely related to concepts to consciousness and nonconsciousness processing. This is a simplified view that captures the high overlap of the situations in which both occur but this view does not detail the different nuanced conditions in which this simplification breaks down. In order to identify the differences one must be clear about the definitions of consciousness and automaticity that one is considering. There are also multiple definitions of consciousness. The distinction between access consciousness and phenomenal consciousness by Ned Block provides a useful distinction to organize the discussion of two types of consciousness. Briefly, ‘access consciousness’ relates information made available to a
86
Automaticity and Consciousness
global workspace, what Block refers to as ‘‘the brain’s ‘consumer’ systems: systems of memory, perceptual categorization, reasoning, planning, evaluation of alternatives, decision-making, voluntary direction of attention, and more generally, rational control of action.’’ Phenomenal consciousness refers to the experience of the event and the ability to introspect, reflect and report on the nature of the event. This metacognitive report may be a different process that provides an experience based on processing events occurring in the global workspace. Table 1 provides a matrix of dual processing concepts and the relationship to consciousness. The upper left cell, automatic nonconscious behaviors, represents the bulk of human cognition. You are unaware (either phenomenological or via access) of the enormous knowledge stored in the connection weights between brain regions. You can become aware of the result of using those connections but not of the connections themselves. For example: What is the name of the first person to walk on the moon? How do you tie your shoes? You likely had neither access nor phenomenological awareness to those two facts before you read the questions. You can drive and make many corrections for changes in the road or other drivers that influence your actions. Yet you may have very limited awareness of much of what is perceptually processed. Dan Simons’ study of ‘change blindness’ illustrates that you are aware generally what is attended to and the gist of the environment. Simons Table 1
examines what is detected in movies between camera cuts. In a conversation of two people much can be changed (e.g., object on a table clothing of the actors, the actors themselves) and occur without awareness. We are aware of the gist of the situation and what we are attentively focused on (e.g., changing an actor from a male to female in a conversation of two people is detected as it changes the expected nature of interaction). The lack of consciousness in nearly full automatic processing is perhaps best illustrated in very high workload, high stress, zombie behaviors, and epileptic seizures and alcohol intoxication. High workload (e.g., driving while engaged in a heated argument) can result in driving with little conscious awareness of the drive (e.g., what stop lights did you stop at). In high stress situations such as combat, there can be a nearly complete disassociation between extensive automatic behaviors (e.g., hand combat motor actions) and the conscious mind activity (e.g., will my mother forgive me if I die in the next minute?). Zombie behaviors are highly trained, automatic, fluid visuomotor behaviors in which people are on automatic pilot (mountain biking, playing sports, driving). Sleepwalking can involve complex behaviors (e.g., rearranging furniture) with no sign of awareness or memory. Automatic capture errors (e.g., exiting an elevator on the wrong floor) illustrate autopilot performance. Robert Penfield describes an epileptic walker: ‘‘Thus, the automaton can walk through traffic as though he were aware of all that he hears and sees, and so continue on his way home. But he is
Relations of consciousness and dual processing modes
Automatic processing
Controlled processing
Nonconscious
Access conscious
Phenomenal consciousness
Connection strength
Pattern activations in working memory
Automatic attention responses Gist
Pattern activation in working memory
Slow processes involving changes in controlled processing events
Most pattern activation Zombie behaviors Sub 100 ms component processes Some extended execution of controlled processing pre results
Selection of information for comparison or remembering
Completed component processes Rehearsed working memory items
Automaticity and Consciousness
aware of nothing and so makes no memory record. If a policeman were to accost him he might consider the poor fellow to be walking in his sleep.’’ These all illustrate situations where past consistent practice has created associations that can execute complex behaviors with little controlled processing or awareness. The behaviors are complex and can be executed for extended durations without the need for controlled processing. Automatic behaviors can place information into working memory. Most of it is not phenomenologically available but can be if attention is shifted to it (e.g., the pressure of the chair you are sitting on as it presses against the leg). One class of automatic processing, an automatic attention response, does place information into both access and phenomenal consciousness. This is best illustrated in visual search tasks. For example, if a subject is consistently detecting a specific set of letters, for example, (‘X’ or ‘S’), after hours of practice, one can detect letters at very high rates (e.g., process displays of 4 letters occurring every 80 ms, or 50 events per second with high accuracy >90%). At very high rates nontrained observers are unable to see character shapes at all. Yet, to the trained observer the letters pop out of the blur perceptually seeming to stop the display and become consciously available. Hearing your name in an ‘unattended conversation,’ missing a step and attending to your feet, or trying to retrieve a reference in memory and it ‘pops’ to mind, are examples of an automatic process activating a stimulus, directing attention to the stimulus, and generating awareness. The phenomenon of spatial neglect illustrates how losing the automatic attention responses can produce a loss of access to information. Spatial neglect is a condition after damage to parietal cortex that produces a deficit in attention to and awareness of one side of space. Patients with neglect do not see stimuli on one half of the visual field or perform actions on half of their body (e.g., not dressing the right side of their body). This is not a deficit in vision because they can see the stimuli (e.g., can read a single word crossing both visual fields but when asked to name random letters will neglect to read the ones on the right side). This lack of stimuli interrupting controlled processing is very debilitating for operating in the normal world making it difficult to dress oneself, perform
87
skills (driving) and interact socially. Controlled processing is tightly associated with conscious processing but there are some control processes that are not conscious. Controlled processing is slow, serial, and involves attention. Given that consciousness is assumed to involve longer processing of hundreds of milliseconds, it is reasonable to expect that the results of controlled processing are both consciously accessible and phenomenologically available. In general, individuals are aware of what they have attended to, targets they have detected, and working memory items that they have rehearsed. In controlled search tasks (ones with novel or varied targets), whenever a target is detected, the subject can report the target. False alarms are rare indicating the detection resulted in phenomenological awareness. There are control processes that seem unavailable to consciousness. A dramatic example of this occurs in a search task when the subject finishes the search, makes a response indicating no target was detected, feels they have quit the task, but then sometimes after a substantial delay (several hundred milliseconds) they have an ‘oops’ response where they realize they missed the target. This suggests that the control process was continuing to check information but there was no conscious phenomenal awareness. Subjects have difficulty reporting details of fast control processes. For example, in search task comparing small numbers (4 or less comparisons, 200 ms comparison time) subjects typically execute an exhaustive search continuing to compare items even after a match is found. When this is discussed in classes, students often disbelieve that they could be using such a strategy. As another example, subjects may perform very predictable search patterns (e.g., in driving around a curve, subjects look at the tangent point of the curve), but few are aware of the strategy they deploy. These are examples of control processes that result in minimal awareness.
Function of Consciousness in Dual Processing Theory Automatic, controlled, and conscious processing operate together with different functions
88
Automaticity and Consciousness
performed by each. The functions of automatic processing are to perform routine, consistent cognitive actions and to alert controlled processing/ consciousness of important stimuli and the gist of the current situation. The functions of control processing are to control attention, perform control process operations such as selective attention, comparison, response release, performing interpretive execution of control programs, set up patterns to allow learning, and to alter/modify the result of automatic processing. Each processing mode has strengths and weaknesses but together they provide greater functionality and survivability to the individual than each separately. There are many disparate views of the functions of consciousness. These views vary in terms of the types of consciousness involved and the breadth to which they feel the term of consciousness applies. Lee Pierson and Monroe Trout view the role of consciousness to explicitly counteract serious problems with purely automatic processing. They state ‘‘Consciousness makes volitional attention possible’’ and that ‘‘Although the ultimate function of volitional attention is to make volitional movement possible, its proximate function is to override automatic attention that is not well-suited to the situation at hand. The capacity for volitional attention gives the conscious organism the flexibility to nondeterministically yet nonrandomly sustain attention on a particular conscious content longer than the default set by neural processes.’’ They continue commenting that consciousness can correct ‘‘misapplied automatisms, infinite loops, and tendencies that were adaptive when they evolved but have outlived their usefulness.’’ In a changing world there is a need to alter some automatic behaviors. A common example might be when a software ‘upgrade’ changes a function of key strokes. The motor actions that were usefully automatic become painfully difficult to inhibit and correct. In search tasks, it can take more time to alter automatic behaviors than it did to learn them. In the case of addictive habits (e.g., smoking) this can make people have enormous difficulty kicking the habit even if they know it is killing them. In a visual search task if you learned to search for numbers and ignore letters, detection of numbers and the search for them becomes automatic and quick, requiring low effort. However, if
you switch to detect letters and ignore numbers, the numbers pop into consciousness and it is difficult to process the letters, making you worse that a novice with no practice at all. Subjects report that after reversal, the task that was easy and perhaps boring is now aversive. They feel their attention is dragged to the old targets when they are trying to attend to the new ones. They are very phenomenally aware that automatic processing is not working to their advantage. The conscious effort to abort processing the old targets and maintain attention on the new letter targets occurs slowly, and after thousands of trials new automatic processing is established. Consciousness awareness of the processing and controlled effort to inhibit the old automatic processing allows for alteration of automatic processes and the eventual replacement of automatic processes with new ones. Many consciousness researchers take a wider view on the role of consciousness. Bernard Baars lists a variety of functions including executive control, decision making, recruiting and control actions, adaptation and learning; prioritizing the cognitive system’s concerns, facilitating problem solving, optimizing the trade-off between organization and flexibility, detecting errors and editing action plans, creating access to the self, facilitating learning and adaptation, and in general increasing access between otherwise separate sources of information. In this broader view of consciousness there is a great deal of overlap between the concepts of control processing and the nature of access consciousness. It is unclear which of these differences are just terminology referring to the identical underlying functions. The access consciousness working memory and controlled processing working memory are perhaps very similar constructs. The functions of phenomenological consciousness provide tasks that suggest more reflective processing systems than is typically associated with controlled processing. This is perhaps a processor that has access to a subset of control operations and working memory (see above controlled processing without phenomenological awareness) but can reflect on those operations and their consequences. Consciousness could have multiple functions operating as a metaprocess to controlled processing. It might not directly perform control
Automaticity and Consciousness
operations but rather monitor the success of those operations and initiate tweaks in control processing. For example, in controlled searches, if observers are missing too many targets, they often slow down the comparison rate. If there is a demand for more speed they might increase the rate. If there are multiple concurrent tasks they may shift more resources to the task with higher payoff. In these situations observers are phenomenally aware of their performance and take volitional actions to alter processing. The metaprocess would be beneficial for accomplishing the broad list of Baars’ expanded functions of consciousness. These functions generally relate to monitoring control process operations, assessing whether current operations are accomplishing the desired goals at the intended levels of performance, determining if the existing control operations are successful and using control state variables to influence controlled and automatic behaviors. To illustrate the metaprocess, consider control processing related to reading instructions to assemble a bookshelf from the instructions. This is not automatic, each step of the instructions must be read, converted to a control process routine, and executed. If the instructions are clear, easy to interpret, accurate and read, there is little need for controlled metaprocessing. The control instructions typically involve search for this part A, move position X on part B, and perform motor action Y on part B until some state Z is reached. Then go on to the next step. However, if a control subroutine fails (e.g., you can not find a part A) or the motor action fails (e.g., the screw does not go in), a metaprocess kicks in to find alternatives to execute the intended step. This could take many forms such as tweaks (look longer for A), rereading/re-encoding the step, deciding if the instructions are wrong and interpreting what they ‘meant to write,’ recalling the last time you performed a similar task, deciding whether to delegate the task to someone else, or deciding to give up. These metacontrol operations have great utility in a world where control processes must be learned or altered across situations. They also have the characteristic of not being part of the control routine that would have been executed if the step had succeeded.
89
Information Access and Transfer between Automatic, Controlled, and Conscious Processing Given the large amount of parallel automatic processing and the small amount of serial controlled and conscious processing it is important to recognize the limited information flow between the three systems. The relationships in Table 1 indicate consciousness has very limited access to the full processing of the brain represented by automatic processing. Schneider and Chein propose that automatic processing occurs in a network of ten thousand hypercolumn populations and at any point controlled processing can monitor a only small fraction of these ( 0. This can be achieved, for instance, by ensuring that each unit is connected to other units in a different way, and yet all units can interact. In short, integrated information captures the information generated by causal interactions in the whole, over and above the information generated independently by the parts. Complexes Finally, by measuring F values for all subsets of elements within a system, we can determine which
407
subsets form complexes. Specifically, a complex X is a set of elements that generate integrated information (F > 0) that is not contained in some larger set of higher F. A complex, then, can be properly considered to form a single entity having its own, intrinsic ‘point of view’ (as opposed to being treated as a single entity from an outside, extrinsic point of view). Since integrated information is generated within a complex and not outside its boundaries, experience is necessarily private and related to a single point of view or perspective. A given physical system, such as a brain, is likely to contain more than one complex, many small ones with low F values, and perhaps a few large ones. In fact, at any given time there may be a single main complex of comparatively much higher F that underlies the dominant experience (a main complex is such that its subsets have strictly lower F). In general, a main complex can be embedded into larger complexes of lower F: a complex can be casually connected, through ports-in and ports-out, to elements that are not part of it. According to the IIT, such elements can influence indirectly the state of the main complex without contributing directly to the conscious experience it generates.
Neurobiology: Accounting for Empirical Observations Can this approach account, at least in principle, for some of the basic facts about consciousness that have emerged from decades of clinical and neurobiological observations? Measuring F and finding complexes is not easy for realistic systems, but it can be done for simple networks that bear some structural resemblance to different parts of the brain. For example, by using computer simulations, it is possible to show that high F requires networks that conjoin functional specialization (due to its specialized connectivity, each element has a unique functional role within the network) with functional integration (there are many pathways for interactions among the elements). In very rough terms, this kind of architecture is characteristic of the mammalian corticothalamic system: different parts of the cerebral cortex are specialized for different functions, yet a vast network of connections allows these parts to interact profusely.
408
An Integrated Information Theory of Consciousness
And indeed, as much neurological evidence indicates, the corticothalamic system is precisely the part of the brain which cannot be severely impaired without loss of consciousness. Conversely, F is low for systems that are made up of small, quasi-independent modules. This may be why the cerebellum, despite its large number of neurons, does not contribute much to consciousness: its synaptic organization is such that individual patches of cerebellar cortex tend to be activated independently of one another, with little interaction between distant patches. Computer simulations also show that units along multiple, segregated incoming or outgoing pathways are not incorporated within the repertoire of the main complex. This may be why neural activity in afferent pathways (perhaps as far as V1), though crucial for triggering this or that conscious experience, does not contribute directly to conscious experience; nor does activity in efferent pathways (perhaps starting with primary motor cortex), though it is crucial for reporting each different experience. The addition of many parallel cycles also generally does not change the composition of the main complex, although F values can be altered. Instead, cortical and subcortical cycles or loops implement specialized subroutines that are capable of influencing the states of the main corticothalamic complex without joining it. Such informationally insulated cortico-subcortical loops could constitute the neural substrates for many unconscious processes that can affect and be affected by conscious experience, such as those that enable object recognition, language parsing, or translating our vague intentions into the right words. Other simulations show that the effects of cortical disconnections are readily captured in terms of integrated information: a ‘callosal’ cut produces, out of large complex corresponding to the connected corticothalamic system, two separate complexes, in line with many studies of split brain patients. However, because there is great redundancy between the two hemispheres, their F value is not greatly reduced compared to when they formed a single complex. Functional disconnections may also lead to a restriction of the neural substrate of consciousness, as is seen in neurological neglect phenomena, in psychiatric conversion
and dissociative disorders, and possibly during dreaming and hypnosis. It is also likely that certain attentional phenomena may correspond to changes in the composition of the main complex underlying consciousness. The attentional blink, where a fixed sensory input may at times make it to consciousness and at times not, may also be due to changes in functional connectivity: access to the main corticothalamic complex may be enabled or not based on dynamics intrinsic to the complex. Computer simulations confirm that functional disconnection can reduce the size of a complex and reduce its capacity to integrate information. While it is not easy to determine, at present, whether a particular group of neurons is excluded from the main complex because of hard-wired anatomical constraints, or, is transiently disconnected due to functional changes, the set of elements underlying consciousness is not static, but form a ‘dynamic complex’ or ‘dynamic core.’ The most common example of a marked change in the level of experience is the fading of consciousness that occurs during certain periods of sleep. Subjects awakened in deep nonrapid eye movement (NREM) sleep, especially early in the night, often report that they were not aware of themselves or of anything else, though cortical and thalamic neurons remain active. Awakened at other times, mainly during REM sleep or during lighter periods of NREM sleep later in the night, they report dreams characterized by vivid images. From the perspective of integrated information, a reduction of consciousness during early sleep would be consistent with the extreme bistability of cortical circuits that occurs during deep NREM sleep: due to changes in intrinsic and synaptic conductances triggered by neuromodulatory changes (e.g., low acetylcholine), cortical neurons cannot sustain firing for more than a few hundred milliseconds, and invariably enter a hyperpolarized down-state. Shortly afterwards, they inevitably return to a depolarized up-state. Indeed, computer simulations show that values of F are low in systems with such a bistable dynamics. Consistent with these observations, studies using TMS in conjunction with high-density EEG show that early NREM sleep is associated either with a breakdown of the effective connectivity among cortical areas, and thereby with a loss of integration, or with a
An Integrated Information Theory of Consciousness
stereotypical global response suggestive of a loss of repertoire and thus of information. Similar changes are seen in animal studies of anesthesia. Computer simulations also indicate that the capacity to integrate information is reduced if neural activity is extremely high and near-synchronous, due to a dramatic decrease in the repertoire of discriminable states. This reduction in degrees of freedom could be the reason why consciousness is reduced or eliminated in absence seizure and other conditions during which neural activity is both high and synchronous. Finally, consciousness not only requires a neural substrate with appropriate anatomical structure and appropriate physiological parameters: it also needs time. The theory predicts that the time requirement for the generation of conscious experience in the brain emerge directly from the time requirements for the build-up of an integrated repertoire among the elements of the corticothalamic main complex so that discriminations can be highly informative. To give an obvious example, if one were to perturb half of the elements of the main complex for less than a millisecond, no perturbations would produce any effect on the other half within this time window, and F would be zero. After say 100 ms, however, there is enough time for differential effects to be manifested, and F should grow.
Specifying the Quality of Consciousness If the amount of integrated information generated by different brain structures or by the same structure functioning in different ways can in principle account for changes in the level of consciousness, what is responsible for the quality of each particular experience? What determines that colors look the way they do, and different from the way music sounds? Once again, empirical evidence indicates that different qualities of consciousness must be contributed by different cortical areas. Thus, damage to certain parts of the cerebral cortex forever eliminates our ability to experience color (whether perceived, imagined, remembered or dreamt), whereas damage to other parts selectively eliminates our ability to experience visual shapes.
409
There is obviously something about different parts of the cortex that can account for their different contribution to the quality of experience. What is this something? The IIT claims that just like the quantity of consciousness generated by a complex of elements is determined by the amount of integrated information it generates above and beyond its parts, the quality of consciousness is determined by the set of all the informational relationships its mechanisms generate. That is, how integrated information is generated within a complex determines not only the amount of consciousness it has, but also what kind of consciousness. Consider again the photodiode thought experiment. As we discussed before, when the photodiode reacts to light, it can only tell that things are one way rather than another way. On the other hand, when we see ‘light,’ we discriminate against many more states of affairs, and thus generate much more information. In fact, we argued that ‘light’ means what it means and becomes conscious ‘light’ by virtue of being not just the opposite of dark, but also different from any color, any shape, any combination of colors and shapes, any frame of every possible movie, any sound, smell, thought, and so on. What needs to be emphasized at this point is that discriminating ‘light’ against all these alternatives implies not just picking one thing out of ‘everything else’ (an undifferentiated bunch), but distinguishing at once, in a specific way, between each and every alternative. Consider a very simple example: a binary counter capable of discriminating among the four numbers: [00 01 10 11]. When the counter says binary ‘3,’ it is not just discriminating [11] from everything else as an undifferentiated bunch, otherwise it would not be a counter, but a [11] detector. To be a counter, the system must be able to tell [11] apart from [00] as well as from [10] as well as from [01] in different, specific ways. It does so, of course, by making choices through its mechanisms, for example: is this the first or the second digit? Is it a 0 or a 1? Each mechanism adds its specific contribution to the discrimination they perform together. Similarly, when we see light, mechanisms in our brain are not just specifying ‘light’ with respect to a bunch of undifferentiated alternatives. Rather, these mechanisms are specifying that light is what it is by
410
An Integrated Information Theory of Consciousness
virtue of being different, in this and that specific way, from every other alternative, from dark to any color to any shape, movie frame, sound or smell, and so on. In short, generating a large amount of integrated information entails having a highly structured set of mechanisms that allow us to make many nested choices as a single entity. According to the IIT, these mechanisms working together generate integrated information by specifying a set of informational relationships that univocally determine the quality of experience. The Shape of Experience To see how this intuition can be given a mathematical formulation, let us consider again a complex of n binary elements that enters a particular state and makes a discrimination that generates a certain amount of integrated information F. Let us now suppose that each possible state of the system constitutes an axis or dimension of a qualia space (Q-space) having 2n dimensions. Each axis is labeled with the probability p for that state, going from 0 to 1, so that a repertoire, that is, a probability distribution on the possible states of the complex, corresponds to a point in Q-space. Let us now examine how the connections among the elements of the complex specify probability distributions, that is, how a set of causal interactions specifies a set of informational relationships. First, consider the complex with all connections among its elements disengaged, so that no causal interactions can occur. In this case, the system could not discriminate among any states, which would all be equally likely, corresponding to the maximum entropy or uniform distribution (the potential repertoire). In Q-space, this probability distribution is a point projecting onto all axes at p ¼ 1/2n. Next, consider engaging a single connection: by itself, each connection specifies an actual repertoire – it generates some information by shaping the uniform distribution into a more specific distribution – for instance, it may create some peaks (increase the probability of some states) and valleys (reduce the probability of other states, perhaps down to zero if it rules them out). In Q-space, this corresponds to a point projecting onto higher p values on some axes and onto lower p-values
(or zero) on other axes. That is, the mechanism implemented by this single connection specifies an informational relationship (a relationship between two probability distributions). This informational relationship can be represented as an arrow in Q-space (q-arrow) that goes from the point corresponding to the maximum entropy distribution to the point corresponding to the actual repertoire specified by that connection. The length (divergence) of the q-arrow expresses how much the connection specifies the distribution (the effective information it generates, i.e., the relative entropy between the two distributions); the direction in Q-space expresses the particular way in which the connection specifies the distribution. Similarly, if one considers all other connections taken in isolation, each will specify another q-arrow of a certain length, pointing in a different direction. Next, consider all possible combinations of connections. For instance, consider adding the contribution of the second connection to that of the first. Together, the first and second connections specify another actual repertoire – another point in Q-space – and thereby generate more information than either connection alone as they shape the uniform distribution into a more specific distribution. To the tip of the q-arrow specified by the first connection, one can now add a q-arrow bent in the direction contributed by the second connection, forming a path of two q-arrows in Q-space (the same final point is reached by adding the q-arrow due to the first connection on top of the q-arrow specified by the second one). Each combination of connection therefore specifies a q-path made of concatenated q-arrows (component q-arrows). In general, the more connections one considers together, the more the actual repertoire will take shape and differ from the uniform (potential) distribution. Finally, consider the joint contribution of all connections of the complex. As was discussed above, all connections together specify the actual repertoire of the whole. This is the point where all q-paths made of q-arrows converge (one for every ordered permutation of all connections). Together, these converging paths in Q-space delimit a quale, that is, a shape in Q-space (a kind of ‘solid,’ or rather the ‘body’ of a polytope in more than three dimensions) whose edges are q-arrows, whose bottom is
An Integrated Information Theory of Consciousness
the maximum entropy distribution, and whose top is the actual repertoire of the complex as a whole. The shape of this solid specifies all informational relationships that are generated within the complex by the interactions among its elements – also known as the effective information matrix. It is worth considering briefly two relevant properties of informational relationships or q-arrows. First, informational relationships are contextdependent, in the following sense. A context can be any point in Q-space corresponding to the actual repertoire generated by a particular subset of connections. It can be shown that the q-arrow generated by considering the additional effects of a connection (how it further changes the actual repertoire) can change in both magnitude and direction depending on the context in which it is considered. Another important property of q-arrows is entanglement g. A q-arrow is entangled (g > 0) if the underlying connections generate more information jointly than considered independently, that is, if it specifies information above and beyond its component q-arrows. An entangled q-arrow specifies a concept in that it groups together certain states of affairs in a way that cannot be decomposed into the mere sum of simpler groupings. Thus, entanglement characterizes informational relationships (q-arrows) that are more than the sum of their component relationships (component q-arrows), just like F characterizes systems that are more than the sum of their parts. Geometrically, entanglement ‘warps’ the shape of the quale away from a simple hypercube (where q-arrows are orthogonal to each other). Entanglement is important in identifying modes, just as F is useful in identifying complexes. By analogy with complexes, modes are sets of informational relationships in Q-space that have denser internal structure (are more tangled) than their surroundings. Modes can be considered as ‘subshapes’ in Q-space.
Some Properties of Qualia Space What is the relevance of these constructs to understanding the quality of consciousness? It is not easy to become familiar with a complicated multidimensional space nearly impossible to draw, so it may be useful to resort to some metaphors. Perhaps
411
the most important notion emerging from this approach is that an experience (a quale in the broad sense) is a shape in Q-space. What gives each experience its particular shape are the informational relationships in Q-space (q-arrows between repertoires) generated by causal interactions among the elements of a complex. Only the informational relationships within a complex (those that give the quale its shape) contribute to experience. Conversely, the informational relationships that exist outside the main complex, for example, those involving sensory afferents – do not make it into the quale, and therefore do not contribute either to the quantity or to the quality of consciousness. By the same token, different experiences are, literally, different shapes in Q-space. For example, when the same system enters a different state (firing pattern), it will typically generate a different shape or quale (even for the same value of F). Moreover, experiences are similar if their shape is similar, and different to the extent that their shapes are different. Note that a quale can only be specified by a mechanism entering a particular state – it does not make sense to ask about the quale generated by a mechanism in isolation, or by a state (firing pattern) in isolation. On the other hand, it does make sense to ask what kind of shapes or qualia the same system (mechanism) can generate when it enters different states. The set of all shapes generated by the same system when it enters different states provides a geometrical depiction of all its possible experiences. Another consequence is that two systems entering the same state can generate two different experiences (i.e., two different shapes). In other words, the notion of a system state, or firing pattern for a neural system, is meaningless without taking into account the mechanism that produced the state. As an extreme example, a system that were to copy one by one the state of the neurons in a human brain, but had no internal connections of its own, would generate no consciousness and no quale. Note also that informational relationships are specified both by elements that are firing and others that are not. That is, a complex includes silent elements that are just as necessary in specifying the shape of the quale as are active ones. An intriguing corollary is that, within the main
412
An Integrated Information Theory of Consciousness
complex, disabling elements (neurons) that happen to be silent should change the quantity and quality of experience, though the system state (firing pattern) remains the same. It also follows that two systems with different architectures can generate the same experience (i.e., the same shape). For example, consider again the photodiode, whose mechanism determines that if the current in the sensor exceed a threshold, the detector turns on. This simple causal interaction is all there is, and when the photodiode turns on it merely specifies an actual repertoire where states [00 01 10 11] have, respectively, probability [0 0 ½ ½]. This corresponds in Q-space to a single q-arrow, one bit long, going from the potential, maximum entropy repertoire [¼ ¼ ¼ ¼] to [0 0 ½ ½]. Now imagine the light sensor is substituted by a temperature sensor with the same threshold and dynamic range – we have a thermistor rather than a photodiode. While the physical device has changed, according to the IIT the experience, minimal as it is, has to be the same, since the informational relationship that is generated by the two devices is identical. Similarly, an AND gate when silent and an OR gate when firing also generate the same shape in Q-space, and therefore must generate the same minimal experience (the two shapes have the same symmetries, i.e., are isomorphic). Thus, different ‘physical’ systems generate the same experience if the shape of the informational relationships they specify is the same. On the other hand, more complex networks of causal interactions are likely to create highly idiosyncratic shapes, so systems of high F are unlikely to generate exactly identical experiences. It is important to see what F corresponds to in this representation. The minimum information partition is just another point in Q-space: the one specified by the connections within the minimal parts only, leaving out the contribution of the connections among the parts. This point is the actual repertoire corresponding to the product of the actual repertoires of the parts taken independently. F corresponds then to an arrow linking this point to the top of the solid. In this view, the qpaths leading to the minimum information bipartition provide the natural ‘base’ upon which the solid rests – the informational relationships generated within the parts upon which are built
the informational relationships among the parts. The F-arrow can then be thought of as the height of the solid – or rather, to employ another metaphor, as the highest pole holding up a tent. For example, if F is zero (say a system decomposes into two independent complexes), the tent corresponding to the system is flat – it has no shape. Conversely, the higher the F value of a complex, the higher the tent or solid, the more ‘breathing room’ there is for the various informational relationships within the complex (the edges of the solid or the seams of the tent) to express themselves. In summary, and not very rigorously, the generation of an experience can be thought of as the erection of a tent with a very complex structure: the edges are the tension lines generated by adding each connection in turn (the respective q-arrow or informational relationship). The tent literally takes shape when the connections are engaged and specify actual repertoires. Perhaps an even more daring metaphor would be the following: whenever the mechanisms of a complex unfold and specify informational relationships, the flower of experience blooms.
From Phenomenology to Geometry The notions just sketched aim at providing a framework for translating the seemingly ineffable qualitative properties of phenomenology into the language of mathematics, specifically, the language of informational relationships (q-arrows) in Q-space. Ideally, when sufficiently developed, such language should permit the geometric characterization of phenomenological properties generated by the human brain. In principle, it should also allow us to characterize the phenomenology of other systems. After all, in this framework the experience of a bat echo-locating in a cave must be just another shape in Q-space and, at least in principle, shapes can be compared objectively. At present, due to the combinatorial problems posed by deriving the shape of the quale produced by systems of just a few elements, and to the additional difficulties posed by representing such high-dimensional objects, the best one can hope for is to show that the language of Q-space can capture, in principle, some of the basic distinctions
An Integrated Information Theory of Consciousness
that can be made in our own phenomenology, as well as some key neuropsychological observations. A short list includes the following: 1. Experience is divided into modalities, like the classic five senses (sight, hearing, touch, smell, taste and several others), as well as submodalities, like visual color and visual shape. What do these broad distinctions correspond to in Q-space? According to the IIT, modalities and submodalities are sets of densely entangled q-arrows (modes) that form distinct subshapes in the quale. As a two-dimensional analogue, imagine a given multimodal experience as the shape of the three-continent complex constituted by Europe, Asia, and Africa. The three continents are distinct subshapes, yet they are all part of the same landmass, just like modalities are parts of the same consciousness. Moreover, within each continent there are peninsulas (sub-subshapes), like Italy in Europe, just like there are submodalities within modalities. 2. Some experiences are homogeneous and others are composite: for example, a full-field experience of blue, as when watching a cloudless sky, compared to that of a market street. In Q-space, though not unimodal, homogeneous experiences translate to a single homogeneous shape, and composite ones into a composite shape with many distinguishable subshapes (modes and submodes). 3. Some experiences appear to be ‘primitives’ that cannot be further decomposed. A typical example are what philosophers call a ‘quale’ in the narrow sense, say a pure color like red, or a pain, or an itch: it is difficult, if not impossible, to identify any further phenomenological ‘structure’ within the experience of red. According to the IIT, such primitives correspond to subshapes (modes) in Q-space that cannot be further decomposed into sub-subshapes (primitive modes). 4. Some experiences are hierarchically organized. Take seeing a face: we see at once that as a whole it is somebody’s face, but we also see that it has parts such as hair, eyes nose and mouth, and that those are made in turn of specifically oriented segments. It can be shown that modes in Q-space lend themselves readily to hierarchical decompositions, where part–whole relationships are
5.
6.
7.
8.
413
represented by entangled informational relationships. We recognize intuitively that the way we perceive taste, smell, and maybe color, is organized phenomenologically in a ‘categorical’ manner, quite different from, say, the ‘topographical’ manner in which we perceive space in vision, audition, or touch. According to the IIT, these hard to articulate phenomenological differences correspond to different basic shapes in Q-space, such as 2n-dimensional grid-like structures and pyramid-like structures, which emerge naturally from the underlying neuroanatomy. Some experiences are more alike than others. Blue is certainly different from red (and irreducible to red), but clearly it seems even more different from middle C on the oboe. In the IIT framework, in Q-space colors correspond to different subshapes of the same kind (say pyramids pointing in different directions) and sounds to very different subshapes (say tetrahedra). In principle, such subjective similarities and differences can be investigated by employing objective measures of similarity between shapes (e.g., considering the number and kinds of symmetries involved in specifying shapes that are generated in Q-space by different neuroanatomical circuits). Experiences can be refined through learning and changes in connectivity. Say one learns to distinguish wine from water, then reds from whites, then different varietals. Presumably, underlying this phenomenological refinement is a neurobiological refinement: neurons that initially were connected indiscriminately to the same afferents, become more specialized and split into subgroups with partially segregated afferents. This process has a straightforward equivalent in Q-space: the single q-arrow generated initially by those afferents splits into two or more q-arrows pointing in different directions, and the overall subshape of the quale is correspondingly refined. Qualia in the narrow sense (experiential primitives) exist ‘at the top of experience’ and not at its bottom. Consider the experience of seeing a pure color, such as red. The evidence suggests that the ‘neural correlate’ of color, including
414
An Integrated Information Theory of Consciousness
red, is probably a set of neurons and connections in the fusiform gyrus, maybe in area V8 (ideally, neurons in this area are activated whenever a subject sees red and not otherwise, if stimulated trigger the experience of red, and if lesioned abolish the capacity to see red). Certain achromatopsic subjects with dysfunctions in this general area seem to lack the feeling of what it is like to see color, its ‘coloredness,’ including the ‘redness’ of red. They cannot experience, imagine, remember and even dream of color, though they may talk about it, just as we could talk about echolocation, from a third person perspective. Contrast such subjects, who are otherwise perfectly conscious, with vegetative patients, who are, for all intents and purposes, unconscious. Some of these patients may show behavioral and neurophysiological evidence for residual function in an isolated brain area. Yet it seems highly unlikely that a vegetative patient with residual activity exclusively in V8 should enjoy the vivid perceptions of color just as we do, while being otherwise unconscious. The IIT provides a straightforward account for this difference. To see how, call ‘r ’ the connections targeting the ‘red’ neurons in V8 that confer them their selectivity, and non-r (⌐r) all the other connections within the main corticothalamic complex. Considering r in isolation at the bottom of Q-space (null context), yields a small q-arrow pointing in a direction that represents how r by itself shapes the maximum entropy distribution into an actual repertoire. Schematically, this situation resembles that of a vegetative patient with V8 and its afferents intact but the rest of the corticothalamic system destroyed. The shape of the experience or quale reduces to this q-arrow, so its quantity is minimal (F for this q-arrow is obviously low) and its quality unspecified: as we have seen with the photodiode, r by itself cannot specify whether the experience is a color rather than something else, such as a shape, whether it is visual or not, sensory or not, and so on. By contrast, subtract r from the top of the complex (the set of all connection). This ‘lesion’ collapses all q-arrows generated by r starting from any context, called aq-fold. In particular,
the lesion collapses the q-arrow, called the upset of nonred, which starts from the full context provided by all other connections r and reaches the top of the quale. This q-arrow will typically be much longer and point in a different direction than the q-arrow generated by r at the bottom of the quale because, the fuller the context, the more r can shape the actual repertoire. Schematically, removing r from the top, a situation resembling that of an achromatopsic patient with a selective lesion of V8, leaves the bulk of the experience or quale intact (F remains high), but collapses a noticeable feature of its shape (the upset of nonred). According to the IIT, the upset of nonred captures precisely the quality or ‘redness’ of red. It is worth remarking that the last example also shows why specific qualities of consciousness, such as the ‘redness’ of red, while generated by a mechanism, cannot be reduced to a mechanism. If an achromatopsic subject without the r connections lacks precisely the ‘redness’ of red, whereas a vegetative patient with just the r connections is essentially unconscious, then the redness of red cannot map directly to the mechanism implemented by the r connections. However, the redness of red can map nicely onto the informational relationships specified by r, as these change dramatically between the null context (vegetative patient) and the full context (achromatopsic subject). To conclude, it is worth pointing out some outstanding issues that will need to be addressed in further developments of the theory. One of these is finding a principled way to determine the proper spatial and temporal scale to measure informational relationships and integrated information. What are the elements upon which probability distributions of states are to be evaluated? For example, are they neurons or minicolumns? Similarly, what is the ‘clock’ to use to identify system states? Does it run in milliseconds or hundreds of milliseconds? A working hypothesis is that the relevant spatial and temporal scales are those that jointly maximize F – different systems will generate maximal amount of integrated information at a particular spatiotemporal scale that is determined by their mechanism. Another important issue has to do with the relationship between complexes and the outside world.
An Integrated Information Theory of Consciousness
The mechanisms of a complex generate integrated information and informational relationships from within – as shown by the dreaming brain, an adult brain does not need the outside world to generate experience. However, the mechanisms inside the complex are what they are, and so is the quality of the experience they generate, by virtue of a long evolutionary history, individual development, and learning. In fact, it appears that as a system incorporates statistical regularities from its environment and learns, its capacity for integrated information may grow. It will thus be important to see how the informational relationships (q-arrows) inside a complex reflect and react to informational relationships existing in the world.
Conclusion To recapitulate, the IIT claims that the quantity of consciousness is given by the integrated information (F) generated by a complex of elements, and its quality by the shape in Q-space specified by all the informational relationships they generate. As suggested here, this theoretical framework can account, at least in principle, for several basic neurobiological and neuropsychological observations. Moreover, the same theoretical framework can be extended to translate phenomenology into the language of mathematics. At present, the very notion of a theoretical approach to consciousness may appear far-fetched, yet the nature of the problems posed by a science of consciousness seems to require a combination of experiments and theories: one could say that theories without experiments are lame, but experiments without theories are blind. Only a genuine theoretical framework can go beyond proposing a provisional list of candidate mechanisms (reentry, synchronization, broadcasting) or brain areas (frontoparietal networks, default system) without a principled explanation of why they may be relevant. And only a theory can account, in a coherent manner, for several key but puzzling facts about consciousness and the brain, such as the association of consciousness with the corticothalamic but not the cerebellar system, the ‘unconscious’ functioning of many cortico-subcortical circuits or the fading of consciousness during certain stages of sleep or epilepsy despite continuing neural activity.
415
A theory should also generate relevant corollaries. For example, the IIT: predicts that consciousness depends exclusively on the ability of a system to generate integrated information, whether or not it is immersed in an environment, it has language, capacity for reflection, attention, episodic memory, a sense of space, of the body, and of the self, contrary to some common intuitions, but consistent, as reviewed elsewhere, with the overall neurological evidence. Of course, the theory recognizes that these same factors are important historically because they favor the development of neural circuits forming a main complex of high F. Finally, a theory should be able to help in cases that challenge our intuition or our standard ways to assess consciousness. For instance, the IIT says that the presence and extent of consciousness can be determined, in principle, also in cases in which we have no verbal report, such as infants or animals, or in neurological conditions such as minimally conscious states, akinetic mutism, psychomotor seizures, and sleepwalking. In practice, of course, measuring F accurately in such systems will not be easy, but approximations and informed estimates are certainly conceivable. The theory also implies that consciousness is not an all-or-none property, but is graded: specifically, it increases in proportion to a system’s repertoire of discriminable states. In fact, any physical system with some capacity for integrated information would have some degree of experience, irrespective of the constituents of which it is made, and independent of its ability to report. Whether these and other predictions turn out to be compatible with future clinical and experimental evidence, a coherent theoretical framework should at least help to systematize a number of neuropsychological and neurobiological results that might otherwise seem disparate. See also: Neurobiological Theories of Consciousness.
Suggested Readings Baars BJ (1988) A Cognitive Theory of Consciousness. New York, NY: Cambridge University Press. Balduzzi D and Tononi G (2008) Integrated information in discrete dynamical systems: Motivation and theoretical
416
An Integrated Information Theory of Consciousness
framework. PLoS Computational Biology 13;4(6): e1000091. Cover TM and Thomas JA (2006) Elements of Information Theory, 2nd edn. Hoboken, NJ: Wiley-Interscience. Crick F and Koch C (2003) A framework for consciousness. Nature Neuroscience 6(2): 119–126. Dehaene S, Sergent C, and Changeux JP (2003) A neuronal network model linking subjective reports and objective physiological data during conscious perception. Proceedings of the National Academy of Sciences USA 100(14): 8520–8525. Feldman J (2003) A catalog of Boolean concepts. Journal of Mathematical Psychology 47(1): 75–89. Gazzaniga MS (2005) Forty-five years of split-brain research and still going strong. Nature Reviews Neuroscience 6(8): 653–659. Hobson JA, Pace-Schott EF, and Stickgold R (2000) Dreaming and the brain: Toward a cognitive neuroscience of conscious states. Behavioral Brain Science 23(6): 793–842; discussion 904–1121. Koch C and Tononi G (2008) Can machines be conscious? Spectrum IEEE 45(6): 55–59. Massimini M, Ferrarelli F, Esser SK, et al. (2007) Triggering sleep slow waves by transcranial magnetic stimulation. Proceedings of the National Academy of Sciences USA 104(20): 8496–8501.
Massimini M, Ferrarelli F, Huber R, Esser SK, Singh H, and Tononi G (2005) Breakdown of cortical effective connectivity during sleep. Science 309(5744): 2228–2232. Posner JB and Plum F (2007) Plum and Posner’s Diagnosis of Stupor and Coma. 4th edn. Oxford; New York: Oxford University Press. Steriade M, Timofeev I, and Grenier F (2001) Natural waking and sleep states: A view from inside neocortical neurons. Journal of Neurophysiology 85(5): 1969–1985. Tononi G (2004) An information integration theory of consciousness. BMC Neuroscience 5(1): 42. Tononi G and Edelman GM (1998) Consciousness and complexity. Science 282(5395): 1846–1851. Tononi G and Laureys S (2009) The neurology of consciousness: An overview. In: Laureys S and Tononi G (eds.) The Neurology of Consciousness. Elsevier. Tononi G and Sporns O (2003) Measuring information integration. BMC Neuroscience 4(1): 31. Tononi G, Sporns O, and Edelman GM (1996) A complexity measure for selective matching of signals by the brain. Proceedings of the National Academy of Sciences USA 93(8): 3422–3427. van Zandvoort MJ, Nijboer TC, and de Haan E (2007) Developmental colour agnosia. Cortex 43(6): 750–757.
Biographical Sketch
Intentionality and Consciousness R Menary, The University of Wollongong, Wollongong, NSW, Australia ã 2009 Elsevier Inc. All rights reserved.
Glossary Consciousness – Can be usefully classified in three ways: (1) phenomenal consciousness, the way things feel to us; (2) access consciousness, a mental state is poised for use in reasoning and direct rational control of thought and action; and (3) reflexive consciousness, the ability to think about one’s own thoughts. Embodied mind – The view that consciousness and intentionality should be understood in terms of our bodily engagements with the world and not by bracketing off the external world. Intentionality – The directedness of mental states, captured by the phrase what are you thinking about? Naturalism – The attempt to give an explanation of intentionality, consciousness, and representation in terms that are consistent with the natural sciences. Representation – States that present, or stand in for an object. Most theories take a representation to have a content that can be specified by a proposition such as ‘‘the Earth is spherical.’’
Introduction: Mind, Consciousness, and Intentionality Conscious minds have both intentional and phenomenal properties. If a mental state is intentional then it is directed toward some object (it is about something). If a mental state is phenomenal then it feels a particular way to us. Cognitive states such as beliefs and desires are obviously intentional. My belief that there is a cold beer on the counter in front of me is directed at an object (the cold beer); my desire to drink that beer is also aimed at an
object (the very same beer). My visual perception of the beer on the counter is also intentionally directed as is the tactile sensation of grasping the cold bottle in my hand. The tactile sensation also feels a certain way; the bottle feels cold. Hence some conscious experiences are both intentionally directed and feel a particular way. One way to think of the relationship between consciousness and intentionality is that all conscious states are also intentional. When I am consciously aware of a sensation (a sound), then this experience is certainly phenomenal; but the experience is also intentional – because the sensation (the sound) is the object or content of my awareness. Many philosophers and cognitive scientists take consciousness to be the defining property of the mental, and they distinguish between consciousness as perceptual awareness (awareness of sights, sounds, smells, etc., which we share with some other animals) and consciousness as awareness of the contents of my experiences and thoughts, as ‘my’ experiences and thoughts. Cognitive states such as beliefs and desires, and perceptual states such as looking at a blue mug or hearing a high C are intentional; they are directed at something other than themselves. Perceptions are intentional in so far as I am aware of them as having sensory objects (at which they are directed), light, sound, etc. Other phenomenal states such as itches and pains are less obviously intentional, when I am aware of my pain then the object of my awareness is the pain. It is not so obvious what the object or content of the pain itself is, consequently some thinkers deny that phenomenal feels (or qualia) such as pains and itches are intentional. Conscious awareness of ‘my’ cognitive states, perceptual states, and phenomenal feels is intentional and this has been recognized for a long time. Take, for example, the following section of Descartes’ second meditation: Finally it is I who has sensations, that is to say, who is aware of objects as though by the senses, since indeed
417
418
Intentionality and Consciousness
I see light, I hear noise, I feel heat. But all these objects are unreal, since I am dreaming. Let it be so; certainly it seems to me that I see, I hear, and I feel heat. That cannot be false; that is what in me is properly called sensation; and in this precise sense, sensation is nothing but thought.
By contrast, Towser the dog may ‘desire’ the bone and may ‘believe’ that the bone is under the bush. But Towser does not think ‘‘I am thinking that there is a bone under the bush.’’ Towser may have beliefs and desires, or equivalent intentional states, but is not conscious of them as beliefs and desires and certainly not as being his beliefs and desires. We can, at this point, distinguish between intentional states, which are cognitive/conceptual such as beliefs and desires and intentional states, which are perceptual/nonconceptual such as sights and sounds. There are those such as McDowell who do not wish to make this distinction, claiming that there is only conceptual intentionality. Conceptual content can be thought of as the content of a thought that can be propositionally described. My thought that the beer in front of me is ice cold is just such a thought with conceptual content. There are strong reasons to favor an explanation of mindedness as being both intentional and conscious (especially with the capacity for selfawareness). However, it is not so clear whether consciousness should be understood in terms of intentionality (some conscious states might not be intentional), nor whether intentionality should be understood in terms of consciousness (Towser’s cognitive states are intentional, but not conscious). What then is the relationship between intentionality and consciousness? Is intentionality a prerequisite for consciousness or is consciousness a prerequisite for intentionality? Opinions, unsurprisingly, differ on these questions. There are further questions about the relationship between intentionality and consciousness. If Brentano is right and intentionality is the hallmark of the mental, then presumably all conscious experiences must be intentional. If some conscious experiences are not intentional, it would appear to follow that they are also not mental. It would appear to be absurd to endorse the view that some of our conscious experiences are not mental phenomena.
However, as noted above, it is not clear that all conscious experiences are describable as having conceptual content. For example, pains are conscious experiences, in what sense might they be thought of as having conceptual contents? Some philosophers, such as Sellars, think that phenomenal feels are not intentional. If we followed Sellars in this then we would mark off conscious states that are intentional, such as beliefs, desires, and perceptions from those that are nonintentional, such as itches and tickles because they do not have content. Furthermore, it may turn out that consciousness outstrips intentionality and conversely that some intentional states, such as representations, might not be accessible to consciousness (although both of these claims are controversial). Consequently the concepts of intentionality and consciousness may overlap at various points, but are not mutually defining: Not all conscious states are intentional; not all intentional states are conscious. These issues will be explored in sections ‘Intentional inexistence,’ ‘Representational theories of consciousness,’ and ‘Is phenomenal consciousness intentional?.’ Finally there is the question of whether intentionality and consciousness can be talked about independently of, or bracketed off from, the body and its relationships with the world. Increasingly research in cognitive science is turning to an embodied and embedded approach to intentionality and consciousness, rather than an approach which solipsistically considers only states internal to the mind or brain. In the embodied approach both intentionality and consciousness become conceived of in terms of our embodied engagements with the world. Therefore, it is still an open and important question to ask what the relationship between intentionality and consciousness is. Before considering this question I shall give more precise definitions of intentionality and consciousness in the next two sections entitled ‘Intentionality’ and ‘Consciousness.’ Then I shall go on to explore some of the puzzles that are produced by Brentano’s theses about intentionality and then tackle the issue of whether phenomenal consciousness is intentional. Finally I shall look at the phenomenological tradition and the embodied approaches to intentionality that draw upon this tradition, as well as the externalism of the analytic tradition.
Intentionality and Consciousness
Intentionality Intentionality is usually defined as the directedness of the mind toward something other than itself. I can be thinking about the ice cold beer sitting in front of me, I might want the beer to be ice cold, or hope that it is. The mental states of wanting and hoping, in this case, are about the beer sitting before me. Consequently, another definition of intentionality is the aboutness of mental states such as hopes, desires, and beliefs; they are about something other than themselves. The notion of intentionality can be neatly captured by the everyday question ‘‘what are you thinking ‘about’?’’ The term intentionality should not be confused with the normal English word intention, I intend to do something. This is despite the fact that both uses are derived from the Latin intentio, which literally means to apply tension, but also to point at or extend toward. Scholastics of the middle ages, such as Thomas Aquinas, used the term in both the more familiar sense of intending to do something and in the restricted technical fashion (direction toward an object) that is still current today. Intentionality becomes interesting when we look at thoughts that are not about objects which are within our perceptual range. For example, I might be sitting in the office thinking about the ice cold beer that I will quaff with gusto at the end of the day. My thoughts are directed at an object that is not sitting before me. My thoughts might also turn to objects that don’t, nor ever will exist such as phlogiston, or Santa Claus or Hamlet. It is quite normal for us to think about nonexistent objects, but what then are my thoughts directed at? This is a puzzle that any account of intentionality must solve, as well as the, more straightforward, intentional phenomenon of directedness at existing objects. The nineteenth century psychologist and philosopher Franz Brentano gave the formulation of intentionality, which underpins contemporary thinking in philosophy and the cognitive sciences today. Brentano claimed that intentionality was the defining feature of the mental, what he called the mark of the mental; all minds are intentional. Furthermore, he claimed that nothing physical exhibited intentionality; there are no physical objects which are intentional. He concluded that intentionality was
419
the feature of mentality, which allows us to differentiate between minds and physical phenomena and that minds were not physical entities. It is one of the central tasks of contemporary philosophy, psychology, neuroscience, and artificial intelligence (what might be collectively described as the cognitive sciences) to explain how physical entities can have minds, or more precisely how physical systems can also be intentional systems.
Consciousness Ned Block has distinguished three different senses of consciousness. The first is phenomenal consciousness (P-consciousness), the familiar sense of conscious experience as what it is like. Conscious experiences have a phenomenal character, or to put it another way, they have sensory qualities (otherwise known as qualia). There is something it is like for me to see the deep blue of the Pacific Ocean and to feel the warm water lapping over my feet, and to smell the briny breeze. The second is access consciousness (A-consciousness). A conscious state is A-conscious if it is poised for use in reasoning and direct rational control of thought and action. P- and A-consciousness although distinct will often interact: my becoming A-conscious of the shirt on the back of my neck changes my P-conscious state. Finally there is reflexive or monitoring consciousness. This amounts to awareness of ones own conscious experiences and thoughts. Classically this was known as introspection and was often thought of as a kind of inner sense directed at conscious experiences, but more recently reflexive consciousness has been thought of as a kind of representation of the contents of consciousness, especially by Rosenthal and Dennett. A major point of disagreement is whether phenomenal consciousness is intentional. This leads to a question about explanatory priority: Is phenomenal experience necessary for intentionality or are all phenomenal experiences simply the way in which intentional states (representations) are experienced? Before proceeding to deal with these questions in sections entitled ‘Is phenomenal consciousness intentional’ and ‘Embodied intentionality,’ I will outline some of the standard approaches to intentionality in the analytic tradition. I shall
420
Intentionality and Consciousness
introduce the phenomenological tradition in the final two sections entitled ‘The phenomenological tradition’ and ‘Embodied intentionality.’
Intentional Inexistence Brentano bequeathed us three important claims concerning intentionality: 1. The directedness of mental states: All mental states are directed at something beyond themselves. 2. Intentional inexistence: All mental states are directed at an object or content that is internal to the mind. 3. Brentano’s thesis: Only mental states exhibit the two above properties, this marks off the mental from the physical. The first two claims are clearly related and jointly provide reason to believe in the truth of the third. While the first claim is easy enough to grasp, what Brentano meant by intentional inexistence is not so clear and the section entitled ‘Analytic approaches to intentionality’ will be devoted to giving an account. Section entitled ‘Analytic approaches to intentionality’ will look at how the first two claims have been dealt with in the analytic tradition, stretching from Frege and Russell to contemporary philosophy of language and mind. Sections entitled ‘Brentano’s thesis’ and ‘Naturalizing representations and content’ will look at the final claim, often referred to as Brentano’s thesis and recent attempts to show that the thesis is false and that physical systems can also be intentional systems. As noted above we can think about things that don’t exist as well as things that do and this makes intentionality a difficult phenomenon to explain. Brentano attempted to explain this by claiming that mental states intentionally contain an object in themselves and this is what he called intentional inexistence. Mental states such as wanting, believing, hoping, and fearing are directed at objects, but these objects don’t always exist, for example I want world peace, but the world is lamentably war torn. To deal with this Brentano claimed that mental states are not directed at actual existing objects out there in the world, but directed at an object or content internal to themselves. Mental states can still be about actual existing objects in the world, but only by being directed at mental contents.
Thus, we find Brentano (in Psycholog y from an Empirical Standpoint) saying that the intentional relation is a quasirelation, because it is not a relation between two existing things – only the thinker’s mind need exist. Therefore, Brentano’s account of intentionality is a form of internalism, the intentional relation is entirely contained within the boundaries of the mind (this should be contrasted with externalism, see section entitled ‘Naturalizing representations and content’). It is not entirely clear what Brentano meant by a mental state intentionally containing an object in itself, an object at which it is directed. More recently Roderick Chisholm has attempted to clarify what he could have meant. When Brentano says that mental states intentionally contain an object within themselves, those states truly have an object even if that object does not exist. So Diogenes’ looking for an honest man would still have the same object (an honest man) even if there are no such things. However, this kind of relation to an object never occurs in nonpsychological phenomena; for example, for Diogenes to sit in his tub there must be a tub for him to sit in. As Chisholm points out although there is a relation between Diogenes and an object in both examples above, the psychological relation is of a peculiar sort because it can hold, even though one of its terms (an honest man) does not exist. Minds can be intentionally related to things that exist within the mind exclusively. This reduces the initial plausibility of the definition of intentionality that it is the direction of the mind toward something external to itself, for it now appears that the mind is directed only at something internal to itself. An alternative view is that there are intentional objects at which the mind is directed, including nonexistent ones. This was famously argued by one of Brentano’s pupils, Meinong, and has been subject to stringent criticism, especially by Russell and Quine (see section entitled ‘Analytic approaches to intentionality’). There are three puzzles that Brentano’s account of intentionality give rise to 1. How can two different beliefs be about the same thing? 2. What are thoughts about nonexistent things about? 3. How can Intentionality be reconciled with a physicalist ontology?
Intentionality and Consciousness
Before looking at contemporary attempts to answer these questions (especially question three) we shall look at the historical antecedents in the analytic tradition.
Analytic Approaches to Intentionality Early analytic philosophy has provided an answer to the first two questions from the end of the previous section entitled ‘Intentional inexistence’ and has provided the impetus for contemporary accounts of intentionality. Frege and Russell were primarily concerned with the logical structure of language, but since most analytic philosophers take language to be the expression of thought, it is also a method for studying the intentional structure of thought. The intentional structure of the propositional attitudes – hope that, fear that, believe that, etc. – can be revealed in an analysis of the sentences used to express them. One revealing claim is that sentences express, or have as their contents, the very same propositional contents of mental states (such as the propositional attitudes). Therefore my belief that Paris is the capital of France has as its content the very same proposition as my utterance: ‘‘I believe that Paris is the capital of France.’’ They have the ‘same’ content (this is what I earlier called conceptual content). We should, therefore, think of the contents of mental states and sentences as independent of those states and sentences. If we share the same belief – that the world is round – then this is not because our beliefs are numerically identical (one and the same); the identity is one of content, we believe the same proposition (that the world is round). Furthermore, the contents of singular thoughts such as beliefs are propositions, not sentences of a language. To play this role, propositions must be both abstract and objective. If I were to translate the belief sentence, (1) ‘‘I believe he is here,’’ into German, I translate it as ‘‘ich glaube er ist hier.’’ If beliefs merely have sentences as their contents (and not propositions) then I am saying (2) ‘‘I believe the sentence, ‘he is here,’ ’’ which would be translated as ‘‘Ich glaube den satz ‘he is here.’ ’’ But we would not translate in this way, (3) it should be translated as, ‘‘Ich glaube den satz ‘er ist hier.’ ’’ The sentence is not what I believe (the
421
sentence token ‘‘he is here’’), but the meaning of the sentence, the proposition. Frege explains this by introducing the distinction between sense and reference. The reference of a name is the object to which that name refers, for example: plant, chair, quark, etc. A classical example to illustrate the difference between sense and reference is the planet Venus. The Greeks believed that there were two different stars that rose in the morning and the evening – not a single planet, Venus. The reference of the words Hesperus and Phosphorus is the same object – Venus. However, the senses of Hesperus, the evening star, and Phosphorus, the morning star, are not the same. Words or expressions which have the same meaning, that is, are synonymous, are called ‘intensionally equivalent;’ and words or expressions which have the same reference or extension are called ‘extensionally equivalent.’ Take the following two statements: ‘creature with a heart’ and ‘creature with a kidney,’ both of these statements are extensionally equivalent, any creature with a heart is also a creature with a kidney. They are not, however, intensionally equivalent, they do not mean the same thing. Our examples show that words or expressions may be extensionally equivalent without being intensionally equivalent. The first puzzle now comes into play. Meno believes that Phosphorous rises in the morning and he believes that Hesperus rises in the evening. These are two different beliefs, they are not intensionally equivalent, yet they are extensionally equivalent – they are both about the planet Venus, how can this be the case? For Frege the sense of a sentence is the ‘proposition’ it expresses, however, the reference of sentences is not simply a spatiotemporal object. This is because Frege held that the reference of a sentence was the truth – value of that sentence. Frege held that a thought is what is expressed in a proposition, and a proposition is a function with a value, which is always a truth value. Frege took his inspiration for this conclusion from mathematics. The reference of a whole expression, such as X2 + 2, is its value. X2 + 2 has as its value the number 6, when we fill in the variable X with the argument 2. The expressions X2 + 2 = 6 and X is bald, have as their values ‘the true,’ when we fill in their variables with the arguments 2 and Socrates.
422
Intentionality and Consciousness
The sense of a sentence or proposition is the thought expressed by it. The reference of the proposition ‘Paris is the capital of France’ is its truth value, its sense is the thought that ‘Paris is the capital of France.’ Thoughts refer to truth values. Thoughts are not subjective. Thoughts and their relations to truth are objective and mind independent. They must be publicly accessible and objective, they are also abstract and nonphysical, much like numbers, according to Frege. The meaning of names and propositions is made up of two elements: their reference and their sense. Returning to the first puzzle (how can two different beliefs be about the same thing?) Frege presents a solution because Meno’s beliefs have different senses, but the same reference. Or, to put it another way, each belief presents its referent differently (senses are modes of presentation) to the thinker (Meno). In considering the second puzzle (what are thoughts about nonexistent things about?) we must turn to Frege’s contemporary Bertrand Russell. Russell made a distinction between two ways of knowing: Knowing by acquaintance and knowing by description. When I know by acquaintance, I, the knowing subject, stands in a direct relation to some object of awareness. ‘‘I say that I am ‘acquainted’ with an object when I have a direct cognitive relation to that object, for example, when I am directly aware of the object itself.’’ When Russell says ‘directly aware,’ we should not think in terms of a direct theory of perception. This is because, the direct objects of acquaintance are things such as: Sense – data, memories (which are particulars); and concepts of redness or roundness (which are universals). Particulars and universals are, therefore, the only things that we can know by acquaintance. When we know something by description we are not directly acquainted with the object of knowledge, and descriptions are introduced by sentences of the form, ‘‘a so-and-so’’ or ‘‘the so-and-so.’’ Physical objects and other minds are not objects of acquaintance, so they must be known by description. The sentence form, ‘‘a soand-so,’’ is an ambiguous description; and the sentence form, ‘‘the so-and-so,’’ is a definite description. As a consequence of Russell’s theory, nouns and proper names are considered to be descriptions. This is because the logical form of proper names can only be revealed as a description.
In sentences that contain names and nouns that have no denotation (i.e., they are fictional), such as ‘the present King of France is bald,’ or ‘the Unicorn is white,’ the subject expression does not denote, yet the sentences are meaningful. Here our second puzzle enters the stage, what does the sentence (or thought) ‘the present King of France is bald’ denote? Russell had earlier accepted Meinong’s solution to this problem (Meinong was a pupil of Brentano’s). Meinong held a realist theory of ‘objects,’ objects both exist and subsist. Those objects that exist form only a small class of objects, whereas, the class of the objects of knowledge, which is often nonexistent, is large and subsists; in the sense that they are the objects of thought and talk. This position requires that all possible negative facts subsist, such as England did not win the Soccer World Cup in 2006, and all impossible objects, such as, round squares. Consequently Meinong held a view that for any thought, it had an intentional object that either existed or subsisted – this was Meinong’s way of explaining intentional inexistence. Russell came to believe that this Meinongian realism offended against his sense of reality. Russell devised the theory of descriptions to avoid Meinong’s ontological profligacy, yet maintained the denotative theory of meaning. Therefore, most proper names and nouns are concealed descriptions. Only logically proper names denote, so how can descriptions be meaningful if one accepts a denotative theory of meaning? Russell’s answer was to apply a logical analysis to sentences that appeared to denote objects that do not in fact exist. The two most famous examples that Russell gave of sentences containing definite descriptions were 1. The present King of France is bald. Russell’s analysis of this phrase will give us: 1a. There is a King of France. 1b. There is not more than one King of France, and 1c. Anything which is the King of France is bald. ð∃xÞðFx & ½ð8 yÞ ðFy ! y ¼ xÞ & GxÞ
1a. is false, therefore 1. is false. 1a. is defined in terms of the existential generalization (there is an x); 1b. gives the uniqueness condition as determined by the use of the definite article ‘the’ (i.e.,
Intentionality and Consciousness
‘the King of France’); 1c. defines the existent x as having a particular property, baldness. Sentence 1 is an example of what Russell meant when he claimed that the grammar of language can be misleading. It appears as if the description has a denotation, however, in the formal paraphrase there are no denoting singular terms, only variables bound by quantifiers (along with predicates and identity). Furthermore because 1a is false there is no need to suppose that there is an object to which 1 refers, it only appears that there is, because we expect there to be an extension of the subject of predication uniquely picked out by the definite article. In his definitive paper ‘On Denoting,’ Russell says: ‘the present King of France is bald’ is certainly false; and ‘the present King of France is not bald’ is false if it means ‘There is an entity which is now King of France and is not bald’ but it is true if it means ‘It is false that there is an entity which is now King of France and is bald.’
However, Russell’s analysis of definite descriptions has been criticized by Strawson and Donellan and recent theories of direct reference have superseded the account to some degree. Frege and Russell produced analyses that were supposed to solve some of the puzzles produced by Brentano’s analysis of intentionality. They also began a century long investigation of intentionality and reference, which has taken a naturalistic turn in more recent philosophical work and this, leads us directly to the next puzzle. Brentano’s Thesis Brentano’s thesis is that intentionality marks off the mental from the physical. The property of intentional inexistence is a property to be found only in mental states and never in physical objects or processes. Chisholm brought this thesis to the attention of the late twentieth century philosophers in arguing that intentional states such as belief could not be accounted for simply in terms of behavior or behavioral dispositions. This is because such behavioral analyses are circular; they depend upon other intentional states. To be able to explain my behavior in terms of my desire for something I also have to explain it in terms of my beliefs about it, therefore if I desire a beer, my behavior is also explained in terms of my beliefs concerning what a beer is and where I can procure one. Chisholm concluded, in line with Brentano’s thesis, that
423
behavioral, reductive, or physicalist explanations of intentionality were doomed to failure. Contemporary philosophers of mind and psychology have not agreed with Brentano and Chisholm’s conclusion. They have taken up the challenge to provide a naturalistic or physicalist explanation of intentionality. One way to answer Brentano’s challenge is simply to demonstrate that there are physical or nonmental entities that exhibit intentionality. One obvious contender is language, sentences of natural language can exhibit intentionality: they can be directed at something other than themselves. A reason for rejecting a linguistic explanation of intentionality has been proposed by John Searle and that is the distinction between original and derived intentionality. Sentences of language do not have any intrinsic meaning or content, they have meaning and content conferred upon them by people who use sentences to express their thoughts and it is these thoughts which intrinsically have meaning or content. Consequently the mental states expressed by sentences have original intentionality, whereas the sentences themselves have merely derived intentionality. Another response has been proposed by Daniel Dennett who takes it that intentional idioms do not describe any actual phenomena, but they do have instrumental value for predicting the behavior of complex physical systems such as human beings. Although there are no actual intentional states or objects on Dennett’s account, the intentional idiom of beliefs and desires and other intentional states is required, to be able to predict and explain the behavior of others, in place of a complex physical story about how our brains and bodies work. An even more ambitious answer to the challenge is to demonstrate that there are physical systems which are nevertheless intentional systems. Fred Dretske has given a detailed information theoretical account, which shows that there are many natural states which indicate features of their local environment. This brings us to the representational account of intentionality. Naturalizing Representations and Content The Representational Theory of Mind (RTM) is an attempt to explain intentional mental states in terms of the concept of representation. As such,
424
Intentionality and Consciousness
intentional states such as beliefs, fears, hopes, and perceptions are inner states of people, perhaps brain states, which have representational content. The hope is that intentionality can be explained by the more familiar notion of representation, and that representations are better candidates for being physical states. The RTM posits semantically evaluable representations that enter into causal connections with other representations and behavior. Beliefs and desires are semantically evaluable, beliefs can be true or false, desires can be satisfied or frustrated, etc. It is assumed that what makes beliefs true or false is their relation to the external world. So, we give beliefs and desires semantic evaluations just because we evaluate them in terms of their relation to the world. However, beliefs and desires are attitudes to ‘something’ and we can call this ‘something’ the belief ’s content (what was above called conceptual content). It is the content, or proposition, (hence propositional attitude) that is semantically evaluable (here we see the continuity with the early analytic tradition exemplified by Frege and Russell). An example: Hamlet believes that his uncle killed his father. The belief has a semantic value, it is a true belief. The content of Hamlet’s belief is that ‘‘Hamlet’s uncle killed his father’’ (the content of the belief is the proposition, ‘‘Hamlet’s uncle killed his father’’). Knowing the content of a belief allows you to know what it is about the world that determines the semantic evaluation of the belief. Hence, intentional states are to be understood in terms of representation. I now turn to two attempts to give a naturalistic explanation of representation. Dretske’s indicator semantics Dretske wants to build his account of representation from the bottom-up. We derive an account of representation and intentionality, from a natural, nonintentional, account of indicators and their causal relation to what they indicate. Dretske begins with an analysis of natural signs or indicators. Tracks in the snow indicate the previous presence of an animal; the width of tree rings indicates the amount of rainfall there was in a year, etc. The indicator is dependent on the presence of the indicated, in other words, there would be no tracks in the snow (indicator) if the animal
(indicated) had not been present. The dependency of indicator on indicated is causal, the indicator covaries with the indicated. Natural signs derive their indicative powers from the way in which they are objectively related to the conditions that they signify. The 24 rings of a tree stump indicate that the tree is 24 years old and the fact that the 24 rings indicate that the tree is 24 years old could not be the case if the tree was not 24 years old. The power of indication is dependent upon the relation between the indicator and what it indicates. The relation should be lawful or causal in character. Therefore, an indicator C indicates E, iff C indicates the presence of E only when E is present in the environment. Hence, for Dretske there is no such thing as misindication. Indicators become representations, which can misrepresent when a representation is biologically supposed to indicate an environmental stimulus. The indicational content of a representation is explained in terms of the adaptive function of that representation. A neural state N has the function of representing food when it indicates the presence of food in the environment, and it is recruited in the service of the function of the organism’s moving toward and consuming the food. How might an internal state I acquire its function of indicating? Dretske is clear on this matter: I is recruited as a cause of some behavior B, because of what it indicates about a state of the environment E. Once I becomes a cause of B it acquires the function of indicating E – it comes to represent E. I becomes a representation when its natural meaning acquires an explanatory relevance. Dretske uses the example of northern hemisphere marine bacteria to illustrate the idea. These bacteria contain magnetosomes, which align the bacteria parallel to the Earth’s magnetic field (the indicators). The bacteria are propelled downward, to the geomagnetic north, away from the oxygen-rich surface water, which is toxic for the bacteria (the indicators are recruited to help perform this function). We must be clear about the notion of recruitment at work here. The recruitment involves a process of selection or reinforcement, which establishes a link between I and B, by virtue of the consequences of producing B in certain circumstances. In the bacteria example, recruitment of the
Intentionality and Consciousness
indicator is selected for to produce movement toward the geomagnetic north. However, for cases where there is a genuine belief, learning is the recruitment process. It is during this period of recruitment that I comes to represent E, I can now misrepresent E even though before the period of recruitment, when it was a mere indicator of E, I could not misrepresent E. Millikan’s biosemantics Millikan’s understanding of proper biological functions allows us to understand how there could be representation in the biological world. Millikan shows that the production and consumption of representational vehicles, what she calls intentional icons, are biological functions and the normativity of representations is derived from the normativity of biological functions, which she calls proper functions. Proper functions are normative, in the sense that a device might have a proper function even though it fails to perform it. Here the possibility of misrepresentation might be made clear. What allows for the continuance of a proper function throughout generations? For a device/ organism to have a proper function, it must share this function in common with its ancestors. You and I both have hearts that pump blood, because we share a common ancestor whose heart had the proper function of pumping blood. Proper functions are copied and reproduced through generations. However, we know (according to the best neo-Darwinian accounts) that no heart is directly copied from any other heart, rather, it is the genes which are directly copied and it is these genes that have the proper function of producing hearts. The normal explanation of the performance of a proper function makes reference to the normal conditions under which, historically, the proper function has been performed and selected for. This is illustrated by Millikan’s bee dance example. There are mechanisms in bees that have the proper function of producing a bee dance. There are also mechanisms in bees that have the proper function of consuming bee dances. The proper function of the bee dance producer (more strictly the relational proper function, because the function is related to something in the organism’s environment) is to produce the consequence that consumer bees fly off in the direction of the orientation of the
425
bee dance. The relational proper function of the bee dance producer is selected for iff in normal conditions it has, historically, led bees to find flowers, pollen, nectar, food – that which optimizes survival value. The consumer mechanism gets selected for iff, under normal conditions it has, historically, produced behavior leading to flowers, nectar, etc. on the basis of the consumed bee dances. The producer mechanism has the function of producing intentional icons for consumer mechanisms and consumer mechanisms have the function of consuming the intentional icons produced by producer mechanisms for some further end. This requires that the producer and consumer mechanisms can only function properly if they are both present and coordinating; this is the normal condition for the mechanisms to function properly. A normal condition of the environment, the location of nectar, has the effect of producing bee dances. These have the effect of sending consumer bees to the location of the nectar. This in turn produces two normal conditions in the environment, the nectar being located in the hive and flowers being pollinated at the first location of the nectar. For the mechanisms to function properly the normal conditions must be in place. It is quite easy to see how contingent factors could interfere with the normal conditions of proper functioning, but this is why biological functions are normative; it is the proper function of mechanisms in normal conditions that is selected for. The next issue concerns why the bee dance is a representation. Is this really representation? Firstly the distinction between proper functioning and malfunctioning looks secure. This could, in principle, underwrite the normative notion of content demanded by the possibility of misrepresentation. Secondly, the relationality of some proper functions gives them an intentional aspect, in Brentano’s sense, they are directed at something beyond themselves. The intentional icons have three properties: 1. They are relationally adapted to some feature of the world. 2. The relation can be seen in terms of a ‘mapping.’ 3. The icons have the proper function of guiding a consumer mechanism in performing its proper function.
426
Intentionality and Consciousness
If the relational conditions for this bee dance (qua intentional icon) are normal, then it will successfully map the location of flowers, etc., call this indicative mapping. If this is successful, then the icon directs the consumer bee to the location of the nectar, call this imperative mapping. It is in the consumption of an icon that the representational function is established. The direct proper function of an icon is the effect it ought to produce (sending consumer bees in the direction of nectar), not what it statistically does produce. Millikan provides an account of biological norms and how representation can arise according to these norms. This is quite different from Dretske’s indicator function theory. The indicator relation is a causal relation that is recruited for some end. On Millikan’s approach intentional icons represent because they are consumed by another mechanism and this relationship constitutes the proper function (hence the norm). Both these naturalistic accounts of representation (hence intentionality) are externalist in character. The content fixing relations and norms are based in the external environment of the organism, hence externalists seek to explain intentionality as a relation between representations and the world and not as a relation between representations and inner mental objects. Unsurprisingly the naturalistic turn to analyzing intentionality in terms of representation has led to representationalist theories of consciousness. I turn now to considering the relationship between consciousness and intentionality.
Representational Theories of Consciousness While Block’s tripartite definition of consciousness provides a useful classification of different states of consciousness there are those who have denied the existence, or at least importance, of one of the types of conscious states. For example, Dennett has, famously, denied that there is a distinct sense of phenomenal consciousness at all. This is because phenomenal consciousness is really just a form of reflexive consciousness. Higher order thought (HOT) theories of consciousness are based on just such a thought, a state is conscious if we have a HOT about that state. Therefore, for a
mental state M to be conscious there must be a HOT M*, which is about M. For a state to be conscious is for us to be aware of that state. To be consciously aware of pain is to have a thought that one is in pain. Thoughts are all potentially unconscious, but are made conscious by our becoming aware of them, by our having a further thought about them. This view of consciousness is not so very far away from the view presented by Descartes in the meditations (quoted in section titled ‘Introduction: Mind, consciousness, and intentionality’): certainly it seems to me that I see, I hear, and I feel heat. That cannot be false; that is what in me is properly called sensation; and in this precise sense, sensation is nothing but thought.
If I am reflexively aware of a sensation of pain, if I have a thought about it, then it is a conscious thought. There is of course the question of whether animals and infants have the capability to form thoughts involving concepts of their sensory states – what they see, hear, smell, and feel. If they do not possess concepts of red, or pain, then they cannot form HOTs about those sensations and cannot, therefore, have conscious experiences. Rosenthal counters this objection by claiming that the conditions for sensory concept possession are so minimal that most animals and infants are likely to possess them. Nevertheless some critics think that HOT theories involve an unnecessary additional layer of mental complexity to direct phenomenal experience, which is unmotivated. If I am in pain or seeing a red object why do I need to posit a further mental state to make the experience a conscious one? A different approach is to accept that conscious experiences are representational, but not make the further move of modeling all of conscious experience on reflexive consciousness. Dretske disagrees with the HOT theorists on this matter, conscious mental states are not objects that we are conscious of having. Mental states make us conscious of something – my experience of a red balloon makes me conscious of the external object, the balloon. To be consciously aware of the balloon (to experience it), I do not have to be aware that I am conscious of the balloon. Dretske’s approach is to apply his account of representation in terms of function of indication to sensory consciousness.
Intentionality and Consciousness
Dretske claims that all mental facts are representational facts and that experiences do not have phenomenal features of their own. The subjective quality of an experience is just the way the experience represents things to be. Dretske’s representational theory of consciousness is compatible with Brentano’s thesis that intentionality is the mark of the mental, but is incompatible with Brentano’s thesis that the mental and the physical are separate categories, because nothing physical is intentional. However, Dretske’s view is in tension with those views that hold that phenomenology is more fundamental to consciousness than representation.
Is Phenomenal Consciousness Intentional? Phenomenal consciousness is a matter of the way things are experienced by us. To understand this aspect of consciousness philosophers have long posited a category of raw inner feels, or the qualitative character of experiences, that they refer to as qualia. Qualia are supposed to be the properties that constitute the phenomenal character of conscious experience, and what it is like for us to experience them. When you see a red object, or feel a stinging pain in your finger, or taste the smokiness of a heavily peated whisky, these experiences all have a particular phenomenal character, or qualia. Two questions arise at this point: Are phenomenal experiences intentional? Does intentionality itself depend upon the phenomenal character of consciousness? One way of answering both questions is Sellars’s view that sensations and thoughts are distinct. The realm of raw sensory feels is nonintentional and not to be confused with the space of reasons where intentional mental states such as beliefs are governed by norms of rationality. Davidson and Rorty also endorse this separation of the phenomenal realm of sensations and the intentional realm of thought. This position separates the phenomenal from the intentional, the answer to both questions is a resounding no. However, as we have already seen with Dretske’s representational theory of consciousness there is an alternative to the separation move, which is to make all phenomenal experience at root intentional, because phenomenal
427
experiences just are the way that things are represented in consciousness. Horgan and Tienson have recently argued that phenomenal experience is thoroughly intentional. my experience of a red balloon is intentionally directed at the redness of the balloon. The unity of our phenomenal consciousness amounts for them to the what-it’s-like of being in the world. However, they also argue for a thesis of phenomenal intentionality: there is a kind of intentionality that is determined solely by phenomenology alone and not by bodily and sensory contact with the environment. If they are right then externalist theories of content, such as Dretske’s and Millikan’s will turn out to be mistaken for these cases of phenomenal intentionality. Therefore, it would show that the strategy of explaining consciousness by giving a naturalistic account of intentionality was doomed to failure, because some of our phenomenology could not be explained in terms of representational content. One might accept that there is a category of phenomenal intentionality where phenomenal experiences are intentionally directed, while rejecting the standard view of phenomenal experiences as being internal ‘raw feels’ or that the phenomenal character of experiences is independent of the way the world is (see the section titled ‘Embodied intentionality’). This would be to reject Horgan and Tienson’s position that phenomenology, and hence phenomenal intentionality, are independent of the world (what exists beyond the skin of the individual). Horgan and Tienson’s position depends upon thought experiments such as brains in vats or deception by an evil demon. Normally our phenomenal experience is dependent upon our sensory modalities and our bodily progress around an environment. However, Horgan and Tienson claim that if you were a brain in a vat, with no sensory or bodily contact with the world, you could have the same phenomenal experiences as if you were really embodied. To argue for the narrowness of phenomenal experience and intentionality on the grounds that such scenarios are imaginable is empirically weak. Our phenomenology is due to our bodily contact with the world and it is in these terms that we will construct empirical theories. We shall not construct such theories by considering imaginable scenarios such as brains in vats. However that may be, Horgan and Tienson have
428
Intentionality and Consciousness
drawn our attention to the close relationship between phenomenal experience and intentionality. These issues are taken up further in the final two sections entitled ‘The phenomenological tradition’ and ‘Embodied intentionality.’ The Phenomenological Tradition Husserl followed Brentano in attempting to analyze conscious thought in terms of intentionality. In his earlier writings (The Logical Investigations) Husserl rejected Brentano’s notion of intentional (or mental) inexistence – the object of thought is always internal (immanent to) the mind. By contrast Husserl thought that the objects of intentional states such as beliefs and desires, often transcend the mind, in the sense that they exist externally to it. However, Husserl also proposed a ‘phenomenological reduction’ of mental phenomena, a method for investigating consciousness that ‘brackets off ’ judgments about the ontological existence or inexistence of the world. Phenomenological investigations were, thereby, restricted to understanding the structure of mental acts. The exact interpretation of the phenomenological reduction is a matter of dispute, and is contested by Heidegger who views intentionality as the ontological structure of ‘being-in-the-world,’ and by Merleau-Ponty, who emphasizes the embodied nature of intentionality. MerleauPonty rejects the idea that there is a purely mental experiencing subject that can be considered apart from the body and the rest of the world. Rather, a conscious subject already presupposes a bodily and temporal existence situated in an environment. The conscious subject can only be understood in so far as he is a bodily subject of experience situated in and related to a preexisting environment. Consequently, for Merleau-Ponty, there can be no ‘bracketing off ’ of the world when performing phenomenological investigations of the intentional structure of consciousness. For Merleau-Ponty an embodied subject is actionoriented to the world and has an intentional relation to that world, even before it starts to reflect on it. We can, therefore, derive from MerleauPonty a kind of prereflective embodied intentionality, one that does not yet involve conceptual content.
Embodied Intentionality The move away from Brentano’s account of intentionality in terms of intentional or mental inexistence is completed in the emerging approach to the mind and cognition, variously labeled: the embodied mind; the extended mind; distributed cognition; and cognitive integration. We saw how an externalist account of intentionality rejected the idea that intentional relations were understandable purely in terms of a pure conscious/ mental subject, intentionally related to mental/ intentional objects wholly encapsulated by the mind. Furthermore, Merleau-Ponty gave us the notion of a prereflective conscious subject as essentially embodied and situated in an environment. From this starting point the conscious subject and intentionality look very different from the Cartesian/Brentanian starting point in which intentionality is an entirely internal affair. Combining a Merleau-Pontian position on embodied intentionality and an externalist position on representation would give us a relationship between intentionality and consciousness that would deal with Brentano’s thesis and relieve the tension between the phenomenological and intentional aspects of consciousness. We would begin by thinking of our primary conscious engagement with the world asanintentionallydirectedbodilyengagement,which also has a phenomenal character in Merleau-Ponty’s prereflective sense. We would then add a layer of cognitive sophistication involving contentful (hence intentionally directed) mental states such as beliefs anddesires. Thisfusionofembodiedphenomenology and externalist theories of content is a promising line of inquiry into the relationship between intentionality and consciousness. See also: Brain Basis of Voluntary Control; Concepts and Definitions of Consciousness; Free Will; Language and Consciousness; Mental Representation and Consciousness; Phenomenology of Consciousness.
Suggested Readings Block N, Flanagan O, and Gu¨zeldere G (eds.) (1997) The Nature of Consciousness. Massachusetts: MIT Press. Brentano F (1874) Psychology from an Empirical Standpoint, Rancurello AC, Terrell DB, and McAlister L (trans.).
Intentionality and Consciousness London: Routledge, 1973 [2nd edn., intr. by Peter Simons, 1995]. Crane T (2001) Elements of Mind: An Introduction to the Philosophy of Mind. Oxford: Oxford University Press. Dennett D (1990) Consciousness Explained. London: Penguin. Dretske F (1988) Explaining Behavior: Reasons in a World of Causes. Massachusetts: MIT Press. Dretske F (1997) Naturalizing the Mind. Massachusetts: MIT Press. Gallagher S (2005) How the Body Shapes the Mind. Oxford: Oxford University Press. Gallagher S and Zahavi D (2008) The Phenomenological Mind: An Introduction to the Philosophy of Mind and Cognitive Science. Oxford: Routledge. Husserl E (1900/1) Logical Investigations, Findlay JN (trans.). London: Routledge.
429
Menary R (2007) Cognitive Integration: Mind and Cognition Unbounded. Basingstoke: Palgrave Macmillan. Merleau-Ponty M (1945) Phenomenology of Perception, Smith C (trans.). New York: Humanities Press, 1962 and London: Routledge & Kegan Paul, 1962 [translation revised by Forrest Williams, 1981; reprinted, 2002]. Millikan R (1993) White Queen Psychology and Other Essays For Alice. Bradford Books/MIT Press. Moore AW (1993) Meaning and Reference. Oxford: Oxford University Press. (Collects papers by Frege and Russell with later developments in the field). Rosenthal D (2005) Consciousness and Mind. Oxford: Clarendon press. Tye M (1997) Ten Problems of Consciousness. Massachusetts: MIT Press.
Biographical Sketch
Richard Menary read for a BA in philosophy at the University of Ulster, an MSc in cognitive science at the University of Birmingham, and then a PhD in philosophy at King’s College London. He has been a senior lecturer at the University of Hertfordshire in the UK and is currently a senior lecturer at the University of Wollongong in Australia. Richard has written articles on cognition and consciousness for journals such as Philosophical Psychology, the Journal of Consciousness Studies, and Language Sciences. He has also edited two books Radical Enactivism and The Extended Mind, and is the author of Cognitive Integration. He is currently writing another book on the philosophy of cognition.
Intuition, Creativity, and Unconscious Aspects of Problem Solving J I Fleck, The Richard Stockton College of New Jersey, Pomona, NJ, USA J Kounios, Drexel University, Philadelphia, PA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Impasse – A mental sticking point, usually during problem solving, after which no further mental progress seems possible. Incubation – A period of time away from an unsolvable problem when the individual makes progress on the problem, likely via unconscious processes, while consciously engaged in other thoughts. Insight – A moment of sudden understanding that accompanies the solution of a problem. This understanding is usually preceded by misdirected thought. Metacognition – The understanding or impression of one’s own thought processes. Restructuring – A change in one’s representation or understanding of a problem. Most commonly associated with the insight experience, restructuring is thought to occur after initial misdirection. Semantic priming – Responses to stimuli are facilitated when items similar in meaning have been previously presented. Synchrony – Large clusters of neurons begin firing together in the completion of a particular task. This synchrony can appear as greater EEG power.
Introduction Creativity, insight, and intuition are just some forms of thought believed to rely heavily on the unconscious, giving these areas special status in psychology. Growing interest in creativity and problem solving has led to an influx of novel approaches in the empirical exploration of topics previously believed to be beyond our reach.
This article reviews existing research and theory concerning the cognitive, behavioral, and neural components of the unconscious in creativity and problem solving. Of importance is understanding the field’s past and conceptualizing its future.
An Historical Account When Archimedes first stepped in the public baths of Syracuse he had no conscious awareness of what was to come, he merely intended to take a bath. But there came the answer to a problem which to that point had baffled him. Archimedes, according to written accounts, was attempting to determine if the crown of Hiero was made of pure gold, or if the crown instead contained a cheaper silver alloy. Though the finished crown was identical in weight to the gold initially provided for the crown’s construction, it had been suggested to Hiero that the crown’s constructor had tricked him. Archimedes was puzzled as to how to determine if the suspicion was true. We cannot be sure of the mechanisms that led to Archimedes’ Eureka moment – an eureka moment is one in which a solution to a problem that previously eluded the individual suddenly pops into mind, often so suddenly that it surprises the person resulting in an emotional reaction. For Archimedes, water splashing over the edge as he stepped into the bath facilitated the sudden and unexpected solution to his problem – the crown and an equal weight of pure gold should exhibit the same volume and, therefore, should displace the same amount of liquid if submerged in water. If the crown contained some amount of silver, it would be larger in volume (because silver is less dense than gold) and would thus displace more water than an equal weight of pure gold. Archimedes is said to have found just that – the crown 431
432
Intuition, Creativity, and Unconscious Aspects of Problem Solving
displaced more water than the pure gold providing proof that Hiero had indeed been duped. The experience portrayed above of Archimedes’s Eureka moment is similar to popular anecdotes of spontaneous discovery relayed in the creativity literature that detail the innovations of Newton, Poincare´, Kekule´, and others. Though what occurs in the minds of persons who produce such discoveries is often unclear, the phenomenological experiences are more widely recognized. Many researchers and laypersons alike have held the belief that creativity and related processes (e.g., intuition and insight) must differ in some fundamental way from other forms of thought (e.g., memory and logic) partly because they are associated with different experiences. For the individual, creative insights seem to come when least expected – after an unsuccessful and at times lengthy period of directed cognition, the insight or answer to one’s problem appears in what seems a spontaneous manner when the person is otherwise engaged. Experimental definitions of insight are diverse; some of the most common include the moment of understanding after a period of mental confusion, generating a solution outside awareness, and generating a solution by overcoming a mental sticking point after changing one’s thinking about the problem at hand. Historically, the mystery surrounding creative insights was so great that creativity was believed to be a gift from the gods. How else could such unprompted events be explained? Graham Wallas first drew our attention to the complex interplay between conscious and unconscious forms of thought as it pertains to creativity. Wallas was an English psychologist and political scientist whose book The Art of Thought, published in 1926, significantly impacted how researchers would study creative thought and its unconscious components. Wallas suggested that creativity was composed of four stages: preparation, a period of directed knowledge acquisition; incubation, a period of unconscious task-related activity occurring while the conscious mind is otherwise engaged; illumination, the moment of insight when the answer emerges into conscious awareness; and finally, verification, when the now-conscious insight is subjected to a period of directed fact checking. Of these stages, incubation and
subsequent illumination have held the most potential and interest for researchers attempting to reveal the thoughts responsible for creative insights. Almost a century of research and theory in this area has provided merely a glimpse into the cognitive and neural components – conscious and unconscious – preceding that final Eureka or Aha! moment. In this article, we address several themes surrounding the study of the unconscious in creativity, problem solving, and related processes. First, we present existing data on the role of the unconscious in problem solving, specifically problem solving by insight, believed by many to elicit unique and unconscious components of thought. Second, we explore the process of incubation, the stage of creative thought proposed by Wallas with the most significant tie to the unconscious. Third, we consider contributions to our understanding of the unconscious in problem solving and creativity that have emerged from advances in neuroscience. Finally, we present a glimpse of the up-andcoming topics sure to influence our future conceptualization of the unconscious in problem solving, creativity, and related processes.
The Study of Creative Insights A Search for Definitions There are few fields in psychology that have been more stymied by the discrepancy over a definition than the field of creativity. A general definition, though unsatisfying to many, is that creativity is a process that results in the generation of something that is both novel and useful (the terms novel and useful have also been much debated). A series of factors have been proposed as part of a complex state preparing the individual for the experience of creative insight; personality characteristics, environment, knowledge acquisition, intelligence, and motivation are just some of these factors. Even mental illness (e.g., bipolar disorder and schizophrenia) has been associated with enhanced creative potential. Added to the mix is the potential relevance of the magnitude of the creative product as well as the eminence of its creator. Should our focus be the creative
Intuition, Creativity, and Unconscious Aspects of Problem Solving
process in eminent individuals, in normal adults, or both; and, how do the creative processes in these two groups relate to each other? Significant in all fields but perhaps more relevant to the study of creativity, is how researchers are able to manufacture experimental situations in a controlled environment that are valid reflections of the natural creative process. Finally, and of greatest relevance here, are the mechanisms that lead to creative thought the result of conscious or unconscious processes? Studying the Unconscious through Insight Insights, most often explored in conjunction with problem solving, are analogous to the illumination stage of Wallas’ creativity stages. Insight problem solving has been used by researchers to explore creative thought in a controlled laboratory environment; however, little research exists that directly explores the relation between insight and creativity. As with illumination, there is strong debate about the components (conscious or unconscious) that precede the insight experience. Insight has been described as the end product of a process in which the individual is initially misdirected by the problem components, with this misdirection leading the individual to a mental sticking point or impasse. Further progress is only possible when the individual experiences a sudden restructuring or change in mental representation of the problem – in fact, many argue that restructuring leads the individual directly to the problem’s solution. The final stage of the process, after the solution enters awareness, is often met by an Aha! experience that occurs when the verification process reveals that the solution in hindsight could have been easily and directly obtained had the solver correctly applied the problem information. Metacognitively, insights have been described as the proverbial lightbulb suddenly going on in one’s head. The Gestalt psychologists initiated the study of insight as an alternative to the trial-anderror approach to learning proposed by Edward Thorndike. While Thorndike described learning as the gradual acquisition or building of knowledge, premier psychologists of the time including Karl Duncker and Wolfgang Ko¨hler suggested
433
that more intelligent forms of thought such as insight were possible. The school of Gestalt psychology originating in early nineteenth century Germany is best known for its theories of perceptual organization, most popular the tendency of humans to organize the visual field as a whole rather than by its constituent parts. For the Gestalt psychologists, the change that led to the insight experience was sudden in nature and was preceded by the misapplication of prior knowledge in situations where the individual’s existing knowledge base was not relevant – in other words, the conscious application of information obtained via long-term memory search impeded progress. Examples of this misapplication or fixation are apparent in two classic insight problems from the time: the candle problem and the nine-dot problem (see Figure 1). Classic insight problems are those originally used in the time-period of the Gestalt psychologists but have remained key problem stimuli in more recent studies of creativity and insight. In the candle problem, it is common for problem solvers to become fixated on the tack box as serving the function of holding the tacks (i.e., box as container) and for individuals not to extend the box to other roles such as a shelf or ledge to support the candle. In the nine-dot problem, perceptual fixation stemming from the inferred perimeter formed by the square configuration of the nine dots reduces the likelihood that participants will extend lines beyond the square when attempting the problem’s solution. Preliminary research into the mechanisms of insight problem solving originally focused on if, and in what manner, solutions to problems solved via insight were somehow the result of processes that were comparable to the processes supporting other forms of problem solving in which strategies were methodically and consciously applied by problem solvers (e.g., analysis-based problem solving). Though many researchers argued for unique cognitive processes in insight, others supported the idea that little more than speed of processing differentiated insight from analysis-based problem solving. Research conducted in the late 1980s and throughout the 1990s revealed several possible differences between the two problem types, especially with regard to the role of conscious processing.
434
Intuition, Creativity, and Unconscious Aspects of Problem Solving
Tacks
(a)
(b)
(c) Figure 1 Classic insight problems used in the study of insight and creativity. (a) The candle problem. Using the items provided, attach the candle to a door so it can burn properly. Reprinted from Cognitive Psychology, Vol 4, Robert W. Weisberg and Jerry M. Suls, An information-processing model of Duncker’s candle problem, pp. 255–276, Copyright (1973), with permission from Elsevier. (b) The nine-dot problem. Without lifting your pencil, connect the nine dots using four straight lines. (c) The two-string problem. Two strings are hanging from the ceiling. The distance between the strings is such that you cannot reach one of the strings while holding on to the other. Using the items provided, tie the two strings together. Reprinted from Cognitive Psychology and Its Implication, J. R. Anderson, p. 247, Copyright (1980), with permission from W. H. Freeman. Problem solutions are available in Figure 4.
First, it appeared to be the case that the metacognitive awareness that existed during analysisbased problem solving was lacking in insight. In an interesting series of studies, Janet Metcalfe and colleagues explored participants’ awareness
of their problem-solving potential and their online perceptions of their solving progress for a set of insight problems and a set of noninsight problems. After participants’ brief viewing of the to-be-solved problems, they were unsuccessful in predicting their abilities to generate solutions to insight problems, although they were quite successful in anticipating their likelihood of generating the correct responses to a series of trivia questions. Differences also emerged during the solving process itself. During analysis-based problem solving, participants correctly perceived and reported that they were inching ever closer to the problems’ solutions. Not so with insightful problem solving; solvers instead were unclear as to their progress until the interval immediately preceding the solution. Comparable findings have also been observed in physiological recordings of heart rate during problem solving, with a gradual increase in heart rate accompanying the incremental solution of analysis-based problems and a sudden increase in heart rate immediately prior to the solution of insight problems. Also of interest, it did not appear that forcing thoughts linked to problem solving into consciousness was advantageous when attempting insight problems. Instructing participants to generate verbal protocols (i.e., to verbalize their thoughts during problem solving by thinking out loud) as a window into their thought processes negatively influenced participants’ solving performance for insight problems but had little effect, or in some cases facilitating effects, for participants providing verbal protocols while attempting to solve analysis-based problems. It should be noted that the interfering effects of concurrent verbalization on the solution of insight problems has been viewed as tenuous with subsequent studies having difficulty replicating these effects. Researchers have extensively examined the restructuring component of insight, thought to result in the sudden awareness or illumination of the problem’s solution. Explorations of the nature of restructuring have been limited primarily to behavioral approaches and have focused on the potential occurrence and characteristics of restructuring itself rather than the factors that facilitate the restructuring. Several classification systems have been proposed detailing the possible
Intuition, Creativity, and Unconscious Aspects of Problem Solving
mechanisms through which restructuring is possible, the most notable of which was proposed by Stellan Ohlsson. According to Ohlsson, forms of restructuring include, but are not limited to, a change in one’s representation of the problem state (the initial problem configuration) or the goal state (the desired ending effect), a change in the operators or problem components available for application, the decomposition of problem components, and perhaps receiving most attention, the relaxation of previously imposed problem constraints (the problem solver’s beliefs about how different components relate to each other or the mechanisms acceptable in the problem’s solution). A change in any one of these things could lead to a problem’s solution, though it has also been suggested that multiple restructurings may be needed to solve a problem depending on problem difficulty. Behavioral studies have been divided on the role of the unconscious in restructuring, with researchers suggesting that both conscious and unconscious components can elicit restructuring and subsequently insight. Proponents of the conscious view suggest that long-term memory search and retrieval initiated by bottom-up (external) or top-down (internal) sources can lead to restructuring experiences. During the problem-solving experience when impasse is reached, the solver consciously begins to search memory for relevant information. This conscious application of new information is only possible, however, if the individual possesses in memory relevant knowledge for the problem’s solution. Some support for this possibility has been found in research employing classic insight problems such as the nine-dot problem. Researchers observed limited restructuring in participants attempting to solve the nine-dot problem even when solution hints, such as the instruction to extend solution lines outside the dots’ configuration, were provided. Solution rates improved, however, when participants were provided with training in activities such as line extension prior to attempting the nine-dot problem. The beneficial effects of prior training suggest that relevant knowledge must be in place prior to problem solving and recent enough in memory to be retrievable if it is to be applied during problem solving.
435
In contrast, others have proposed that restructuring occurs via the unconscious spreading of activation stemming from problem components. Spreading of activation occurs when a node or representation is activated in the brain, resulting in the spread of activation to related nodes or concepts. Support for this theory appeared first in research reporting the detrimental effects on insightful problem solving when individuals’ thought processes are forced into consciousness. As noted above, some researchers have observed that verbalization resulted in a significant decrease in solution rates for insight problems when compared to participants attempting the same problems but during silent conditions. These results led the researchers to suggest that certain components needed for insight could not be verbalized. Some of the most recent research in the debate on the conscious versus unconscious nature of restructuring has explored the role of working memory in the solution of insight problems. Working memory is most often defined as the online storage and processing, including directed attention, of information during conscious thought. In order to specifically isolate the restructuring component of insight, researchers manipulated the amount of processing required in the initial phases of problem solving. With many problems, solvers must explore a number of possible solution approaches before reaching the impasse state thought to precede restructuring. When participants attempted to solve problems that isolated the restructuring component of insight (i.e., the initial solving space was small so that participants quickly exhausted their initial ideas) working memory was not a significant predictor of success. In contrast, when participants were presented with problems that included a much larger initial solving space (i.e., multiple solution paths requiring consideration prior to restructuring), working memory was a significant predictor of solving success. Similar research exploring the role of working memory in the solution of insight versus analysis-based problems revealed that working memory was a significant predictor of success for analysis-based problems, whereas short-term memory capacity (storage only) was a significant predictor of success on insight problems. Together, this line of research suggests that the directed
436
Intuition, Creativity, and Unconscious Aspects of Problem Solving
attention and processing present in working memory may only serve a role in the initial stages of insight problem solving when the individual is attempting to rule out possible paths to the solution; of course, the solver must have the impression that all reasonable possibilities have been attempted in order to achieve impasse. After this mental sticking point is achieved, working memory is not a factor that influences whether or not the individual will experience a restructuring of the problem’s components – that instead may be directed by processes outside awareness. Incubation – The Unconscious at Work? Though the unconscious has been the cornerstone of many theories of creativity and insight problem solving, evidence of its influence has been difficult to obtain. One of the most fundamental questions that researchers faced was whether incubation, the unconscious period in Wallas’ stages of creativity, was passive, advancing thought as a mere byproduct of the passage of time or was an active process requiring additional mental processing such as the search of long-term memory or the integration of new information from the environment. Passive theories suggest that time away from the primary task allows original memory traces to weaken in favor of more fruitful ideas. Empirical support has been found for this theory through the early work of Steven Smith and Steven Blankenship who demonstrated that time away from a problemsolving task resulted in higher solution rates when participants later returned to the task of interest. Solving rates during this second solving window after a break were greater than those achieved when the second problem-solving window followed immediately after the initial solving window. This effect was later extended in research demonstrating that both short and long breaks were equally effective in increasing problemsolving success. A similar theory focusing instead on cortical arousal suggests that a decrease in cortical arousal during incubation could effectively alleviate mental sticking points by generating an environment where more diffuse attention is possible. Focused attention could cause problem solvers to become fixated on their initial solution ideas, therefore
resulting in impasse. Undirected cognition or cognition directed toward the completion of another task would cause a decrease in arousal, thereby reducing the focus on one’s conscious solution ideas allowing novel solution possibilities, perhaps outside awareness, to be considered. Alternative theories of incubation suggest more active forms of processing. Most active theories to date imply that the application of external hints could be sufficient to facilitate novel solution ideas. One such idea is the opportunistic assimilation hypothesis proposed by Colleen Seifert, David E. Meyer, and colleagues. Opportunistic assimilation proposes that when an impasse or sticking point has been reached, a mental flag or failure index is linked in the brain to existing task-related information. As the individual then moves through his or her daily events, external stimuli encountered initiate new activation that spreads to associated areas in the brain. New solution ideas may enter into consciousness if this novel activation reaches the stored failure indices, indicating the relevance of the external stimuli in the problem’s solution. Research evaluating the role of external stimuli in incubation has generated mixed results. Some of the first research exploring the value of external hints in the unconscious facilitation of problem solving was conducted by Norman Maier, using the two-string problem (see Figure 1). During the solving period, a confederate entered the room where the participant was working and, within the participant’s view and in an apparently unintentional manner, brushed one of the strings causing it to swing from side to side. The participant group exposed to the confederate exhibited higher solution rates than those attempting to solve the problem without the confederate’s presence, though most participants did not report the actions of the confederate as influencing their solution ideas. In this instance, external cues influenced unconscious but not conscious thought. Although participants in Maier’s research were able to apply the external hints in generating the solution, other researchers have had less success in replicating this effect. For example, participants were not successful in applying solution hints to problems from Sarnoff Mednick’s remote associates task (RAT), a popular test of creativity
Intuition, Creativity, and Unconscious Aspects of Problem Solving
developed in the 1960s. In the RAT, participants are presented with three words and must identify a fourth word that can link all three (e.g., problem words: letter, puppy, true – solution: love). During the incubation phase, participants were asked to perform a word generation task in which they needed to recombine the letters of a stimulus word to form new words; the stimuli for the task contained the solution words for the RAT problems. Participants were not able to apply these hints to the RAT problems during the second solving period unless they were explicitly told of the hints’ value in solving the problems. They instead seemed to compartmentalize the initial problem-solving phase and the word-generation task or incubation phase that followed. Recent research has suggested that a key factor in the effectiveness of the incubation period is the solver’s expertise in the topic area of the problem. For example, experts were less likely to solve misleading problems during the initial solving period than novices presumably due to fixation resulting from their application of irrelevant prior knowledge; however, after an incubation period, they exhibited a reduction in this fixation and improved solution success. Interestingly, an incubation period did not facilitate the solution of problems for which their expertise was relevant or when they were asked to solve neutral problems outside their areas of expertise. Further, no incubation effects were present in novices who were not experts in fields related to the problems and, therefore, were not fixated during the initial problem presentation. These results suggest that mental fixation by the solver during the initial solving phase may be a necessary prerequisite for the occurrence of the incubation effect. Evidence for an Active Unconscious Some of the first behavioral evidence for an active unconscious came from research conducted by Kenneth Bowers and colleagues. In an interesting series of studies, researchers were able to demonstrate that lack of awareness of the solution on the solver’s part did not coincide with lack of processing. Bowers presented participants with two three-word sets in a task he called The Dyads of Triads Task. In one of the sets, all three words were associates of a fourth (not presented) target word.
A Playing Credit Report
B Still Pages Music
437
A Card
Figure 2 Sample dyads of triads task. Only triad A is coherent. Each word in A can be combined with the target word card. Reprinted from Cognitive Psychology, Vol 22, Kenneth S. Bowers, Glenn Regehr, Claude Balthazard, and Kevin Parker, Intuition in the context of discovery, pp. 72–110, Copyright (1990), with permission from Elsevier.
In the other set, only two of the words were associates of the same word; the third was not (see Figure 2). When possible, participants were asked to provide the target word for the coherent triad. If a solution was not possible, they were then asked to choose which of the two sets was the coherent triad and to rate their confidence in this decision. For unsolved items participants were able to choose the coherent dyad at rates above chance. Even more interesting was the observation that participants demonstrated increased accuracy during the forced-choice decision when the solution word for the coherent dyad performed a semantically similar role for all three words than in instances where the dyad words were all linked to the same solution word, but that word carried different meanings as an associate. These and other results led Bowers and associates to suggest that the summation of activation to the target word on the unconscious level led to the correct choice for solvers during the forced-choice decision even though this information was not yet available in consciousness.
Neural Underpinnings of Creative Thought The Creative Right Hemisphere Prior to the use of modern imaging techniques such as functional magnetic resonance imaging (fMRI) and high-density electroencephalography (EEG) which will be discussed below, much was learned about the brain’s involvement in creative thought through research using priming paradigms previously developed for language research. In priming, judgments regarding a target word are faster if that target word is preceded by a
438
Intuition, Creativity, and Unconscious Aspects of Problem Solving
semantically related source word (i.e., the prime) than when judgments are made about the same target word without the preceding prime (e.g., people are faster in making judgments about the target word dog when it is preceded by the prime word cat). In semantic priming tasks, primes presented to the right visual field (i.e., going to the left hemisphere) have been associated with priming effects for primary word meanings and direct semantic associates. Primes presented in the left visual field (going to the right hemisphere) are instead associated with priming effects for more diffuse semantic connections, including secondary word meanings and remotes associates. It has been further suggested that the structure of the neural networks in the hemispheres may be uniquely constructed to support these two forms of priming. Researchers utilizing priming paradigms have observed that priming effects in the right hemisphere are linked to the creative solution of compound remote associate (CRA) problems. CRAs, similar in nature to the problems from Mednick’s RAT, are three-word problems in which a solution word can be combined with each of the problem words to generate a compound word form (e.g., problem: pine, crab, sauce – solution: apple – pineapple, crabapple, and applesauce). Participants were given 15 s to generate the solution for the three problem words presented in the center of the computer screen. If participants were unsuccessful in generating the problem’s solution, then a target word was briefly presented (180 ms) in either the right visual field or the left visual field; this target word was either the problem’s solution or an unrelated word. Participants were required to determine as quickly as possible if the target word was the correct or incorrect solution for the problem. After making their solution/nonsolution judgment, participants were asked to report the nature of the thought processes surrounding the judgment; was their choice the result of a deliberate, conscious strategy (low-insight/low-creativity) or instead made without conscious awareness of the processes behind their decision (high-insight/high-creativity). A right-hemisphere priming effect (i.e., faster judgments) occurred in the solution/nonsolution identification task for decisions that were reported by participants as high insight. Similar effects have
been observed in research using classic insight problems. When participants were initially unsuccessful in generating the solutions to classic insight problems, solution hints presented in the left visual field facilitated the creative solution of unsolved problems. Together these results suggest that the activation of right-hemisphere semantic networks may be more likely to produce an environment where the sudden solution of verbal problems outside awareness is possible. While the above priming research began to suggest hemispheric differences between creative and noncreative thought, most behavioral research was limited in its ability to definitively demonstrate that creative thought was reliant upon unique, perhaps unconscious, mental processes. At the rise of the cognitive neuroscience revolution in the early 1990s, many researchers held the theory that creativity stemmed from unique processes but had no way to substantiate that claim. Supporters of the business-as-usual approach to insight and creativity were also restricted in their ability to demonstrate that the processes in creative and noncreative thought did not differ. Creativity and the Unconscious – The Right Hemisphere and Beyond Since expanding to a more diverse set of tasks and techniques, researchers have suggested that a variety of different regions of the brain support creativity and related unconscious processes; the frontal lobes, the anterior cingulate, posterior regions of the cortex, and the cerebellum are just some of these regions. A number of researchers have used EEG to examine neural differences between convergent and divergent thought. With EEG, electrodes placed on the scalp record changes in electrical brain activity for significant numbers of neurons. These changes stem from large numbers of neurons firing in synchrony reflecting mental processing in the brain. Divergent thinking has traditionally been viewed as a form of creative thought. Typical divergent thinking tasks are similar in nature to the tasks originally proposed by Joy Paul Guilford. Guilford’s assessments of divergent thinking required participants to generate as many unusual uses for
Intuition, Creativity, and Unconscious Aspects of Problem Solving
common objects (e.g., shoe, matchstick) as possible. Responses were scored based on the number of responses generated (fluency), and their originality from other responses from the same individual (flexibility), as well as from the responses of other participants in the study (originality). In contrast, convergent thinking tasks often included activities such a mental arithmetic, the completion of verbal math problems, and sequence completion tasks, all tasks requiring a specific response. EEGs recorded from participants completing divergent and convergent tasks revealed greater EEG complexity over frontal, central, and parietal regions when participants were engaged in divergent thought. Greater EEG complexity often suggests that there is less synchrony in activity between different regions of the cortex or that multiple neural assemblies (large groups of neurons) are necessary to support the task of interest. Interestingly, when participants were subdivided into high and low performance groups based upon divergent thinking ability, high performers exhibited less complexity over frontal/central regions than did low performers. A similar effect was observed when participants were provided with training in divergent thinking (nine 30 min sessions). In fact, following training, participants exhibited greater low-alpha power (an EEG frequency indicating cortical inactivity) over frontal regions when compared to the control group. Together these results suggest that as individuals are more skilled in divergent thinking, fewer mental resources, especially in frontal regions of the brain, are needed to search for relevant strategies to apply during divergent-thinking tasks. Researchers exploring alpha power throughout the cortex during divergent thinking have found general support for theories proposing that reduced cortical arousal facilitates creative thought. In general, greater alpha power has been observed over central and parietal regions during divergent thinking; this increase was more enhanced during segments of the EEG associated with ideas rated by participants and external raters as highly original. Patterns of alpha activity observed during creative thought generally support theories highlighting the critical role of decreased cortical arousal in the generation of novel ideas. Therefore, it would appear that high arousal strengthens prepotent
439
(initial) response ideas in consciousness, whereas lower arousal adjusts the playing field of ideas allowing ideas in the unconscious to enter consciousness. A common component of several theories of creative thought is that the individual must be able to generate a substantial number of solution ideas linked to the task at hand – some have proposed that posterior regions of the cortex are responsible for this prolific generation. Enhanced alpha power over posterior regions of the cortex may reflect a decrease in conscious processing in the region, allowing the combination of problem components to occur on the unconscious level. After the unconscious combination of problem components a solution or idea could then be selected from among this collection and subsequently implemented. New Tasks, New Methods: Exploring the Moment of Insight One of the most critical facets in understanding the unconscious and its role in creativity is understanding the moment of insight – when the work of the unconscious enters awareness. Before considering research in this area, it is important to note that by no means is there universal agreement regarding the neural substrates of unconscious processing supporting creativity and problem solving. For example, some researchers have proposed that the basic process of plasticity (neural changes occurring in response to learning or experience) may be enough to produce the unconscious incubation effects that lead to the moment of insight. In these theories, the strong activation of problem components facilitates additional spreading of activation and modification of connections on the unconscious level while the problem solver is otherwise engaged. Then, when returning to the original problem, the new neural networks constructed via the brain’s own plasticity may be more effective than during the initial attempt. Studying the exact moment of insight, when unconscious thought becomes conscious, has proved difficult. One key problem noted above is that unconscious processes are not directly reportable by participants reducing the utility of verbal and behavioral measures of thought. Further, the classic problems that shaped the initial study of creativity,
440
Intuition, Creativity, and Unconscious Aspects of Problem Solving
chiefly the comparison of solutions in insight versus analysis, posed several problems. First, there was concern that problems labeled as insight-based and those as analysis-based may differ in ways other than their mechanisms of solution: concerns over differences in task difficulty, the ambiguity of the goal state in insight versus analytic problems, and the general lack of control that stems from using different problems are just some of the criticisms lodged. An even greater concern, however, has been the reliance by researchers on the assumption that insight problems are indeed solved via insight and that analytic problems are solved independent of insight. In response to these criticisms new stimuli have been introduced in the study of creativity and unconscious problem solving. A change in task stimuli was also a requirement for neuroscience applications, where a substantial number of standardized trials are necessary in order to identify activation patterns or components unique to creativity. We will explore just some of the tasks that have gained popularity. In recent years, Jing Juo, Xiao-Qin Mai, and colleagues have explored neural correlates associated with the solution of riddles and brain teasers using fMRI and EEG techniques (e.g., ‘‘The thing that can move heavy logs, but cannot move a small nail’’ Luo and Knoblich, 2007, p. 79 – solution – ‘‘a river’’). The use of these techniques together can provide the researcher with a more complete picture of the neural underpinnings of creativity and related processes; fMRI provides spatial data indicating where in the brain differences are present, whereas EEG provides temporal data signifying when differences are occurring. To compare brain mechanisms associated with creative versus noncreative processing, Luo manipulated the type of hint presented to participants for unsolved riddles; hints were either those intended to elicit restructuring or those that added more detail to the problem without changing solvers’ mental representations. Unique activation following restructuring cues was observed in the anterior cingulate, a region previously linked to processing cognitive and emotional conflicts. Though noteworthy, the above research could be criticized for its reliance on the assumption that insight processes (e.g., restructurings) occur following the presentation of restructuring hints
and not following the other hint types. To address this concern, Mai, Luo, and colleagues presented participants with verbal riddles, half of which were difficult to solve increasing the likelihood of impasse and subsequent Aha experiences, and half which were easy by comparison and not likely to elicit Aha experiences. It should be noted that riddles of both difficulty types could be solved with or without insight. During the 8 s solving interval, participants were to generate a solution if possible. Two seconds after the solving period elapsed, the solution to the riddle was presented. Participants then self-reported whether the solution presented matched the one they had generated during the solving interval (no-Aha trials) or if the solution, while clearly correct, was different from the solution they had generated or made sense even though they were unable to generate a solution (Aha trials). It should be noted, however, that many researchers question the classification methods used in the Aha/No-Aha categorizations noted above. For example, it seems possible that a participant’s idea that matches the presented solution could have been generated via insight; it also seems possible that a nonmatching idea generated during the solving window could have been the result of an analysis-based strategy. The researchers’ analysis of segments of interest recorded using high-density EEGs during the problem-solving task revealed early visual eventrelated potential (ERP) components for both Aha and no-Aha trials. However, a later negative component unique to Aha trials was observed over midline electrodes. Subsequent analysis localized the source of this component to regions in and around the anterior cingulate. These data corroborate the results of their prior fMRI study in identifying the anterior cingulate as a structure linked to the restructuring component of insight. If we can accept the assumption that restructuring hints facilitated insight/creative solutions, the question becomes how activation in the anterior cingulate is related to processing during these experiences. Because the anterior cingulate is involved in conflict processing, it may be that activation following restructuring hints stems from the individual considering multiple solution ideas – the initial prepotent response, as well as other ideas that are being considered but are still at the unconscious level.
Intuition, Creativity, and Unconscious Aspects of Problem Solving
A second set of experiments conducted by Mark Jung-Beeman, John Kounios, and colleagues has been instrumental in expanding our understanding of differences in the brain during creativity and related processes. These researchers have used fMRI and high-density EEG to isolate neural differences when participants solve CRA problems (described above) using different strategies. Participants were asked to attempt 180 CRA problems, and were given up to 30 s to generate each problem’s solution. If successful, participants were asked to report the type of strategy they used in determining the problem’s solution, either insight or noninsight. A noninsight solution was one in which a participant could identify the strategy used in generating the problem’s solution – a conscious application. In contrast, an insight solution was one in which the solution suddenly popped into the participant’s awareness without conscious access to the strategy yielding the solution. Both spatial and temporal information distinguishing the two solution types was obtained. Results of the analysis of fMRI data for the brief time interval immediately prior to a solution response revealed a significant difference in right-hemisphere brain activity for insight and noninsight problem types – greater activity was present in the anterior superior temporal gyrus (aSTG) for CRA problems solved via insight than for those solved via noninsight strategies. EEG data, more sensitive to temporal information, were highly valuable in this paradigm due to the sudden nature of insight solutions. Isolation of the window immediately preceding the solution response corroborated and extended the fMRI results. A surge of high-frequency (gamma-band) activity was observed over a region of the STG in the right hemisphere for insight solutions about 0.3 s prior to the solution response (see Figure 3); the gamma frequency band has been associated in prior research with perceptual and semantic integration. Activity in earlier time intervals preceding the solution provides a clearer picture of the insight/noninsight distinction. When comparing the two problem types for the second preceding the insight burst (i.e., 1.5 to 0.3 s preceding the solution response), gamma activity did not significantly differ between insight and noninsight solutions. Finally, the analysis of EEG data in the above research includes a finding in the alpha frequency
441
Figure 3 Differences in neural activity immediately preceding insight and noninsight solutions. The left head model shows a region of greater gamma activation for insight solutions over the right superior temporal gyrus immediately prior to the solution response. Reproduced from PLoS Biology, Vol 2, Mark Jung-Beeman, Edward M. Bowden, Jason Haberman, Jennifer L. Frymiare, Stella Arambel-Liu, Richard Greenblatt, Paul J. Reber, and John Kounios, Neural activity when people solve verbal problems with insight, pp. 500–510, Copyright (2004), with permission from the Public Library of Science, open source.
band that further supports Martindale’s theory of hypothesis selection as a critical component to the creative process. Greater alpha activity was present prior to solution over posterior regions of the cortex for solutions reported to result from insight solution strategies; this activity occurred immediately prior to the right-hemisphere gamma burst. Posterior alpha is linked to sensory gating (the reduction of visual inputs) and could reflect the inhibition of inputs from the environment that might otherwise interfere with the successful retrieval of a weakly activated (unconscious) solution idea.
Special Topics in the Study of Intuition, Creativity, and Insight Is the Unconscious Better Suited for Creativity? Recent publications in the field of creativity suggest that creativity can result from either conscious or unconscious processes. Beginning with Mednick, the idea that creative thoughts are more likely to be achieved via flatter, diffuse associations throughout the brain has been a popular theory. Central to this theory is the idea that each concept is associated to a large number of
442
Intuition, Creativity, and Unconscious Aspects of Problem Solving
other concepts with relatively equal strength. This is in contrast to a steep gradient of association in which each concept is strongly associated to only a small number of other concepts. The need for flatter associations makes sense if we consider that unique combinations of problem subcomponents, whether verbal or structural in nature, often spring from distantly rather than directly related items. For example, the solution to tasks such as remote associate problems (e.g., CRAs) often relies on the solver to generate unique and relatively distant associates of target words. Another popular idea in creativity theories is the importance of the unconscious in generating an extensive list of novel ideas for consideration. For example, Dean Keith Simonton suggested that creative products may be the end result of the recombining of problem components on the unconscious level. According to this theory, the prolific unconscious generates combinations of problem components in search of a solution. Because the unconscious is not directed in the combinations composed, chance plays a considerable role in the ideas produced. Cognitive theories addressing the general mechanisms of unconscious thought suggest that the unconscious may be uniquely suited for this role. The unconscious is believed to have more available resources than the conscious which is limited to processing items within the scope of awareness. In addition, the unconscious is capable of processing information in parallel making sufficient resources available for idea generation, less likely to occur in the serial processing that limits conscious thought. Further support for the role of the unconscious in creative thinking comes from recent studies using neuroimaging techniques which have suggested that gating of external stimuli may be necessary to facilitate creative thought. As noted above, the research of Jung-Beeman, Kounios, and colleagues demonstrated a period of increased alpha power over posterior regions of the cortex immediately preceding insights. This alpha power, linked in prior research to sensory gating, may reflect the need to devote conscious resources to the selection of the best alternative from among those formed by the unconscious. It should be noted, however, that in several theories of creative thought, such as opportunistic assimilation, broader perception (i.e., attention to a
wider variety of external stimuli) can stimulate insight experiences. Few studies have directly explored whether unconscious or conscious thinking is more likely to produce creative products. Preliminary research in the area suggests greater potential for creative output by unconscious processing. In a series of experiments conducted by Ap Dijksterhuis and Teun Meurs, participants were asked to perform generation tasks (e.g., name as many cities as possible that begin with the letter ‘A’) immediately after the instruction was given, after 3 min of directed thinking about the generation task (conscious condition), or after 3 min of performing a distractor task (unconscious condition). Participants in the unconscious condition listed more atypical cities than did participants completing the task under the other two conditions. A similar pattern of results was observed when the alternate uses task (e.g., generate as many uses as possible for a brick) was performed under the same three conditions. Though these results seem promising, the observed effect in the unconscious condition could simply be the result of the passage of time allowing prepotent thoughts to weaken in favor of more remote alternatives. If so, a passive period of incubation may be sufficient to produce these effects rather than the generation of solution ideas by an active unconscious. When we consider whether the unconscious is sufficient for creativity or insight, the answer is no. No matter how prolific the unconscious, creative thought demands that the individual select an idea for implementation. In insight, conscious thought reduces the initial problem space such that the individual is able to reach impasse, setting the stage for restructuring. In creativity, an extensive period of knowledge acquisition, often in excess of 10 years for significant creations, is needed as preparation for creativity. Finally, it is only through illumination that ideas formed by the unconscious can be evaluated as to their merit in Wallas’ final stage of creativity, verification. Equal Preparation for Equal Results? Predictors of Creativity There is a long history in cognitive psychology suggesting that individual differences in attention, working memory, fluid intelligence, and other
Intuition, Creativity, and Unconscious Aspects of Problem Solving
processes translate into differences in task performance. This is no different in creative processing, where factors such as diffuse attention and diffuse semantic networks have been highlighted as key areas contributing to enhanced creative potential. For example, normal adults scoring high on psychometric measures of creativity exhibited weaker performance on measures of selective attention than their less creative counterparts. Also of interest is the finding that normal adults who exhibited strong performance on the RAT were more likely to apply hints presented outside the scope of focused attention when solving subsequently presented creative problems. Thus, possessing reduced selective attention (or increased diffuse attention) may ultimately facilitate the occurrence of creative thought by allowing the individual to consider ideas outside the scope of focused attention, even those ideas generated at the level of the unconscious. Several affective and clinical factors have been examined as to their effects on selective attention. Research has demonstrated that positive affect results in the more diffuse allocation of attentional resources, facilitating creative problem solving, as well as problem solving via insight. A predisposition toward psychopathologies, such as schizophrenia, has been found to coexist with altered cognitive processing in both attention and language and may facilitate insight and creativity. For example, normal adults who exhibit high levels of traits often present in schizophrenia such as sensitivity to light and belief in the paranormal exhibit more diffuse allocation of attention resources. These individuals also present more diffusely organized language networks than do normal adults with low levels of these traits. Finally, a number of correlational studies report that higher levels of schizophrenic-like traits are associated with improved performance on psychometric measures of creativity. A novel area of research has explored brain states that may be linked to a predisposition for creative forms of thought. Researchers have used EEG to examine the link between baseline (resting) activity in the brain and the strategies utilized most often by participants during the solution of anagrams. Anagrams are brief puzzles in which participants must unscramble the letters provided to form a word (e.g., problem: vleo – solution: love). Participants were asked to solve a series of
443
anagrams and, for each solution, to indicate whether that solution was the result of an insight (i.e., sudden) or a conscious search strategy. Interesting differences were observed between participants whose self-reports indicated high levels of insight-based solutions versus those with lower rates of insight-based solutions. Less high alpha and low beta activity was observed over occipital regions in participants exhibiting high rates of insight than in their low-insight counterparts; these differences support prior theories of creativity suggesting that diffuse attention is a key factor in the occurrence of insight-based solutions. Further, high-insight subjects exhibited more right-hemisphere activity over frontal and temporal regions and low-insight subjects exhibited more left-hemisphere activity over frontal and temporal regions. These differences suggest that brain activity during a resting state may influence the form of strategy implemented during problem solving. Researchers have also observed links between the neural activity immediately preceding the presentation of a problem (the pretrial interval) and the type of strategy (insight or search) applied during that problem’s solution. In sum, these results suggest that cognitive, affective, and neural factors may be predictive of creative thought. It should be noted, however, that studies exploring individual differences in this area have focused primarily on the divergent-thinking components of creativity and insight, leaving many aspects of creativity and related processes unexplored.
New Thoughts and Directions Conventional thought concerning the unconscious and its role in creativity and related processes has at times restricted progress in this exciting area. Researchers have been criticized for being narrow-minded when considering the cognitive (diffuse attention) and neural (right-hemisphere) components linked to creative insights. For a substantial period of the field’s history there was little variety in the types of stimuli used in research or, until recently, the techniques at our disposal to study the unconscious and its role in creativity. A re-interest in creativity and related fields such as insight, has coincided with a change in thinking
444
Intuition, Creativity, and Unconscious Aspects of Problem Solving
about the subject area leading to promising research in new areas. It is difficult to review all of the interesting new developments in the study of creativity so only a selection will be provided here. One area of interest is the recent exploration of the purported link between creativity and psychopathology, previously supported almost exclusively via anecdotal data. Although the results of several correlational studies were promising, suggesting a link between creativity and psychopathology, there was little to indicate the exact source of the relationship. Research has begun to examine the connection – studies exploring sensory gating in the first-degree relative of schizophrenics and in normal adults high in creativity have shown that both groups exhibit reduced levels of sensory gating when compared to normal adults who are low in creativity. Also of interest is the finding that normal adults high in traits associated with schizophrenia and those high in creativity both exhibit a more diffuse network connecting semantic concepts than is present in normal adults lower in creativity, thus providing support for the diffuse association network proposed in Mednick’s theory of creativity. Another area gaining momentum is the study of neurotransmitters (chemicals that allow communication between neurons) and how altering neurotransmitter levels affects cognitive processes, such as creativity. Researchers have explored the effect of propranolol, a beta-adrenergic antagonist that blocks the effects of the neurotransmitter norepinephrine, on the ability to creatively solve problems. Over-activity by the noradrenergic system which supplies norepinephrine throughout the brain has been linked to an increase in anxiety and a decrease in cognitive flexibility. The administration of propranolol which blocks the effects of norepinephrine decreased solution times in individuals solving anagrams. In contrast, administration of a beta-adrenergic agonist increased effects in the brain linked to norepinephrine, thus decreasing cognitive flexibility during problem solving. Similar research has been conducted manipulating the neurotransmitter dopamine and exploring the effects of these manipulations on problem-solving ability. A final area to note is research exploring how one’s mindset, altered through the use of counterfactuals, affects creative thought. Counterfactuals
are situations in which individuals reflect on how past events could have had a different outcome with only a slight alteration to the actual events. Counterfactuals can be those in which the thinker alters the situation by adding an event (i.e., if only this would have happened) or by subtracting an event (i.e., if only this had not happened). Counterfactual thinking has been shown to influence the nature of an individual’s thought processes subsequent to the counterfactual, thereby changing one’s mindset or approach to upcoming stimuli. Research conducted by Laura Kray, Keith Markman, and others has demonstrated that counterfactual thinking that is additive in nature facilitates the solution of creative problems, such as RAT problems, whereas counterfactual thinking that is subtractive in nature facilitates analytic thought. These results add promise to the theory that it is possible to alter our abilities for creative thought. Though many new lines of research now exist in the study of the unconscious in creativity and problem solving, this remains a field in which
(a)
(b)
Figure 4 Classic insight problems – solutions. (a) The
candle problem. Empty the tacks from the box and use the box as a ledge to support the candle. Reprinted from Memory & Cognition, Vol 32, Jessica I. Fleck and Robert W. Weisberg, The use of verbal protocols as data: An analysis of insight in the candle problem, pp. 990–1006, Copyright (2004), with permission from The Psychonomic Society. (b) The nine-dot problem. Reprinted from the Journal of Experimental Psychology: General, Vol 110, Robert W. Weisberg, and Joseph W. Alba, An examination of the alleged role of ‘fixation’ in the solution of several ‘insight’ problems, pp. 169–192, Copyright (1981), with permission from the American Psychological Association. (c) The two-string problem. Tie the pliers to one of the strings and get the string to swing back and forth like a pendulum. You would then be able to pull the second string over as far as possible and wait for the first string to swing within your reach.
Intuition, Creativity, and Unconscious Aspects of Problem Solving
there are more unknowns than knowns. It is, therefore, an exciting time to be part of this growing research area. An expanding view of creativity and insight suggests that both conscious and unconscious aspects of thought may result in creativity. Further, it appears that both forms are necessary in some capacity. But is there a fundamental difference between creative achievements that rely primarily on the unconscious and those that result predominantly from conscious thought? Another question of considerable interest is how can we increase the likelihood of creative achievements? Regardless of the domain, business, education, the arts, or the military, the need to think creatively and flexibly is a necessity for successful performance. How can we alter our abilities to think creatively if it is the unconscious and not the conscious that affects our output? These are just some of many the questions waiting to be answered as we move forward. See also: Attention: Selective Attention and Consciousness; Mental Representation and Consciousness; Unconscious Cognition.
Suggested Readings Bowden EM, Jung-Beeman M, Fleck J, and Kounios J (2005) New approaches to demystifying insight. Trends in Cognitive Sciences 9: 322–328.
445
Bowers KS, Regehr G, Balthazard C, and Parker K (1990) Intuition in the context of discovery. Cognitive Psychology 22: 72–110. Dijksterhuis A and Meurs T (2006) Where creativity resides: The generative power of unconscious thought. Consciousness and Cognition 15: 135–146. Fink A, Benedek M, Grabner RH, Staudt B, and Neubauer AC (2007) Creativity meets neuroscience: Experimental tasks for the neuroscientific study of creativity. Methods 42: 68–76. Jung-Beeman M, Bowden EM, Haberman J, et al. (2004) Neural activity when people solve verbal problems with insight. PLoS Biology 2: 500–510. Kounios J, Fleck JI, Green DL, et al. (2008) The origins of insight in resting-state brain activity. Neuropsychologia 46: 281–291. Luo J and Knoblich G (2007) Studying insight with neuroscientific methods. Methods 42: 77–86. Metcalfe J and Wiebe D (1987) Intuition in insight and noninsight problem solving. Memory & Cognition 15: 238–246. Ohlsson S (1992) Information-processing explanations of insight and related phenomena. In: Keane MT and Gilhooly KJ (eds.) Advances in the Psychology of Thinking, pp. 1–44. New York: Harvester Wheatsheaf. Seifert CM, Meyer DE, Davidson N, Patalano AL, and Yaniv I (1995) Demystification of cognitive insight: Opportunistic assimilation and the prepared-mind perspective. In: Sternberg RJ and Davidson JE (eds.) The Nature of Insight, pp. 65–124. Cambridge, MA: MIT Press. Smith SM and Blankenship SE (1989) Incubation effects. Bulletin of the Psychonomic Society 27: 311–314. Wallas G (1926) The Art of Thought. London: Cape. Weisberg RW (2006) Creativity: Understanding Innovation in Problem Solving, Science, Invention, and the Arts, pp. 386–446. Hoboken, NJ: John Wiley & Sons, Inc.
Biographical Sketch
Jessica Fleck is currently an assistant professor of psychology at the Richard Stockton College of New Jersey. She completed her undergraduate education at Shippensburg University, majoring in psychology and criminal justice. She then obtained her PhD in brain, behavior, and cognition from Temple University, studying under Robert Weisberg. Prior to joining the faculty at Stockton, Fleck was a postdoctoral research associate in Drexel University’s Electrophysiological Laboratory, working with John Kounios.
446
Intuition, Creativity, and Unconscious Aspects of Problem Solving
Her published research includes topics such as the cognitive and neural components of problem solving, creativity, and working memory, as well as the relationship between creativity and personality. At present, Fleck is researching cognitive and neural factors in the link between creativity and schizophrenia.
John Kounios attended Haverford College, where he received his BA with a double major in psychology and music theory/ composition. He then enrolled in the University of Michigan’s graduate program in experimental psychology, where he received his PhD under David E. Meyer. Kounios has held research and faculty positions at Princeton University, Tufts University, the Boston Veterans Affairs Medical Center, the University of Pennsylvania, and Drexel University. He has published research on a variety of topics in cognitive psychology and cognitive neuroscience, including knowledge representation, insight in problem solving, creativity, and episodic memory. He is currently a professor of psychology at Drexel University where his research focuses on the neural and cognitive bases of semantic information processing, problem solving, and creativity. His research has been funded by grants from the National Institute of Mental Health, the National Institute of Deafness and Other Communication Disorders, and the National Institute of Aging. His research has been reported by National Public Radio, US News & World Report, Scientific American, The Times (London), The Wall Street Journal, and other print and electronic media.
Language and Consciousness P Gordon, University of Columbia, New York, NY, USA ã 2009 Elsevier Inc. All rights reserved.
Introduction Although language is often considered to be the very medium by which our thoughts are made conscious, the extent of conscious control of language may be more illusion than reality. In thinking about the role of conscious and unconscious processes in language, one is reminded of Freud’s metaphor of the mind being like an iceberg: Consciousness is represented as the relatively small exposed surface of the iceberg that is supported by vast structures below the surface representing the unconscious mind. Although Freud’s metaphor refers to the dynamic processes driving our psychosocial interactions, the metaphor is nevertheless apt in thinking of the vast hidden structures that support the ability to produce and comprehend utterances in one’s native tongue. To explore the relationships between language and consciousness, we must first understand what kinds of categories of conscious experience are linguistic in nature, which experiences are made available or highlighted by language, and which are left under the waters of the unconscious mind.
Conscious Linguistic Processes Linguistic processes that impinge on our conscious experience require a medium through which the phenomena or qualia of consciousness can be experienced. Such media include the familiar modalities of speech, hearing, reading, and writing. In nontraditional groups or cultures, members might use other means of communication such as signing in the deaf, Braille in the blind, and Tadoma in deaf-blind persons. In some languages, the familiar speech segments of consonants and vowels are augmented by meaningful tonal structures, where the same syllable spoken with low, mid, high, rising, or falling intonation might each have distinct meanings. There is a classic sentence
in Thai that means: ‘‘Who sells chicken eggs?’’ but to most nontonal speakers the sentence sounds like the same syllable repeated four times. Consciousness of language therefore extends to consciousness of distinctions in one’s native language and is affected by the meanings we assign to different sounds and qualities of sound. In other cultures, such as the Piraha˜ of Lowland Amazonia, communication with children may occur through the use of ‘hum speech’ that finesses a cultural convention against directly speaking to children. In hum speech, speakers hum the tonal structure of the intended utterance without use of vowels or consonants. Whistle speech in the Piraha˜ is a similar mechanism for communication between adults when hunting in the jungle or communicating over long distances where the discrete syllabic elements of speech are deleted and the overall tonal contour of the utterance is communicated instead. Lest we think that such communicative modes are exotically restricted to the jungles of Brazil, consider the communicative act that consists of: uh uh uh (low, high, rising), where the message is clearly understood as ‘‘I don’t know,’’ often uttered by bored, put-upon teenagers. In such cases, the conscious experience of such signals is completely different when they are interpreted as linguistic than when they are received by someone unfamiliar with the language and its conventions. In addition to the modes of communication that transmit the actual interpretable symbolic structures of the language, we also communicate with paralinguistic gestures that do not contain discrete linguistic elements such as words, affixes, and sentences, but nonetheless modulate the meaning of a message. These paralinguistic elements include manual, facial, and intonational gestures that may accompany or replace spoken language. An utterance can have the completely opposite meaning depending on whether one uses a neutral tone or a sarcastic one (compare ‘‘That’s really nice of you’’
447
448
Language and Consciousness
spoken in an enthusiastic/neutral tone versus a sarcastic tone). Emphasizing one word in a sentence can also radically alter the intention of the message (note the different intentions of a hat sales person saying: ‘‘You want that hat?’’ and ‘‘You want THAT hat!’’). In a similar manner, facial gestures, smiling, showing disgust, or doubt, can alter the intended meaning of an utterance as well. We can infer that experience of messages that have distinct meanings in this way are distinct conscious experiences to the extent that meaning modulates or scaffolds the conscious experience of linguistic qualia. Manual gestures are also known to be a rich source of information accompanying speech. While such gestures normally augment the spoken utterances, research has shown that the two modalities are sometimes mismatched within the same communicative exchange: People can say one thing while gesturing the opposite. In a remarkable experiment by Susan Goldin-Meadow and her colleagues, children were asked to explain how quantities are affected by changing the configuration of the items in Piaget’s famous conservation task. Children who were going through the beginnings of a cognitive reorganization employed manual gestures that indicated knowledge of the correct, more advanced, solution, even though their simultaneous, spoken explanations were more immature reflecting earlier stages of cognitive development. This mismatch between language and gesture suggests that unconscious thought processes can bleed through into modalities that are not being utilized by the more conscious process of communication. The idea that there can be uncoordinated parallel processing of information where only one channel can be verified through conscious linguistic report is reminiscent of earlier studies by Roger Sperry and Michael Gazzaniga and their colleagues on split brain patients. In these studies, patients had undergone commissurotomy – a surgical procedure that severs the corpus callosum, a bundle of nerve fibers connecting the two cortical hemispheres of the brain. This procedure, which is carried out to relieve massive neural cross firing during epileptic seizures, made it impossible for the two sides of the brain to talk to each other.
Experiments with these patients examined how the two hemispheres processed information by presenting stimuli separately using a split screen presentation so that information presented in one half of the visual field could not be seen by the other side of the brain. Information presented in the right visual field was sent only to the left hemisphere, which is the hemisphere that is primarily responsible for language. Similarly, information presented in the left visual field was processed only by the right hemisphere, which is more responsible for analyzing spatial and other holistic information about the world. When words were presented to the left visual field, split-brain patients were unable to name them. However, because the right hemisphere is able to read words, but not speak them, the patients were able to identify the meaning of a word like ‘spoon’ by reaching onto a table and picking up the appropriate object. Like the case of the asynchrony between language and gesture, the case of split brain processing of language suggests a similar split in the consciousness of reading processes. Other phenomena exist where language lurks beneath conscious experience and can suddenly pop up into consciousness without deliberate control. The ‘tip-of-the-tongue’ phenomenon involves the familiar experience of searching for a word to fit a particular context that you just cannot find in your mental lexicon. In an early study of this phenomenon, the psychologist, Roger Brown, showed that the word can be accessed and brought to consciousness by providing clues such as the initial sounds of the word. Sometimes, trying to find a word seems to actually inhibit success, and it is not until we stop trying that the word itself pops into consciousness. This phenomenon suggests that when one actively searches for a word whose semantic properties (meaning) fit in a particular context, one is reactivating the semantic features that are already activated by the context. It is possible that reactivating the semantic features in this manner inhibits activation of the phonological features that allow you to access the form of the word. The strategy of searching for the word by going through the alphabet can often be more fruitful in such cases because it focuses on phonologically rather than semantic properties. In the tip
Language and Consciousness
of the tongue phenomenon, we see a particularly striking experience that highlights the border between unconscious and conscious processing of language. In summary, consciousness of language requires a sensory medium through which the linguistic signals can be experienced at the level of qualia. These signals can take on many forms and are as varied as the extent of human ingenuity. Conscious awareness of language is not just of the experience of shapes, movements, and sounds that are used to transmit the signals, but it is the meaningful interpretation of those shapes, movements, and sounds as words, sentences, and speech acts. Meaning can be like an attentional spotlight that focuses the mind on certain aspects of the signal, albeit unconsciously, but the resulting conscious experience can be very different depending on how the behaviors are interpreted. Awareness of the intent of the message often means interpreting subtle cues from multiple sources. Gesture, expression, and tone of voice conspire to modulate the meaning of the message as a whole. We might say, in such cases that there is an effect on the level of the ‘speech act.’ For the most part, unless a gesture or expression is completely obvious or highly conventionalized such as the profane cultural forms (middle fingers, V-shaped fingers, pumping the bicep, etc.), the processing of information from these paralinguistic channels is often not done consciously and requires expert scientific study in order to fully understand the meanings behind them. Sometimes, a totally artificial form of encoding language might come along, and not only are the signals experienced within the sensory domain of their transmission, but higher order properties are also picked up through the analysis of invariant properties across situations of use. For example, radio operators who are experts in using Morse code – a sequence of short and long beeps – are able to recognize the identity of the person sending the message based on their signature rate and rhythm in hitting the telegraph key much in the same way as if they were recognizing a familiar voice. In fact, when one recognizes a familiar voice, there is a special conscious experience of this voice in which the conceptual properties of the whole person are experienced at some level.
449
Unconscious Aspects of Language Below the level of communication at which the conscious signals are processed, there are multiple levels of unconscious processing of linguistic inputs and outputs. These include levels of phonology (sound structure), morphology (word structure), syntax (sentence/utterance structure), semantics (meaning), and pragmatics (appropriate and conventional social use of language). When we attempt to bring these levels of language processing to conscious introspection, we refer to this as metalinguistic knowledge (knowledge about our implicit knowledge of language). Because of the unconscious and inaccessible nature of linguistic structures and processes, attempts at conscious reflection of language are often fraught with possibilities for error. More often than not, those who claim to be language experts – the so-called ‘language mavens’ are not trained in formal linguistic analysis – show confusions about many aspects of grammar and thereby fail to grasp the reason why certain constructions might or might not sound acceptable in the language. One of the major advances in our understanding of language came in the latter half of the twentieth century from Noam Chomsky’s approach to linguistics, which became known as ‘Generative Grammar.’ Chomsky’s proposals represented the beginnings of modern cognitive science in which cognition as mental representation is characterized as a set of algorithms or rules in which symbols are formally manipulated (moved, rearranged, rewritten, etc.). Implicit in the theory of generative grammar is the idea that languages have a mathematical-like structure that generates the sentences and utterances that we use in everyday life, and that facts about the underlying grammar can be discovered by using a formal analysis of linguistic structure. Grammar is said to consist of category symbols like N(oun) and V(erb), which are like variables in algebra (x, y). Rewriting category symbols as words (N ! dog) is like substituting a variable with a numerical value (x ¼ 5). The method for discovering the systematic principles of grammar involves formulating rules and seeing if they generate grammatical or ungrammatical sentences when words are substituted as values for the category symbols. Thus, grammaticality
450
Language and Consciousness
judgments became the metric for judging whether a particular grammatical hypothesis was a good fit to the language. Grammaticality, the conscious perception of a sentence as either sounding ‘good’ or ‘bad,’ became the data for the science of linguistics. Sometimes the judgments were extremely subtle and led to remarkable insights into the nature of grammatical competence. For example, changing the wording every so slightly in the following two sentences leads to strikingly different perceptions of grammaticality: Who did Mary tell the man that Bill hit? *Who did Mary tell the story that Bill hit?
The asterisk indicates the common judgment that the second sentence sounds bad compared to the first. Yet both would appear to be derived from questioning of perfectly grammatical sentences: Mary told the man that Bill hit Tom Mary told the story that Bill hit Tom
In both cases, the questioned element (Tom) is substituted with a wh-question form (who) and is moved to the front of the sentence. Yet, for some reason, it is only the first sentence that leads to the experience that it is part of the English language. In this instance, we have a consciously differentiated experience of two sentences that are remarkably similar in their actual wording. In one case we experience a regular sentence that is part of our language, and in the second case, we experience a string of words that is jarring and outside of the permissible word combinations of the English language. In such cases, speakers of a language generally agree about their experiences of whether something sounds good or sounds bad, but we are completely unable to articulate the reasons that we share those common judgments, since their origins lie within the hidden depths of our unconscious language processing systems. As mentioned previously, when people attempt to formulate rules of language through their own conscious introspection and intuition, they do not access the level of analysis explained by generative grammar, and hence tend to invoke prescriptive rules that are made up by language mavens and other guardians of the mother tongue. Sometimes, language experts make disastrous pronouncements
about what is, or is not, ‘proper grammar.’ For example, African American English was called ‘illogical’ because it failed to show verb agreement with the subject of the sentence (e.g., I be going); whereas a deeper analysis of such vernaculars revealed a language that was every bit as complex and logically consistent as the received dialect of ‘Standard English’; it was just a different set of rules. In cases like this, formally trained linguists generally agree that the explanation of why language sounds good or sounds bad, should not be left to the language mavens, but should be explored using the formal tools of linguistic analysis that can gain insight into the unconscious representations of grammar that undergird our conscious intuitions about grammaticality. The rise of generative grammar in the late 1950s and early 1960s stood in stark contrast to the then dominant behaviorist approaches within psychology that eschewed any study of mental processes as mediating stimulus and response. Along with the concomitant rise of the digital computer in society, there was a trend for researchers in the cognitive sciences to begin to think about the mind as a kind of computer. Unconscious thought processes were like the hidden software codes that generate the programs that we interface with on the computer screen. The computer metaphor for the human mind is apt in many ways for thinking about how consciousness interfaces with the rest of cognition. What appears on the monitor screen of a computer is a useful, if not complete, analogy for what makes it to consciousness in the human mind. In a computer, there are whole unseen worlds of complex computation slaving away inside the memory and processing chips that allow us to interface at only a superficial level with what appears on the screen as we breeze through writing a document using a text editor interface. If all of the computational codes were to appear on the screen, we would not be able to function effectively. So it is with language, that mental processes are churning away to allow us to use language with little or no effort, yet we are not consciously aware of the machinery that permits us to speak in grammatical sentences and to present thoughts in a conversation. When Chomsky began to characterize the extremely complex nature of grammatical computation, he noted the unconscious nature of the
Language and Consciousness
knowledge underlying grammar, and likened it to the idea of ‘tacit knowledge’ explored by the philosopher, Michael Polanyi. In his writings, Polanyi explored ways in which knowledge can guide our behavior in a manner that is not open to conscious introspection. So it is, that the Freudian iceberg is appropriate for describing the unconscious computation that goes into producing the conscious experience of language that is quite marginal in many ways.
Modularity and Consciousness The relationship between conscious and unconscious aspects of language can be usefully addressed by employing the concept of ‘modularity,’ which was proposed by the psychologist and philosopher, Jerry Fodor in the 1980s. Fodor proposed that the mind is composed of a set of horizontal and vertical faculties. The horizontal faculties are generalpurpose processes that cut across domains. These might include mental abilities such as memory, reasoning, and judgment. For all intents and purposes, such faculties can potentially draw on information of any kind to form associations and to make inferences. Vertical faculties, on the other hand, are restricted to a specific domain of computation. In Fodor’s original conception, these vertical faculties, which he called modules, were small in number, consisting of a few core processes such as perceptual processing in the sensory domains (vision, hearing, etc.) and language. Modules are characterized as being dumb but fast. They are automatic and epistemologically encapsulated. In other words, the mind accomplishes rapid processing within the module by processes that specialize in formal computation over the elements in the system and are immune to external influences arising from general knowledge. A classic example of modularity in the domain of vision is the Muller–Lyer illusion shown in Figure 1. In this illusion, we perceive two lines as being unequal in length even though we know them to be of equal length (this can be easily demonstrated by removing the biasing end pieces). Even the most seasoned vision scientists, who have seen the illusion thousands of times cannot persuade themselves to consciously perceive the lines
451
Figure 1 The Muller–Lyer illusion.
as equal. Instead, the visual system is bound by automatic processes that cannot help but produce a conscious perception of unequal length. This is because the vision module can only follow the unconscious computational rules of seeing and is destined to ignore any conscious knowledge we may possess of the actual state of world (i.e., consciously knowing that the lines are equal in length). In this case, we see that dual consciousness – being conscious of the state of the world, and conscious perception of a stimulus – does not guarantee their mutual influence and coherence. Rather, the conscious output of the modular process takes precedence and is immutable, and multiple mutually incompatible conscious states can coexist. The claim that language processing is similarly modular in nature suggests that the outputs of linguistic processes are also automatically computed and are isolated from general reasoning processes. Some researchers have suggested that parsing biases in language show characteristics of modularity. For example, there is an extensive literature on a phenomenon known as ‘minimal attachment.’ This is a principle that concerns a certain bias in the human grammatical parser for speakers of English. For example, in certain constructions that contain relative clauses, we may find sentences where the meaning is ambiguous depending on what part of the sentence the clause is thought to modify. For example, in the sentence: ‘‘The spy saw the cop with the binoculars,’’ there is an ambiguity as to whether the spy or the cop had the binoculars. Experiments with such sentences suggest that there is a preference to attach the subordinate clause, ‘with the binoculars,’ so as to produce the least amount of additional structure. In this case, it means that the preference is to interpret the sentence as the cop having
452
Language and Consciousness
the binoculars despite the possible bias in meaning that would associate the spy seeing with binoculars. Sometimes the effect of modularity on the conscious perception of language is more obvious at the level of phonology – the perceived sounds of language. One example of how language channels conscious perception of language is in the difference between perceiving the sounds of one’s own native language and that of an unfamiliar foreign language. Fodor gives the example that native speakers of English generally know what Chinese ‘sounds like’ and what French ‘sounds like,’ but it is harder to have a sense of what English sounds like – and similarly for whatever native language one might speak. The inability to perceive a general sound of one’s native language based on its generalized tonal and rhythmic properties suggests that, for competent speakers of a language, there is an obligatory processing of that language whereby sounds are converted into meanings and there is no conscious access to generalized phonological properties that would constitute the ‘sound’ of the language. Such experiences are afforded only when the sounds are left uninterpreted. When listening to one’s own native language, processing of that language into meaning is obligatory. We cannot help but convert linguistic signals into meanings in the same way that we cannot help but see an object as an object when it is presented to our visual system. The output of a language comprehension module is a set of structured representations of meanings of the propositions encoded by the sentences being received. Conscious perception of the abstract sound characteristics of the speech in which this message is encoded is minimal. This effect is rather like the inability of the artistically untrained eye to perceive the formal structure of an object and its relation to the perspective of a scene. The awkward drawings of the untrained amateur are notoriously unsuccessful because they are guided by the attempt to capture the ‘meaning’ of an object in its environment rather than capturing its actual formal linear properties, which require training to divorce oneself from the scene and see the lines and the structure of objects in space rather than to obey one’s intuition. The old trick of copying a drawing upside down serves
to excise meaning from a scene and allows the budding artist to concentrate on form. In like manner, a copy editor might read a text backwards to detect typos like repeated words that might otherwise go undetected because the language processor focuses on meaning rather than form. Modularity makes language processing an obligatory process. It is only when we hear an unfamiliar language that we consciously perceive the detailed contour of the sounds unaffected by meaning. We cannot choose to perceive English or some other native language as simply a sequence of sounds, but rather are required to interpret those sounds as propositions and speech acts, first and foremost. In the written domain, there is a similar effect in the Stroop task, where words naming colors are presented in colored fonts that are distinct from the name of the color encoded in the word. Suppressing the color word being read and naming only the color of its font is one of the most difficult tasks in the psychological test arsenal. Thus, in the modality of reading, the processing of language as meaningful is obligatory and suppresses conscious attention to the physical characteristics of the stimulus, in the case of the Stroop task, the color of the font.
The Role of Consciousness in Symbolic Representation While we have explored the idea that unconscious mental computation of linguistic forms sends up to consciousness a mental representation of the meaning of a word or utterance, there is almost no scientific understanding of the notion of ‘meaningfulness’ itself. Language researchers who study meaning or semantics often analyze how meanings relate to each other and how we can draw inferences from linguistic utterances. For example, researchers might develop a theoretical framework that explains how we represent that a dog is a kind of animal, and that telling someone to do something and persuading them to do so have different implications regarding whether or not they actually did it – only the latter entails that the action was completed. However, such approaches only analyze the formal relations between meanings, not the nature of meaning itself. Similarly, the
Language and Consciousness
analysis of grammatical structure and the claims about modularity of such processes rely on a formal analysis of the syntactic structure of the language. Formal analyses of grammar can tell us how we put strings of words together to form sentences, but the analyses do not really tell us anything about how those sentences allow us to refer to objects, events, feelings, and so on that relate language to the real world. Nobody really understands how to capture what it means for a brain to possess meanings for words such as dog, persuade, or virtue. Such puzzles have plagued philosophers since the time of Plato and do not have recommendable solutions. Imagine trying to build a robot that was able to recognize instances of virtue in the world. What would its software program look for and how would it interact with the world to be able to do so? Perhaps the strongest critique of the purely formal analysis of language comes from the philosopher, John Searle, who illustrates the emptiness of formal linguistic analysis with the thought experiment known as the Chinese room. In this thought experiment, we imagine a person who does not speak Chinese, and is placed in a room where they receive pieces of paper with Chinese writing on them. When these pieces of paper are received, the person in the room consult a book containing a set of rules that allows them to respond appropriately, again using Chinese symbols. The moral of this thought experiment is that, while the inhabitant of the Chinese room functions in a manner that is indistinguishable from an actual speaker of Chinese, we cannot say that this person understands Chinese, or that there is anything about the room itself that understands Chinese. This case extends as an analogy to the approach in cognitive science that characterizes linguistic comprehension and production as the manipulation of formal symbolic structures. According to Searle, understanding a language requires a conscious individual that stands in a particular relationship to the formal symbolic structures in the language. He notes that a symbol is a formal element that represents something (the word dog is a symbol that represents instances of dogs in the world, and the word virtue represents instances of virtue in the world). But symbols cannot function to link language and the world
453
without some sentient individual who is able to understand the relationship between the symbol and what it represents. In other words, a symbol is not a symbol without some individual who can recognize it as such. Otherwise, it is just some meaningless shape or sound. In this respect, Searle suggests that language understanding cannot occur in the absence of consciousness. Otherwise, we are dealing with a version of the Chinese room inside our heads. The argument is a powerful indictment of the claim that we can understand language and cognition in general as a process of symbol manipulation alone.
How Language Affects Conscious Perception of the World In previous sections, we have illustrated how language can shape the forms of consciousness that we experience when we hear language spoken or see language written or gestured. First, the very fact that we perceive sounds as meaningful language means that we hear those sounds in a very different way than if they were uninterpreted. Since the very early study of speech sound processing, researchers have been interested in how language affects our perception of speech sounds, as opposed to other sounds. The perception of distinctions between speech sounds is shaped by whether those sounds are used in language or not. We are all familiar with the common alterations in the phonological systems that are made by foreign speakers attempting to speak English, and, of course, the same would be true of English speakers attempting to speak in a foreign language. Speakers of a particular language tend to show common changes to a foreign tongue. French speakers might say ‘ze’ instead of ‘the,’ Japanese speakers might say ‘psychorogy’ instead of ‘psychology.’ The commonality of such transformations by foreign speakers highlights the differences between the inventory of sounds employed in the native tongue versus the phonology of the language in which they are attempting to speak. Since the ‘th’ sound is not available in French, speakers fall back on the closest equivalent, which is produced with the tongue slightly further back in the mouth resulting in the ‘z’ sound.
454
Language and Consciousness
In addition to being unable to produce the target sound of a foreign language, such attempts are often confounded by the inability to hear sound distinctions required in the foreign language. For example, when Japanese speakers substitute ‘r’ for ‘l’ (and vice versa) in English words, it is partly because they do not hear this distinction. Instead, the perception is slotted into the existing Japanese phoneme (speech sound) that consists of a sound somewhere in between ‘r’ and ‘l.’ In this case, we see that speaking a particular language focuses our conscious perception of speech sounds into categories of sound that are used in our own language and this makes us unaware of sound categories and distinctions that are used in other languages that employ different sound distinctions. Research on speech perception in infants has tracked the developmental trajectory of speech perception in exquisite detail. In many cases, the ability to distinguish between speech sounds is not acquired when a particular language is learned but is lost when the native language does not make use of a particular distinction. For example, Japanese infants would have no problem differentiating between river and liver, but that ability would be lost sometime between 6 and 12 months of age, as the language they are exposed to continually fails to make use of such a distinction. On the other hand, if the same babies were adopted into English-speaking families, then they would not lose their ability to discriminate these sounds, and they would be incorporated into the English sound system they would be acquiring. Of course, the same is true of infants born into Englishspeaking homes, who can distinguish sounds made in various exotic languages that their parents cannot perceive and they too would lose within the second half of the first year of life. One other aspect of speech perception that has been of particular interest is the notion of ‘categorical perception.’ Perception of speech sounds, in some respects, is similar to the perception of colors. In the case of color, we know that electromagnetic radiation reflected from objects impinge on our eyes as continuous differences in wave forms, the whole range of which can be seen in a rainbow or when a prism is used to disperse white light into its component wavelengths across the spectrum. Our perception of these wavelengths,
however, is not as a continuously changing phenomenon, but as consciously distinct qualia, in the form of discrete colors. Our visual system cuts the continuous spectrum into discrete colors seen in the rainbow or the refraction of a prism. By analogy, there are also variations in speech signals that can be presented as continuously changing along a temporal or frequency spectral dimension, yet we perceive the changes not as continuously changing but as discrete categories much like we perceive colors as discrete. Such categorical perception occurs when people are presented with sounds that vary in the timing between the onset of a consonant and the vibration of the vocal cords signaling the following vowel. Hard, or voiced consonants like /b/ /d/ /g/ show almost no interval between the consonant and following vowel, when we hear a syllable like ‘bah.’ The softer, ‘unvoiced’ consonants like /p/, /t/, and /k/ have a period of silence between the vibration burst of the initial consonant and a following vowel, as in the syllable ‘pah.’ When the timing of the consonant and vowel, know as the voice onset time (VOT), is artificially manipulated, humans perceive the change, not as a series of gradually changing sounds, but as one of two distinct speech sounds, either /b/ or /p/ depending on whether the VOT is greater than about 30 ms. Therefore, much in the same way that a continuously changing stimulus is perceived as a discrete set of categories, speech perception is like color perception in this respect, and the qualia of conscious perception in speech are those categories that are differentiated by the auditory system.
Perceiving the Structure in Language: Critical Periods in Language Learning Effects of exposure to language found in speech perception can presumably be found in other modalities in which language is communicated. However, the visual modality of sign language presents special opportunities to study the effects of being exposed to language at different ages. Few children learn sign language at the age or with the amount of natural exposure that one would learning a native spoken language. Some learn sign later in
Language and Consciousness
life when exposed to a signing community. Hearing signers usually choose to learn sign at a relatively late age to communicate with family members or friends or for professional reasons. Extensive research on the acquisition of sign language by Elissa Newport, Ted Supalla, and their colleagues reveals differences in the way that early learners and late learners perceive and produce sign language. Research shows that a fully articulated sign language like American Sign Language (ASL) has the richness and discrete structure of any spoken language. Newport and Supalla have demonstrated that mime-like sign sequences that appear iconic rather than arbitrary in nature contain discrete elements that are often imperceptible to the hearing parents of deaf children who learn sign language later in life to communicate with their children. Young children who learn sign language as a native language are able to pick out the discrete structure in the sign system that is imperceptible to their parents and other late learners. For example, when a sign such as a flat hand shape is used as a sign for a vehicle, it can undergo motions that include paths in straight lines, zigzags, ups and downs, left and right turns, and so on. To the late learner, such motions appear iconic in nature. That is, they appear to refer by resemblance to their message. To the native signer, such sequences appear as discrete finite elements in a systematic use of limited hand motions. These motions are used to indicate linguistic elements of path, manner and direction, much in the same way that we use prepositions like at, across, to, toward, and so on, in English. A similarly striking phenomenon has been found in Nicaraguan Sign Language (NSL), which only emerged as a conventional sign language in the 1980s after the Sandinista government attempted to develop centralized comprehensive education for deaf children and adults, who had previously been isolated from each other around the country. Initial attempts by educators to teach the students how to communicate were not successful. Instead, NSL spontaneously grew out of the attempts of young children to communicate with other deaf children. Because of this unique situation, it was possible to compare those who came to the deaf community early in life with those who came later in their development. The striking finding was
455
that only the early learners managed to fully incorporate the grammatical structure of the new language into their communications. Later learners, whilst still communicating with others, were unable to internalize the grammatical functions in the same way. Deaf signers who are isolated from a natural sign language develop their own ‘home sign’ systems. These are iconic gesture systems that each individual deaf child invents to communicate with other (hearing) people around them (signs for objects will either involve pointing at the object if present or mimicking its shape or function; signs for actions generally involved mimicking the actions). Such gesture systems lack the subtle and complex grammatical structure of a conventional language. Consequently, those who relied on this deficient grammatical system and never learned to communicate fluently in NSL missed out on the subtle functions encoded by the grammatical structure. Annie Senghas, a researcher who studied the emergence of NSL in the Nicaraguan deaf community, developed a simple task in which Nicaraguan deaf children and adults were tested for their ability to sign spatial positions of people and objects. Participants viewed a sequence of pictures of a man and a tree in nine different combinations of orientation and relative position (e.g., man to the left of the tree facing forward). Their task was to describe to another participant which of the nine configurations was present in each picture. Late learners of NSL were confused and could not signal relative configuration, nor could they see the directional components in the messages of other signers. On the other hand, the children who learned NSL with native-like competence developed a fast and efficient way to uniquely specify relative position and direction that was easily transmitted to other early signers of NSL. This ability to communicate rapidly and in agreement with others is not something that was explicitly articulated within the community as a prescribed form. In other words, there was no set of individuals who decided what the gesture forms were going to be that would encode the spatial functions within this system. It simply emerged from the natural process of young children being put into a communicative environment that employed gesture as a medium. As a result, a real language forms
456
Language and Consciousness
that is capable of fully encoding discrete and complex meanings that are only interpretable by others who acquired the NSL in this manner. Those who came late to the language could not break the code. In addition to the spontaneous nature of how this system emerged within the deaf community, native signers cannot easily explain the structure of the system. That is, they are not consciously aware of what they have created. For example, they cannot explain how the system of signs works that allows them to specify the relative positions of the elements in the Senghas task. This is rather like speakers of English who often fail to articulate the basic principles of grammar underlying their spoken competence, where their consciousness of language at the metalinguistic level fails. It is as if native signers of NSL come to a common consensus merely by dint of being exposed to the language early as a first language. It appears then that conscious perception of language differs depending on the circumstances under which that language is acquired. To summarize, the perception of linguistic messages can be highly influenced by developmental experience with the language in question. Regarding effects of age of exposure, it appears that being exposed to a language early in life leads to induction of the structure underlying the formal structure of the language. This, in turn, leads to the perception of functions that are encoded by the structure of the grammar that are often not encoded by those exposed to language later in life. In other words, if language exposure occurs late in the developmental history of an individual, the grammar does not develop fully, and so the messages conveyed by the language heard or seen are deficient. Hence, the conscious interpretation of language appears to depend on the very subtle grammatical clues that are conveyed perhaps at a non-conscious level, but affect the conscious experience of the message.
Language, Thought, and Consciousness The way that language itself is consciously experienced in its various forms is affected by the way
that language shapes the medium within which it is transmitted. Whether it be the categorization of sounds or the ability to perceive discrete structure in the gestures of sign language, as humans we are highly suited (perhaps evolutionarily adapted) to pick up and create the regularities of linguistic structure. If experience with language can shape the way that we consciously perceive linguistic signals in their own medium, can experience with a particular language also shape the manner in which we experience reality outside of the context of speaking, signing, and listening to language? It is interesting to note, in the context of the Nicaraguan experience, that children and adults who learned the conventional sign systems of NSL also appeared to experience a change in the way that they thought about the world in general. In one telling comment from a BBC documentary about the phenomenon, a deaf Nicaraguan woman who had learned to sign commented that for the first time, she knew what it was to think. Is there a sense in which language causes us to think or at least shapes the way in which we think? Although researchers generally reject the idea that we think using language as a general rule, there remains the question of whether the structures of language create possibilities for thought that would not exist without the existence of language. Before generative grammar became the predominant theory in linguistics, a highly influential approach to the study of language and thought in the middle of the twentieth century was the Sapir-Whorf hypothesis, named after the linguists Edward Sapir and his student, Benjamin Lee Whorf. This hypothesis proposed the principles of ‘linguistic relativity’ and ‘linguistic determinism.’ Linguistic relativity proposed that languages could differ radically in the manner by which they encode reality, and linguistic determinism proposed that language can causally determine how reality is perceived in nonlinguistic domains. Although highly influential for many years, the ideas developed by these linguists fell out of favor amongst linguists and psychologists for various reasons. In Whorf ’s writing about this topic, he often assumed what he set out to prove. That is, if a language encoded a particular mode of expression in a manner that seemed radically different from English, it was assumed that this shaped the way
Language and Consciousness
that speakers thought about the world. Sometimes this involved taking grammatical elements too literally without knowing how they are actually interpreted in the minds of people in the culture. For example, when we say: ‘‘I ‘will’ go’’, there is no implication that there is any ‘will’ (i.e., intentionality) involved in how we interpret this expression. Instead, it is just a neutral construction for future tense. Ironically, Whorf took the existence of a similar construction in the Navajo language to claim that Navajos did not have a conception of future time. Unfortunately, he failed to notice that the English future tense construction was entirely parallel and his reasoning would imply the English speakers also cannot conceive of future time. The general problem with the approach was a lack of independent evidence that differences in linguistic expression were accompanied by perceptual or conceptual differences that contrasted between speakers of radically different languages. Studies that used empirical methods to test whether independent measures of perception and cognition were correlated with differences in linguistic structure focused initially on the domain of color names. It is well known that languages in some cultures use only sparse naming systems for colors and do not cover the color categories of the spectrum found in English. A language might have a word for ‘dark’ and ‘light’ and, perhaps, ‘red’ and ‘green.’ Obviously, in naming other colors, such terms would bleed over into yellow, blue, and so on, and so the direct translation into the English terms should not be taken literally. In the 1970s, Eleanor Rosch studied color memory in the Dani tribe of Melanesia, who possess only two color terms. In reporting the results, Rosch focused on the similarities between Dani speakers and English speakers. In particular, she found that Dani speakers showed better memory for focal, saturated colors when compared to nonfocal, unsaturated colors, even though they had no names for these colors. These results were used to argue that language was not necessary for organizing the color spectrum around the focal colors of the color categories found in English. More recently these results have been questioned and it has been pointed out that, when memory for colors is tested, there is a clear deficit found in cultures that have few color names. In addition, Rosch original results had design problems. When
457
corrections were made for order effects in the tasks, there was a failure to replicate the similarity between the Dani and English speakers with respect to the advantage in memory for focal over nonfocal colors. Debi Robertson studied color naming and memory in the Berinmo tribe of Papua, New Guinea. As in the case of the Dani, she similarly found very poor memory for colors and no effect for focal colors. Instead she found that performance was determined by how the language formed boundaries between colors. Unlike English, the Berinmo language showed no blue-green boundary, but instead had a boundary between names for colors ‘noi’ and ‘wor.’ Although inhabiting roughly the same area of the color spectrum as blue and green, these two color names covered very different ranges of colors. The category boundary between these two color names was what determined their memorability and naming properties. In other words, the way that words in the language carve up perceptual spaces can have a strong effect on how those spaces are perceived and committed to memory. In a clever study, Jonathan Winawer showed that Russian speakers are faster than English speakers in a discrimination task where one is required to distinguish between shades of blue. He attributed this difference to the fact that Russians have distinct words for light and dark blue, which are not found in English. Language also has effects on the perception of number. Several cultures do not have the fully elaborated counting systems that we take for granted in developed western societies. This difference allows us to ask whether the lack of number words affects perception of number. Cultures that live in small communities with limited need for trading do not require the ability to keep track of numerical values for objects and individuals. One can keep track of individuals without actually counting them, but generally accurate exact tracking of individuals is restricted to small numbers up to about 3 or 4 depending on the demands of the task. The ability to keep track of small numbers is often called ‘subitizing,’ which is thought to involve an immediate apprehension of the numerosity of a set in the absence of counting. In experiments with people who can count, researchers have designed tasks to prevent them from counting
458
Language and Consciousness
using very rapid presentation of stimuli and/or having participants repeat the word ‘the’ to prevent sub-vocal counting. In a culture that does not count, such measures are not necessary, and arrays of objects can be inspected for up to several seconds before performing a task. Cross-cultural studies of numerical cognition have been performed with cultures that lack elaborated counting systems, including the Piraha˜ and Munduruku tribes of Amazonia. The Piraha˜ have two words for numbers, ‘ho´i’ and ‘hoı´’ (falling and rising tone respectively), which are relatively imprecise, but are conventionally described as a ‘1-2-many’ system. The Munduruku have number words up to 5, but again the values are also imprecise. In tasks that require exact appreciation of numerosity, speakers of these languages look much like undergraduates in experiments where counting is suppressed. They tend to show accurate appreciation for numerosities for quantities up to 3 or 4, and then show increasing error in their estimations with increasing set sizes of stimuli. In fact, the relation between set size and error, known as the ‘coefficient of variation,’ is exactly the same in the Piraha˜ who cannot count under any circumstances and undergraduates in New Jersey when they are prevented from counting. On the other hand, if the undergrads had been given the ample opportunity to inspect the arrays without counting suppression, they would have been flawless. In one of the Piraha˜ studies, participants viewed an array of 4 objects for an extended period, but then failed to pick that array when later presented with a choice of 4 versus 5 objects to obtain a reward. Such studies suggest that conscious perception of numerical quantity differs radically from that of counting cultures. It is almost impossible for us to fail to perceive 4 or 5 as a quantity given enough time to encode the numerosity of the set (see Figure 2). For a culture that does not count, the distinction between 4 and 5 is not at all distinct. The failure to distinguish quantities of 4 and 5 when the language does not encode such a distinction is quite parallel to the case of failing to distinguish between phonemes such as /r/ and /l/ when the language does not make use of such a distinction. Other effects of language on concepts have been found in studies by Lera Boroditsky who
Number stimuli used with the Piraha˜. A candy is placed in a box with five fish, and then presented alongside a box with four fish. Participants were very poor at picking the correct box.
Figure 2
compared languages that differentially assign masculine or feminine gender to arbitrary nouns. For example, if one language assigns feminine gender to the word for table, and another language assigns masculine gender to this word, speakers of the two languages will show differential responses when asked to choose adjectives that are associated with the word. When table is feminine, speakers will associate more feminine descriptors with tables, such as soft, gentle, and warm. When the same word is assigned to masculine gender by the grammar, speakers associate the word with masculine descriptors such as strong, aggressive, and dominant. In English, since we refer to boats as ‘she’ we often find similar gender effects that have nothing to do with the biological gender of the objects themselves. However, whether this reflects a difference in the conscious experience of the objects, as opposed to merely producing covert conceptual associations is unclear.
General Conclusions Language emerges into consciousness primarily through the medium in which it is transmitted. Such media are richly varied reflecting the ingenuity of the human mind to develop channels for communication. The structure of the communication system that we refer to as natural language requires a vast supporting structure of symbolic structures, rules, and principles that allow us to communicate in conventional ways. Much of this structure is unconscious in nature. One way to understand the interface between conscious and
Language and Consciousness
unconscious linguistic processes is within the framework of modularity, which suggests that the linguistic processor acts in a primarily unconscious mode of symbol manipulation much like the inner workings of a digital computer. Where language does play a role in affecting conscious experiences is in focusing attention on aspects of sound, structure, and meaning. Experiments comparing languages that differ in how they use the available sounds for language, how they structure the
459
elements of language, and how they denote conceptual spaces of color, number, and other domains suggests that having a language can have profound effects on how speakers consciously perceive the sights, sounds and concepts in their mental construction of conscious reality.
See also: Inner Speech and Consciousness; Intentionality and Consciousness; Meta-Awareness.
Memory: Errors, Constructive Processes, and Conscious Retrieval W Hirstein, Elmhurst College, Elmhurst, IL, USA ã 2009 Elsevier Inc. All rights reserved.
This article focuses on aspects of the neuroscience and psychology of memory related to memory errors, the processes by which we form and reconstruct memories, as well as the conscious retrieval of memories. We examine two neurological syndromes in which patients report false memories, Korsakoff ’s syndrome and aneurysm of the anterior communicating artery, as well as false memory syndrome, which can affect normal children and adults. We are most interested in those aspects of memory that relate to consciousness, the place where memory and perception come together, an idea captured by Edelman’s apt phrase, ‘the remembered present.’ Thoughts held in consciousness naturally attract relevant memories, so that they also enter into consciousness. Once the mnemonic material is retrieved, however, it must be carefully checked for accuracy and actual relevance to the current situation.
Implicit and Explicit Memory The brain’s many memory systems can be divided into two main types: implicit memory systems and explicit memory systems. The primary difference between the two is their relation to consciousness: explicit memory systems deliver information to consciousness in the form of thoughts or images, whereas implicit memory largely bypasses consciousness. Procedural memory is an example of a type of implicit memory. Procedural memory allows us to acquire skills, such as playing the piano or playing a sport. It functions largely without consciousness, and in fact conscious awareness can interfere with its workings, as evidenced by the infamous nasty trick sometimes played on fellow golfers by asking them whether they inhale or exhale when they swing. When the victim attempts to discern the answer it can cause the intricate
pattern of muscle activations to disintegrate. Classical conditioning, of the type discovered by Pavlov, is also a form of implicit memory. We will focus here on a type of explicit memory known as episodic or autobiographical memory. Can you remember having breakfast this morning? You need to employ your ‘autobiographical memory’ in order to do this. Autobiographical memory is a rough record of our personal experiences, usually from our point of view. Most researchers agree that autobiographical memory is either the same as, or a subset of, episodic memory. The autobiographical record is fragmentary, whole hours, then days, weeks, and even years can fall into the abyss of forgetting. It is an especially personal form of memory, not only because it records most indelibly those things of importance to us but also because losing it means losing a sense of ourselves, as anyone who has ever watched someone succumb to Alzheimer’s can testify. The Alzheimer’s patient eventually forgets you and claims you are someone else, or a stranger, and this flags another function of the autobiographical memory system: it records information about other people, places, and things that are significant to us. With some of the things you know, the knowledge of when and where you first acquired that information is long gone. You know that cats have claws, but you most likely have no idea when you learned this. Other information does bring with it what researchers call a source memory: a type of episodic memory about when and where a memory was acquired. Source memory is a fragile thing, and we are prone to characteristic errors in source memory tasks. In one study, normal people and hospital patients with frontal cortical lesions learned the answers to a set of trivia questions. When they were tested a week later, the frontal patients had normal memory of the answers
1
2
Memory: Errors, Constructive Processes, and Conscious Retrieval
themselves, but showed poor source memory, often claiming they had learned the answer at some earlier point in life. The other type of explicit memory which does not necessarily come with a source tag is called semantic memory. It includes knowledge of facts, such as that the Eiffel Tower is in Paris, that Truman was a US president, and so on. Notice that these facts are represented from an impersonal perspective, whereas autobiographical memory is essentially personal. So far, researchers have been unable to cleanly separate the neural loci of semantic memory and episodic memory, and perhaps for good reason. The two memory systems interact in several ways, and some have suggested that they are merely different levels of categorization in the same memory store.
The Neuropsychology of Autobiographical Memory The medial temporal lobe memory system includes the hippocampus formation and the adjacent parahippocampal and perirhinal cortices. The hippocampus is not the place where the content of memories is stored, but rather appears to contain a set of neural links to the content, which is distributed widely throughout the cortex. Memories of an episode in one’s life typically contain information from more than one modality: vision, hearing, and even taste, touch, and smell. Each of these components is stored in a unimodal sensory area, for example, the visual components of an episodic memory are stored in the visual cortex in the occipital lobe, while the auditory components are stored in the auditory cortex in the temporal lobe. These distributed representations are linked to a central index in the hippocampus. When recent episodes are retrieved, the index is reactivated, causing activation to spread to each of the associated unimodal areas. This is more correct of recent episodes, however. Once a representation of an episode has been fully consolidated, activation can spread between the separate features themselves, so that hippocampal activation is no longer needed. We are also beginning to gain an understanding of the brain areas that comprise the frontal components of the medial temporal lobe memory
system. Medial temporal and hippocampal regions tend to be more involved in spatial context memory, while the frontocortical region, the diencephalon, and the temporal lobes are involved in temporal context memory. Much has also been learned about the neural bases of short-term memory systems located in the frontal lobes. Psychologists have had trouble determining whether there is one type of short-term memory, or several. The time span involved – exactly what ‘short’ means – is also not widely agreed upon. In the 1980s, however, neuroscientists began exploring a large area in the dorsolateral portion of the prefrontal lobes. This area seems to be responsible for a sort of memory that has been called ‘working memory’ – a concept that at least overlaps with the psychologist’s concept of short-term memory. The recent history of autobiographical memory research begins in the 1950s with the study of H.M., a man who developed severe amnesia after bilateral surgical removal of the medial temporal lobes (including most of the hippocampus, the parahippocampal gyrus, and the amygdala) in an effort to lessen the severity of his epilepsy. H.M. retained his basic intelligence and his personality but lost the ability to remember anything that happened to him after the operation. Researchers noticed, however, that H.M. could retain information for a short time, and could also acquire new motor skills such as mirror writing, solving puzzles, or tracing mazes, without knowing that he was doing so, a form of procedural memory.
Korsakoff’s Syndrome In 1887, Sergei Korsakoff observed mental problems in a group of alcoholic patients he was treating, including memory loss, anxiety, fear, depression, irritability, and confabulation. We have since learned that the syndrome is caused by a lack of vitamin B1, or thiamine, and not directly by alcohol itself. Korsakoff ’s syndrome can come on quickly, after an alcoholic coma, or it can progress slowly over many years. The syndrome occurs primarily in alcoholics, but may also occur in other cases where the patient’s digestive system fails to absorb B1, in conditions known as the malabsorbtion syndrome and regional enteritis, as well as with cancer of the stomach. The means by
Memory: Errors, Constructive Processes, and Conscious Retrieval
which alcoholism causes Korsakoff ’s is not fully understood, but alcohol is known to interfere with transport of thiamine in the gastrointestinal tract, and chronic liver disease can also affect the liver’s ability to store thiamine. Thiamine itself is important for proper brain metabolism, because chemicals derived from thiamine play a role in the synthesis of neurotransmitters, particularly acetylcholine, as well as gamma-amino-butyricacid (GABA). The memory loss in Korsakoff ’s is anterograde – the patients are unable to form new memories. Procedural memory is intact; for instance, the patients can still drive. Korsakoff ’s patients tend to underestimate the time they have spent in the hospital, as well as their own ages. Korsakoff himself successfully traced the memory reports of his patients to actual memories but found that the memories had been displaced in time by the patients. In the early phase of their illness the confabulations of Korsakoff ’s patients is typically an internally consistent account concerning the patient. The contest of this account is drawn fully or principally from the patient’s recollection of his actual experiences, including his thoughts in the past. Korsakoff said of one of his patients, Telling of a trip she had made to Finland before her illness. . . [the patient] mixed into her story her recollections of the Crimea, and so it turned out that in Finland people always eat lamb and the inhabitants are Tartars.
Anterior Communicating Artery Aneurysms Aneurysms of the anterior communicating artery (ACoA) – which distributes blood to portions of the ventromedial frontal lobes (including parts of the orbitofrontal lobes) and related structures, including the basal forebrain, the fornix, the septum, the anterior cingulate gyrus, and the corpus callosum – can also cause confabulation. The ACoA is a common site for brain aneurysms, which occur when the walls of a blood vessel are weakened by some insult, such as infection or degenerative illness. Often the vessel will rupture, causing a hemorrhage and destruction of the surrounding brain tissue. The ACoA forms the
3
anterior link of the Circle of Willis, which is a ring of arteries that spans the inner portions of the two hemispheres, and interconnects the two large anterior cerebral arteries. While the ACoA is small, it feeds a variety of brain areas and organs, and damage to it may also seriously affect blood flow in one or both of the anterior cerebral arteries. The important cognitive features of the classical ACoA syndrome are 1. Memory loss. Patients show anterograde amnesia as well as retrograde amnesia, often for a few years preceding the aneurysm. As in Korsakoff ’s, short-term memory appears to be intact. In tests of recognition memory, they can often correctly recognize, for example, famous people, at a normal level, but they can exhibit something called ‘pathological false recognition,’ that is, cases where they claim to recognize a stimulus they are actually seeing for the first time. 2. Changes in personality. Just as do Korsakoff ’s patients, ACoA patients have social interaction problems. Impulsivity, impatience, disinhibition, emotional lability, depression, problems in decision-making, and poor judgment in social situations have also been observed. 3. Executive deficits. These include perseveration, poor concept formation, problems with set shifting, reduced verbal fluency, and impairments in cognitive estimation. 4. Confabulation. Confabulation appears in the acute phase right after the aneurysm, and very often remains in the chronic phase. It can change from implausible ‘spontaneous’ confabulation in the acute phase to more plausible ‘provoked’ confabulation in the chronic phase. De Luca reports that questions such as ‘‘What did you do last night?’’ will usually produce confabulations from the patient. One reported, ‘‘We went to dinner in New York City with my brother who came in to visit us’’; something that had happened, but years before. When a patient was asked how he would respond if he was told his report was wrong, he said, ‘‘I’d be surprised ‘cause my experience, what I learn from my eyes and ears tell me differently. . ..I’d want some evidence.’’ The memory deficits caused by anterior communicating artery aneurysms and by lifelong drinking
4
Memory: Errors, Constructive Processes, and Conscious Retrieval
in Korsakoff ’s syndrome hold a special interest for memory researchers because such deficits indicate that there are also important frontal components to the memory system. The anatomists confirm the idea that the frontal and temporal lobes work together to achieve memory, by finding that the areas that constitute the medial temporal lobe memory system have strong, reciprocal connections to at least two frontal areas. The sites of lesion in Korsakoff ’s and anterior communicating artery syndrome are clearly different from those involved in medial temporal lobe amnesia. There are corresponding differences between the temporal and frontal memory patients, the most important being that medial temporal lobe patients do not confabulate, and will admit their poor memories and pursue compensatory strategies. Medial temporal lobe patients have been found to be less likely than normal people to produce false memories on tasks, specifically designed to elicit them (see below). Medial temporal lobe amnesics show much higher latencies in giving their answers and make many more selfcorrections than confabulating frontal memory patients in memory tasks, probably indicating intact executive processes struggling to correct degraded memories.
False Memories We all experience the memory malfunctions one sees in different types of amnesics. We sometimes remember what we intended to say or do, rather than what we actually said or did. We frequently displace events in time upon recalling them. And then there is the interesting error of mistaking events that were merely dreamed about for real events, and the rarer case of mistaking memory of real events for memory of dreamt events. Recent trends in memory research have strongly confirmed what memory researchers have always known, that is, memorizing something is not at all like recording it, and the act of remembering is not the mere replaying of this recording. Memory is a selective and reconstructive process, which can go wrong in any of several ways. The phrase ‘false memory’ is a bit of a contradiction in terms, of course, given that it is reasonable
to hold that something is not a memory of any sort if it is not correct, but its meaning is clear enough. False memories can easily be produced in children by asking them leading questions, and by a procedure such as the following: Children were presented with a deck of cards, each of which described an event. Some of the events had actually happened to the children, while others had not. When they were repeatedly asked whether the false events had happened to them, a majority of the children eventually agreed that they had, and many of them embellished the event with confabulated details. Apparently the memory systems themselves have a basic accuracy level; then we use different frontal checking procedures to increase this level. As we noted, normal correct memories are rational reconstructions, in that the reconstruction process is guided by what seems rational to the person. One can see this in certain patterns of error in false memories, where something odd in an event is misremembered as something more normal or rational. Young children exhibit some of the same patterns of memory problems that frontal patients show. This may be due to the fact that the frontal lobes are among the last cortical areas to mature. A large part of the development of the frontal lobes occurs between ages 5 and 10, and they do not fully mature until the teenage years. Perhaps nature’s plan is that the checking processes described above will be instilled after birth, during the long training period we humans require, principally by our parents. What begins as an external loop is made internal: The child confabulates, the parent corrects, the child changes what he said. As we mature, we internalize these corrections so that the loop runs completely within our brains, although it shares some of the same dynamics: there is still a candidate claim, and there is still a check that has the power to inhibit the claim from actually being made. Adults are also prone to false memories, however, given the right circumstances. The ‘misinformation effect’ is a way to induce false memories in adults in laboratory settings. In a typical such experiment, the subject will be shown a video depicting a staged crime, and then exposed to false information designed to interfere with his or her memories of the event. Subjects show a strong tendency to incorporate
Memory: Errors, Constructive Processes, and Conscious Retrieval
this information when asked later to recount the event. Loftus and her colleagues have also shown that exposure to prejudicial information after having witnessed an event can influence the subject’s later recall of that event. Garry’s research group has shown that both imagining events that never happened and paraphrasing descriptions of such events can make one more likely to later report that those events actually happened. In another type of experiment, normal subjects are presented with a list of words related to sleep, excluding the word ‘sleep’ itself: ‘bed,’ ‘rest,’ ‘awake,’ ‘tired,’ ‘dream,’ ‘wake,’ ‘snooze,’ etc. When they were later tested, between 30% and 40% of the subjects claimed that they had seen the word ‘sleep.’ Researchers observed the brains of normal subjects using positron emission tomography (PET) as they performed a task in which subjects first heard a list of related words, and then were tested for memory of the words. The researchers were able to successfully differentiate correct from incorrect memories by their different patterns of activation. Subjects of hypnosis may also confabulate when they are asked to recall information associated with crimes, causing researchers to warn criminologists about the dangers of obtaining information from hypnotized subjects. There are also anecdotal reports of hypnotized subjects confabulating when asked why they did something in accord with their hypnotic suggestion. For instance, a stage hypnotist gives his subject the suggestion that he will wave his hands whenever he hears the word ‘money.’ When asked later why is he is waving his hands, he replies ‘‘Oh, I just felt like stretching.’’
Theories of Confabulation Confabulation caused by frontal brain injury is the best-studied type of confabulation, and there are hints that some sort of frontal damage is essential for confabulation, so this is the best place to start our review of theories of confabulation. This damage produces confabulations about past events in the patient’s life, which either did not happen, did not happen to him, or did not happen to him when he believes they did. A man with Korsakoff ’s syndrome might claim that he was at work at his supermarket, finishing up the year-end inventory, for example, when he had been in bed at the hospital
5
the entire time. Stated in term of individually testable criteria, the recommended definition of ‘confabulation’ is as follows: S confabulates (in claiming that p) if and only if: (1) S claims that p; (2) S believes that p; (3) S’s thought that p is ill-grounded; (4) S does not know that her thought is ill-grounded; (5) S should know that her thought is ill-grounded; (6) S is confident that p. The concept of ‘claiming’ (rather than, for instance, ‘saying’ or ‘asserting’) is broad enough to cover a wide variety of responses by subjects and patients, including nonverbal responses, such as drawing and pointing. The second criterion captures the sincerity of confabulators. If explicitly asked, ‘‘Do you believe that p?,’’ they invariably answer yes. The third criterion refers to the problem that caused the flawed response to be generated: Processes within the relevant knowledge domain were not acting optimally. Criterion number four refers to a cognitive failure at a second phase, the failure to check and reject the flawed response. The fifth criterion captures a normative element in our concept of confabulation: If the confabulator’s brain were functioning properly, she would know that the claim is ill-grounded, and not make it. The claims made are about things any normal person would easily get right. The sixth and last criterion refers to another important characteristic of confabulators observed in the clinic, that is, the serene certainty they have in their claims, even in the face of obvious disbelief by their listeners. This epistemic approach eliminates a problem endemic to the falsity criterion in the traditional definition, according to which confabulations are false memory reports: A patient might answer correctly out of luck. The problem is not so much the falsity of the patients’ claims but rather their ill-groundedness and consequent unreliability, at least in the affected domain. In short then, in this epistemic view, to confabulate is to confidently make an ill-grounded claim that one should, but does not, know is ill-grounded. With the increasing information available about how our memory systems work, the discussion of memory-based confabulation has grown increasingly sophisticated. One theme of great interest that comes up frequently in the literature is the idea that these types of confabulations might be caused by two separate malfunctions. First, the
6
Memory: Errors, Constructive Processes, and Conscious Retrieval
patients have a memory problem, which they share with medial temporal lobe patients. But second, the patients have what is typically referred to as an executive problem, which is responsible for the failure to realize that the memories they are reporting are fictitious. In a particular case of confabulation, the two problems manifest as two ‘phases’: first a false memory is produced, but then frontal areas fail to perform functions that would allow the person to realize the falsity of the memory. Notice that this implies that the thoughts that give rise to confabulations exist as genuine beliefs in the patient’s mind (as opposed to the patient merely finding a certain claim coming out of his mouth, without his actually believing it). This implies that the patient’s confabulations are accurately reporting his (disordered or ill-grounded) conscious experience. Theories of the nature of the problem in memory confabulation are divided into two categories, depending on which of the two problems they emphasize: 1. Retrieval theories: According to these theorists, confabulation is caused by a deficit in the ‘strategic retrieval’ of memories. This causes a loss of the sense of the temporal order of one’s memories, and of their ‘sources’ – the places and times they represent. Theories of this type trace all the way back to Korsakoff. 2. Executive theories: These theories typically acknowledge that there is an amnesia present, but add that confabulators are to be differentiated by their additional frontal damage. According to these theories, two different processes are damaged in confabulation: (1) A memory process and (2) An executive or ‘monitoring’ process. The executive processes fail to correct the false memory. Cognition requires both representations and processes for manipulating those representations – these are executive processes. Executive processes perform many different operations on representations. Your memory itself is just a huge collection of representations; executive processes must control the search and reconstruction process that take place when we remember. Marcia Johnson’s view, according to which confabulation is attributed to a deficit in a more general executive function she calls ‘reality monitoring,’ – the ability to distinguish
real from imagined events – is an example of an executive theory. Normal people are able to differentiate real from spurious information at high success rates. This seems to be a learned, or at least a developed ability. Real memories, according to Johnson, can often be distinguished from mere imaginings by the amount of perceptual detail they contain, as well as by the presence of supporting memories, including information about where and when the remembered event occurred – source memory. The search for the neural locus and functional nature of these executive processes begins with what is already known about executive processes affected when the lateral portions of the prefrontal cortex are damaged. One line of inquiry involves determining how often confabulation tends to occur in the presence or absence of other mental functions known to require frontal activity. In the Wisconsin Card Sorting Test, the patient must first sort cards by the color shown on them, then the rule is changed so that the cards are to be sorted by the shape of the figures on them, etc. Patients with lateral frontal damage get ‘stuck’ responding the first way, and are unable to change to the new sorting rule, a phenomenon known as ‘perseveration.’ However, several confabulating patients have been found who perform normally on the standard tests of frontal function, such as the Wisconsin Card Sorting Test, causing speculation that confabulation and failure on such tests are the result of damage to different frontal areas. It may just be that retrieval theories and executive theories are merely directed at different parts of the confabulation process. The first phase involves the production of a false memory. The second phase involves failure to notice and correct the falsity. Retrieval theories focus on the failure to access the correct memory, while executive theories focus on the correction failure. The executive theorists typically attribute confabulation to a failure in what they call self-monitoring, or self-awareness.
Reality Monitoring Confabulation may be due to a broader failure to test representations, whether they are from memory or not. The term ‘reality monitoring’ seems
Memory: Errors, Constructive Processes, and Conscious Retrieval
misleading, since what is being monitored (or not) is a representation, often one to which no reality corresponds – ‘representation monitoring’ would be more accurate. According to Johnson, episodic memories of an event bind together elements of several different types, some of which represent properties of the event, while others represent features of us, for example, our thoughts or emotions in reaction to witnessing the event. These different properties include the following: colors, sounds, tastes, emotions, objects, and locations, and also information contained in semantic memory. Recall of any one of these features is often enough to draw the entire autobiographical memory back into awareness. When thoughts that have this rich detail present themselves as memories, this can be sufficient to make us regard them as genuine. Because of this, in the mind of a person with a vivid and detailed imagination, memories of imaginings can be mistaken for memories of actual events. According to the proponents of this approach, vivid imaginings can be mistaken for memories, for example, when one believes they said or did something that they only thought or imagined saying or doing. Further checks can be made: We can check the consistency of the candidate memory with our set of beliefs. Monitoring can involve the noting of inconsistencies among representations currently in consciousness or between those and long-term knowledge. Confabulation patients tend not to notice or worry when they contradict themselves. One patient, for example, contradicted himself in the same sentence, saying that he had just visited a store he formerly owned, and then acknowledging that the store no longer exists. As early as 1915, Pick noted that Korsakoff ’s patients also have a deficiency in the need to correct contradictions. People can intentionally tighten their monitoring standards when motivated to do so. Researchers often report that simply admonishing memory patients to be more careful can work to increase the accuracy level of their reported memories. It is interesting to note that we tend not to voluntarily loosen our standards; rather, it tends to happen unconsciously and spontaneously. Johnson and her colleagues distinguish between ‘heuristic’ checking of candidate memories, which is usually operating automatically when we are remembering,
7
and ‘systematic’ checking, which is intentional. Heuristic processing consists of fewer component processes and uses readily available information, such as familiarity, perceptual detail, and schemas (e.g., world knowledge, stereotypes), typically activated by a cue. Systematic processing is made up of more component processes and may also involve the retrieval of other memories and knowledge not initially activated. Systematic processing requires selective attention: the person must explicitly attend to the candidate memory. It also includes self-provided memory cues. We often cue our own memories: When I want to remember someone’s name, I imagine her face. I produce a cue for my memory system to use in retrieving the name. I then monitor any representations that the cue gives rise to. I may need to use other information to reject candidate names that come up. Often this cuing process must be used several times in order to reconstruct the memory correctly. As to the neural locus of these monitoring processes, researchers point to bifrontal areas.
The Suppression of Irrelevant Memories Armin Schnider’s research group hypothesizes similarly that the problem in memory confabulation is that the orbitofrontal cortex and its limbic connections are not performing their function of suppressing or inhibiting recalled memories that are not relevant to the current task. They argue that the posterior medial orbitofrontal cortex sorts out the mental associations that pertain to ongoing reality by suppressing memory traces that have no current relevance. As to the question of what the crucial lesion is for producing confabulation, Schnider points to an orbitofrontal-mediodorsalamygdala circuit, claiming that spontaneous confabulation appears to emanate from interruption of the loop connecting the posterior orbitofrontal cortex directly (via ventral amygdalafugal pathways) and indirectly (via the mediodorsal thalamus) with the amygdala. Schnider’s way of connecting confabulation in Korsakoff ’s with that found in ACoA patients is to point out that the basal forebrain lesions one sees in ACoA
8
Memory: Errors, Constructive Processes, and Conscious Retrieval
aneurysms often include the posterior medial orbitofrontal cortex. Schnider’s localization is supported by two of his findings. First, patients with lesions involving the posterior medial orbitofrontal cortex and basal forebrain confabulate for much longer periods (several months) than patients with anterior medial orbitofrontal lesions. Schnider’s group also found posterior medial orbitofrontal cortex activation in normal subjects who performed a memory task requiring that they carefully separate relevant from similar but irrelevant memories. Some frontal theories of confabulation can be expanded to fit other confabulation syndromes not involving memory problems, but others are so strongly memory based that they cannot be expanded. Theories that attribute confabulation specifically to a memory retrieval problem only apply to memory patients: Korsakoff, for instance, argued that confabulations are simply veridical memories placed out of context. Schnider et al. conclude similarly that spontaneous confabulations arise due to a failure to know the proper temporal place of the recalled memory. Alternatively, many of the executive theories have the potential to be expanded beyond mnemonic confabulation, such as Johnson’s view.
Domain Specificity The question of the domain specificity of confabulation becomes important in assessing and improving the different theories of confabulation. Several cases have now been reported of patients whose confabulation was restricted to autobiographical memory, or who at least confabulated mainly about autobiographical memory. But there are exceptions, for instance, one patient showed spontaneous confabulation across episodic, personal, and general semantic memory. Another patient with frontal damage who only showed confabulation in ‘personal’ semantic memory – memory of the important facts of one’s life. While retrieval theories are supported by confabulation restricted to certain memory domains, executive or reality monitoring theories have trouble with domain-specific confabulation, because the general understanding of executive processes is that they can be applied in more than one
domain. Executive processes may have their own domain specificities, however. If there are several different checking or monitoring processes, this predicts that there are several different types of confabulating patients, depending on which processes are damaged. This also implies that several different neurological tests need to be used to assess each of the types. Dalla Barba and his colleagues developed a confabulation battery that contains questions touching on personal semantic memory (such as one’s address, age, or occupation), questions relating to autobiographical memory, questions about orientation in time and place, and questions touching on general semantic memory, such as historical facts. There are also several extremely difficult questions in the battery (e.g., ‘‘Who was Marilyn Monroe’s father?’’), to which normal subjects would be likely to answer ‘‘I don’t know,’’ on the assumption that confabulators will fail to know that they do not know.
Separating Amnesia and Confabulation A large body of evidence has accumulated showing dissociations between problems with temporal memory and confabulation. The two phenomena also show different time courses: The confabulation in Korsakoff ’s often clears while the amnesia remains. Similarly, confabulation clears in most ACoA patients after weeks or months, but the memory problems do not improve. Sometimes it works in the other direction, memory improves, while confabulation continues. Two phenomena militate against the idea that amnesia alone in Korsakoff ’s patients can explain the confabulation: First, confabulation tends to disappear as patients progress from the confusional stage to the chronic phase, while the amnesia remains. Second, some chronic amnesics patients confabulate a great deal, others less or not at all; and there is no correlation between their propensity to confabulate and the gravity of their amnesia. If memory confabulation results from two independent lesions, this indicates that there are two types of patients: 1. Patients who sustained the memory system lesion first. This is a patient who should admit
Memory: Errors, Constructive Processes, and Conscious Retrieval
his memory problem until the executive problem develops, at which point he should deny it and commence confabulating. 2. Patients who sustained the executive lesion first. The course of this disease might be rather subtle. We also need to leave open the possibility that there are some people who simply do not develop the executive processes needed to check memory reports, and make do with their memories alone and tolerate a high rate of errors. This opens up the interesting possibility that the problem with some Korsakoff ’s patients is that they are confabulatory before losing their memory. They have lost the ability to check thoughts or candidate memories. It may pass unnoticed because the patient is substantially correct in what he says. But once the amnesia sets in, the problem becomes painfully obvious.
Episodic Memories of Others A person’s episodic memories are memories of that person from her point of view. The representations that are stored in episodic memory are egocentric in the sense that they represent events as we experienced them, hence they are also called autobiographical memories. Autobiographical episodic memories combine several different subrepresentations including representations of our bodies moving through different spaces and environments and representations of the people and objects we have significant interactions with. We carefully represent each aspect of a particularly significant interaction, exactly what was said, and in which tone of voice. In addition to representations of emotions, episodic memories may also contain representations of other conscious states, such as our thoughts, motives, and intentions at the time of the event. The typical autobiographical memory representation is of a person with a conscious mind, moving through space, and interacting with people and objects. The episodic memory system is also able to aggregate its information into our existing concepts of important people and things. Once this information enters the system of concepts it becomes part of the semantic memory system also, and is then
9
accessible to the process of thinking itself. Thus the episodic memory system can feed the semantic memory system. For instance, if I travel to Paris, the episodic memories I amass as I see important sites in the city also add information to my semantic representation of Paris, the Eiffel Tower, the Arch of Triumph, etc. All of this information tends to be either conceptual or allocentric in form, but I suggest that the egocentric realm has its own ability to aggregate its representations into a full-blown simulation of persons’ minds. As I accumulate information about someone over the course of many interactions with her, I also accumulate information about her thoughts, moods, and emotions, using the egocentric representation system in other mode, to represent her. We might call such memories, ‘biographical memories.’ These accumulations of simulated mental states constitute representations of the minds of the significant people in our lives. If this is right, our representation of a significant person would contain an allocentric component with representation of how he looks and sounds, and an egocentric representation of his mind, body, and environments from his viewpoint. Our awareness of the minds of others when the egocentric system is representing them is nonexplicit and faint in our minds. Perhaps this is a reason why distinguishing whether an emotion or a simulation of a mind is missing is difficult for the patients. Perhaps another reason why the activity level of the egocentric system needs to be low when it is operating in the other mode is because full-blown conscious representation leads to actual external actions, not merely to represented actions, in much the same way that the dreaming mind malfunctions during REM sleep disorder, when dreamed perceptions cause real actions. Simulations cannot fully employ this representation system without danger of causing real actions, with potentially disastrous consequences. The experimental and clinical data show that this biographical memory system may have a special fragility in the way it sorts incoming memories according to persons’ identities. The system needs to be able to create new concepts when new people are encountered, and to properly segregate this concept from its concepts of other people. For instance, if I start a new job and spend several hours with a new coworker, my brain will begin
10
Memory: Errors, Constructive Processes, and Conscious Retrieval
creating a representation of this person. But when this person leaves and another new coworker enters, my brain needs to close the first file it opened and start another one. There needs to be a firm partition between these two representations, so that I do not confuse the identities of my coworkers.
Conclusion The many types of different memory systems that the brain’s evolutionary development has created testifies to the value these forces of design place on learning from the past. Several questions remain: Do memory confabulations belong to the larger set of completion phenomena, such as the filling-in of the visual blind spot? The executive processes located in the prefrontal lobes require clear, unambiguous information in order to achieve their primary task, that is, the creation of effective actions. We typically do not have the time to spend examining gaps in our perceptions and memories. Very often in real life, when memories occur, we make a quick plausibility check, sort out any obvious contradictions or impossibilities in the memory, and move forward with the belief that the memory is correct. Another interesting question is whether memory confabulations might involve a type of self-deception. One piece of evidence for this is the way that so many of the recalled memories present the patient in a rosy light. In addition, the idea that merely imagining or thinking about an event can cause us to later confabulate that the event happened may be revealing one of the mechanisms behind self-deception to us: thinking that something is so, usually something positive for us, can work to convince us that it really is so. On the one hand this removes some of the blame for selfdeception from us, since it shows that unconscious processes not normally under our control can produce a plausibility effect. On the other, it shows the power that conscious rehearsal of information can have on our rather fragile and complex memory systems. See also: The Control of Mnemonic Awareness; Autobiographical Memory and Consciousness; Consciousness and Memory in Amnesia.
Suggested Readings DeLuca J (2001) A cognitive neuroscience perspective on confabulation. Neuro-Psychoanalysis 2: 119–132. DeLuca J and Diamond BJ (1995) Aneurysm of the anterior communicating artery: A review of the neuroanatomical and neuropsychological sequelae. Journal of Clinical and Experimental Neuropsychology 17: 100–121. Fischer RS, Alexander MP, D’Esposito MD, and Otto R (1995) Neuropsychological and neuroanatomical correlates of confabulation. Journal of Clinical and Experimental Neuropsychology 17: 20–28. Fuster JM (1995) Memory in the Cerebral Cortex. Cambridge, MA: MIT Press. Garry M, Manning CG, Loftus EF, and Sherman SJ (1996) Imagination inflation: Imagining a childhood event inflates confidence that it occurred. Psychonomic Bulletin and Review 12: 359–366. Hirstein W (2005) Brain Fiction: Self-Deception and the Riddle of Confabulation. Cambridge, MA: MIT Press. Johnson MK, Hayes SM, D’Esposito MD, and Raye CL (2000) Confabulation. In: Grafman J and Boller F (eds.) Handbook of Neuropsychology. New York: Elsevier. Kopelman MD (1987) Two types of confabulation. Journal of Neurology, Neurosurgery, and Psychiatry 50: 1482–1487. Korsakoff SS (1889) Psychic disturbance in conjunction with peripheral neuritis, Victor M and Yakovlev PI (trans.). Neurology 5(1955): 394–406. Loftus EF (1991) Made in memory: Distortions in recollection after misleading information. In: Bower GH (ed.) The Psychology of Learning and Motivation: Advances in Research and Theory. San Diego: Academic Press. Moscovitch M (1995) Confabulation. In: Schacter DL (ed.) Memory Distortions. Cambridge, MA: Harvard University Press. Parkin AJ and Leng NRC (1993) Neuropsychology of the Amnesic Syndrome. Hove: Lawrence Erlbaum Associates. Schacter DL, Reiman E, Curran T, Yun LS, Bandy D, McDermott KB, and Roediger HL (1996) Neuroanatomical correlates of veridical and illusory recognition memory: Evidence from positron emission tomography. Neuron 17: 267–274. Schnider A (2001) Spontaneous confabulation, reality monitoring, and the limbic system: A review. Brain Research Reviews 36: 150–160. Scoville WB and Milner B (1957) Loss of recent memory after bilateral hippocampal lesions. Journal of Neurology, Neurosurgery, and Psychiatry 20: 11–21. Stuss DT, Alexander MP, Lieberman A, and Levine H (1978) An extraordinary form of confabulation. Neurology 28: 1166–1172. Talland GA (1965) Deranged Memory. New York: Academic Press.
Memory: Errors, Constructive Processes, and Conscious Retrieval
11
Biographical Sketch
William Hirstein is professor and chair of the Philosophy Department at Elmhurst College, in Elmhurst, Illinois, USA. He received his PhD from the University of California, Davis, in 1994. His graduate and postdoctoral studies were conducted under the supervision of John Searle, V. S. Ramachandran, and Patricia Churchland. He is the author of several books, including On the Churchlands (Wadsworth, 2004), and Brain Fiction: Self-Deception and the Riddle of Confabulation (MIT, 2005).
Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness R Custers and H Veling, Utrecht University, Utrecht, The Netherlands ã 2009 Elsevier Inc. All rights reserved.
Glossary Closed-loop process – Process that uses information about its outcomes as input. Declarative memory – Memory for knowledge about facts. Episodic memory – Autobiographical memory, part of declarative memory. Open-loop process – Process that does not use information about its outcomes as input. Perceptual-motor learning – Learning of motor skills that rely on perceptual input. Procedural memory – Memory for knowledge about how actions are executed. Schema – General knowledge structure in memory about a particular type of action. Semantic memory – Memory for meaning, part of declarative memory. Skill – Overlearned behavioral routine resulting from practice.
Introduction Anybody who has ever learned to ride a bicycle, or play the piano, will admit that mastering such skills requires lots of practice. And when asked how we acquire those skills, that is about all we have to say about it. We may be able to describe the principles by which a bike functions, or the structure of chords, but unlike such declarative knowledge, the knowledge about how our actions are executed – that is, procedural knowledge – that we acquire during skill development is not open to introspection. This seems even more remarkable if one considers the challenges that are faced in such learning processes. We never make the exact same bike ride twice, or play the same arpeggio on the piano, yet somehow we are able to extract general knowledge about the motor actions involved, store it in
memory, learn to plan and tune those actions based on perceived feedback, and learn how they can be combined into complex sequences of actions without much awareness of what is going on.
Perceptual-Motor Learning The problems with storing an action in memory already become apparent if we look at the simple act of, say, grasping a coffee mug. In any instance, we have never executed the exact required action before (the location of the mug and its handle are different) and even more problematic, we can execute the action in an infinite number of ways. One can flex and extend the different joints that are orchestrated by many different muscles in an infinite number of ways with the same result. We are, for example, still able to grasp the mug when holding a phone between our ear and the shoulder of the grasping arm and get the job done in an awkward, but effective, way. This degrees-of-freedom problem suggests that knowledge about actions cannot be stored in terms of exact knowledge of muscle tension and joint positions, but has to be represented in a more general way. It turns out that what is stored in memory is the general pattern or schema that captures the essential structure of the action, with specific parameters to be filled in during the planning and execution of the action. This requires integration of knowledge from memory and information resulting from perception. The crucial role of visual feedback in the execution of actions has, for example, been demonstrated in studies in which perception was blocked by turning off the lights once participants had planned and started to execute the movement. Except for extremely quick, reflex-like actions – which are more ballistic in nature – performance significantly deteriorated without visual feedback. This demonstrates that even performance of a simple action is 13
14
Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness
not an open-loop process, in which the action is only planned and then blindly executed, but a closedloop process that requires integration of perceptual as well as motor information. When it comes to executing skilled behaviors, it is especially this tuning of action, based on closedloop processes, that we are mostly unaware of. We may consciously initiate a turn with our bike, but the processes by which we maintain our balance and stabilize our course in the new direction operate largely outside of awareness. This lack of consciously accessible knowledge about tuning becomes apparent in the following phenomenon. When you ask participants in the laboratory to demonstrate the steering movements that they usually make to change lanes with their car, most people turn the wheel in the required direction, and then turn it back to the starting position. In reality, it takes an equal turn in the opposite direction after changing the lane to keep the car on the road. Apparently, although people are consciously aware of the movement they must make to initiate action, they lack conscious knowledge of the subsequent compensating movement that relies on tuning in response to visual feedback.
Sequence Learning In addition to this lack of conscious knowledge of tuning, people are often oblivious to how sequences of actions are learned. Researchers have since long been intrigued by the question of how rapid sequences of actions in playing piano, or typing, can be learned and executed, especially because there is no time to receive the feedback of one action before performing the next. Although it was first believed that such sequences were stored in memory as action chains, in which each specific action that was retrieved from memory would trigger retrieval and execution of the next associated action, by now it has become clear that people do not so much store the chain of the exact responses, but rather a more abstract, higher-level pattern. This was revealed by studies showing that execution of a specific action-pattern benefited from an earlier learning phase, even when the exact actions during that learning phase (pressing a series of buttons with one’s hand) were executed by a different means
in the test phase (e.g., pressing the same series of buttons with one’s elbow). Learning of such action sequences is one of the pillars of skill development and has been shown to occur even without conscious awareness. In classic studies on implicit sequence learning, participants responded with different actions to different stimuli that occurred in a fixed, but rather complex pattern. Even though participants did not consciously detect any pattern, they speeded up over time and performance significantly deteriorated when the stimuli were no longer presented according to the fixed pattern.
Organization of Skills in Memory But even when such sequences are learned there is room for improvement in execution. One of the characteristics of skills is that they keep improving, although ever less dramatically, over time. An important mechanism that contributes to this effect may be more efficient storage of action patterns in memory. Typing lessons capitalize on this knowledge by having people repeat often occurring patterns of keys (e.g., ‘T-H-E’) over and over again. Although it is clear to the student what has to be learned, the pattern of movements has to be ‘stamped in’ through practice. This strategy promotes the grouping of specific actions in memory into chunks, which also happens over time without specific practice. These chunks can then be retrieved from memory and executed as one unit.
Brain Processes, Skill-Learning, and Awareness Studying skill acquisition in people with specific brain lesions, as well as modern neuroimaging methods such as positron emission tomography (PET) and functional magnetic resonance imaging (f MRI) have made it possible to gain knowledge about the areas of the brain that are involved during several stages of skill development. One striking finding is that brain structures that are crucial in memory for events have little to do with skill acquisition. Studies with patients suffering from anterograde amnesia have demonstrated that although
Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness
those patients are unable to form new memories about events, they develop many skills just as well as control participants. It has been demonstrated that this skill-learning goes even further than perceptual-motor skills. In one study, amnesic and control participants practiced reading of words that were presented in mirror image in a series of learning blocks that extended over a 2-week period. Some specific words were repeated from block to block, whereas the rest of the words were always novel in each block. Although the reading time per word for the repeated words dropped faster for control participants than amnesiacs, mirrorreading skills for novel words were shown to improve equally fast for both groups. Hence, although control participants clearly benefited from the memory of having seen the specific repeated words (which they could consciously recall) in previous blocks, amnesiacs developed the same mirror-reading skill without any recollection of those words. These findings demonstrate that procedural and declarative memories are supported by distinct brain areas. In declarative memory, the hippocampus has been found to play a crucial role. Bilateral damage to the hippocampus and parahippocampal regions indeed causes retrograde amnesia, with loss of episodic memory and semantic memory. These structures also play a crucial role in certain types of learning, such as trace conditioning, where there is a relatively long time lag between a neutral conditioned stimulus (CS) (e.g., a tone) that predicts an unconditioned stimulus (US) (e.g., a puff of air in the eye). Although after conditioning in such a task animals with an intact hippocampus react with an eye blink to the CS, such learning does not occur when the hippocampus is damaged. The same effect has been demonstrated in humans, where this type of learning is accompanied by activation of the hippocampus and usually with conscious awareness of the relation between the CS and the US. In sum, the hippocampus plays a crucial role in declarative memory, but also in particular forms of learning, where integration of separate events in memory is required. The brain areas involved in procedural learning, and in particular the learning of motor skills, change as learning of the skill progresses. In the early stage of motor-skill learning, the prefrontal cortex (PFC),
15
anterior cingulated cortex (ACC), and posterior parietal cortex (PPC) direct attentional and control processes that are necessary for grasping the basics of the task and the planning of deliberate, intentional responses. As skills start to form, the cerebellum plays a crucial role in creating ‘mental shortcuts’ between perception and actions. It integrates afferent signals coming in from the sensory systems with efferent signals that produce motor actions. As such, it links together new and already acquired motorprograms, forming more complex response patterns and assigning them to particular perceptual patterns. Activation of the cerebellum is typically found in the early stages of skill-learning and shifts from the cerebellar cortex to the nuclei as learning advances, until it almost ceases when a basic skill is acquired. In this stage, the striatum has been found to play an important role in the detection of errors that are produced. When erroneous responses occur, the inappropriate programs are inhibited, a process in which the PFC is again thought to play an important role. As a result, connections between perceptual input and the appropriate motor output are further strengthened. Thus, the skill is further polished until it is executed almost flawlessly. If this skill is then used frequently over a longer period, the skill becomes overlearned, which will finally lead to changes in the motor areas that play a role in the execution of the specific motor actions. Such cortical plasticity has been demonstrated in, among others, musicians and professional sportsmen, in whom the mapping of crucial motor programs has been reorganized in service of their skills. As such, overlearned skills become more and more hardwired in people’s brains, and conscious attention or awareness are no longer required to conduct them.
Acquired Skills, Goals, and Conscious Control Although the execution of a skill may eventually occur without conscious awareness, skills are often executed in the service of a conscious goal. Grasping a mug and bringing it to one’s mouth, for instance, usually serves to take a sip of the fluid contained by the mug (e.g., coffee). Because acquired
16
Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness
skills are stored as abstract high-level patterns, skill execution such as grasping the mug can be elicited by merely perceiving a skill-relevant cue (e.g., a coffee mug) under many different circumstances (e.g., under variations in distance toward the mug, or size of the mug) once a goal is set. As a result of overlearning, little or no conscious control is needed to execute the individual action sequences that capture the essential structure of the action (e.g., when and how much a hand should be opened when reaching for the mug). In fact, it has been shown that focusing conscious attention on the execution of specific components of a complex motor skill can impair performance. For instance, experienced golfers are better in putting a ball when they are distracted by a secondary task (e.g., monitoring whether a specific tone sounds through a headphone) than when they are instructed to focus attention on their swing. In a similar vein, experienced soccer players, but not novices, handle the ball better with their dominant foot when they are distracted from executing a skill (e.g., dribbling) than when they consciously focus on specific components of the actions that they are executing. An explanation for this effect is that by attending to separate components one overrules the more efficient organizational structure of the skill, causing the building blocks of the skill to function as separate components, in pretty much the same way as before the skill was acquired. Once the organizational structure breaks down, each component is executed separately, which costs more time, and leaves more room for error, than when the abstract high-level pattern is used to execute the separate components. It has been proposed that conscious attention to the step-by-step components of motor skills can lead to choking under pressure, for instance when a soccer player misses a penalty kick. Conscious control can, however, be beneficial if skills are not yet overlearned. It can, for example, improve performance in experienced soccer players when they handle the ball with their nondominant foot, or in novices in general, indicating that conscious attention toward action components can be helpful in practicing, or tuning, an underdeveloped skill. Although conscious control can be detrimental to performance of complex goal-directed skills, conscious control may be experienced over the overall action if one reflects on one’s behaviors, as
long as the outcome of the action (e.g., the golf ball rolls in the hole) is congruent with the goal that started the execution of the skill (e.g., putting the ball in the hole). Thus, even though awareness – and conscious control – of specific components of overlearned skills can be very restricted, conscious control can nevertheless be experienced when an intended outcome is reached. It appears that what is needed for an experience of control is that the outcome of an action is congruent with the mental representation of an intended outcome, irrespective of whether one actually consciously controls each and every step of the action sequence. The lack of conscious control over our skills, however, becomes apparent if our skills produce other outcomes than we consciously intended. As noted above, skill execution can be easily elicited by perceiving a skill-relevant cue when one has a goal in mind. If you want to use your computer, for instance, perceiving the log-in screen may trigger the responses required to log-in (e.g., typing in the correct password) without much conscious deliberation, if this procedure has been performed a great number of times before. Although this efficiency has a great advantage, as it relieves consciousness from tedious tasks such as retrieving a password from declarative memory, it also comes with a cost: It can be hard to prevent skill execution when a goal requires a new response that is different from an overlearned behavioral routine. For instance, when the password of your computer is changed, it is possible that the old password is entered erroneously the first couple of times that you start the computer. This example illustrates that an overlearned behavior (typing in the old password), which is no longer instrumental in attaining the goal (starting your computer), is still easily triggered by the environment. When there is a discrepancy between the outcomes produced by our skills and our intended outcome (starting the computer), we become aware of the situation, and conscious control can be recruited to choose a new path of action in order to reach the intended outcome (e.g., consciously retrieving the new password from declarative memory). Of course, with repeated practice, conscious control will again become less necessary and perceiving the log-in screen will be sufficient to trigger the adjusted skill.
Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness
Conclusion
17
Acknowledgment
From perceptual-motor to sequence learning, procedural knowledge is acquired as skills develop, although most of the time this knowledge is not accessible to consciousness. Acquiring a skill takes a lot of practice, during which different brain processes are involved in creating stable mental maps of the skill. Although it may seem to be a disadvantage in that there is no other way of acquiring this knowledge than through practice, the major benefit of procedural memory over declarative memory (which relies on different brain areas) is that it is highly stable. Once one has learned to ride a bike one never forgets. Although sometimes these skills may produce different outcomes than consciously intended, on the whole they serve us well, by freeing up precious capacity for other conscious processes. This process is so effective that conscious attention to skills may even deteriorate performance. Thus, skills make it possible to behave very effectively without much conscious interventions. In this light, the lack of introspection in how we acquire them may be a small price to pay.
The preparation and writing of this article was financially supported by the Netherlands Organization for Scientific Research (VENI grant 45106-014, VICI grant 453-06-002). See also: Habit, Action, and Consciousness; Perception, Action, and Consciousness.
Suggested Readings Anderson JR, Bothell D, Byrne MD, Douglass S, Lebiere C, and Qin Y (2004) An integrated theory of the mind. Psychological Review 111: 1036. Cooper R and Shallice T (2000) Contention scheduling and the control of routine activities. Cognitive Neuropsychology 17: 297–338. Jeannerod M (1997) The Cognitive Neuroscience of Action. Oxford, England: Blackwell. Lewicki P, Hill T, and Czyzewska M (1992) Nonconscious acquisition of information. American Psychologist 47: 796. Powers WT (1973) Feedback: Beyond behaviorism. Science 179: 351–356. Squire LR, Knowlton B, and Musen G (1993) The structure and organization of memory. Annual Review of Psychology 44: 453–495.
Biographical Sketch
Ruud Custers received an education in human–technology interaction at Eindhoven University of Technology, where he graduated cum laude on work investigating the role of memory in the formation of judgments about environments. Subsequently, he moved to Utrecht University to pursue his PhD in experimental social psychology. He received his PhD cum laude in 2006 for his dissertation on the underlying mechanisms on nonconscious goal pursuit, which mainly focused on the role of affective signals in this process. He published several papers in fundamental journals, of which one was regarded as the best paper of the year on social cognition by the International Social Cognition Network in 2006. As an assistant professor at Utrecht University, he continues to study the processes that allow people to pursue goals without conscious awareness.
18
Memory: Procedural Memory, Skill, Perceptual-Motor Learning, and Awareness
Harm Veling is trained as an experimental social psychologist at Radboud University Nijmegen where he worked on intention memory and inhibitory control. He received his PhD in 2007 for his dissertation on the inhibitory processes that facilitate execution of previously formed intentions. Next, he continued to work as a postdoc at Utrecht University. His work deals with several topics related to the role of goals and actions in automatic processes of social cognition, with an emphasis on inhibitory processes, and is published in several fundamental journals. One recent discovery in his research concerns the notion that stopping an action that is initially triggered by a rewarding stimulus results in devaluation of the rewarding stimulus. This chain of events suggests a functional behavior regulatory dynamic. He continues to study this intriguing and important topic.
Mental Representation and Consciousness D D Hutto, University of Hertfordshire, Hatfield, UK ã 2009 Elsevier Inc. All rights reserved.
Glossary Enactivism – The view that experiences are not inner events but are constituted by the activity of organisms engaging or interacting with some environment or other. Mode of presentation – The manner or aspect under which an individual, object or state of affairs is apprehended by a thinker or subject. Modes of presentation are thought to be the basis of cognitive mediation. Narrow content – A kind of content that is nonrelational and dependent only on the local, portable properties of psychological individuals. Phenomenal character – The distinctive quality associated with token experiences (perceptions, sensations, feelings, moods); what-it-is-like to undergo such experiences. Phenomenal consciousness – Subjective states of mind that involve having experiences with specific phenomenal characters. Strong representationalism – The claim that the phenomenal character of experience is exhausted by, identical with, or entirely determined by representational content or properties. Subjectivity – The property of being an experience for something or of someone, often equated with the having of an idiosyncratic first-personal point of view, or perspective. Supervenience – A nonreductive relation of covariance holding between specified relata (concepts, properties, entities, etc.). Supervenience relations are generally believed to involve asymmetric dependence or determination. Accordingly, if ‘x’ and ‘y’ properties exist, such that ‘y’ depends on or is determined by ‘x’ then any relevant change in ‘x’ also incurs a change in ‘y.’ For example, aesthetic properties are thought by some to depend on physical properties, such that
modifying an object’s material aspects in relevant ways would automatically modify its aesthetic aspects, that is, enhancing or marring its beauty. Supervenience relations can vary in their modal force (as reflected by operators such as ‘necessity’ and ‘possibility’) and scope. Weak representationalism – The claim that experiential facts are, or supervene on, representational facts. Wide (or broad) content – Content that is necessarily individuated by environmental factors, or kinds that exist outside the bounds of subjects or organisms.
Introduction Intentionality and consciousness are the fundamental kinds of mental phenomena. Although they are widely regarded as being entirely distinct some philosophers conjecture that they are intimately related. Prominently it has been claimed that consciousness can be best understood in terms of representational facts or properties. Representationalist theories vary in strength. At their core they seek to establish that subjective, phenomenal consciousness (of the kind that involves the having of first-personal points of view or perspectives on the world – perspectives that incorporate experiences with specific phenomenal characters) is either exhausted by, or supervenes on, capacities for mental representation. These proposals face several serious objections.
Intentionality and Mental Representation Intentionality and consciousness are the most fundamental and philosophically interesting kinds of 19
20
Mental Representation and Consciousness
mental phenomena. Explaining how they fit into the world order poses powerful (and some think insuperable) challenges for naturalists who rely exclusively on the resources of hard sciences for prosecuting that task. Intentionality is the capacity to have thoughts, feelings, and other states of mind that are about, or directed at, particular worldly offerings and states of affairs. Examples include having the thought that ‘‘It is going to rain’’; harboring a desire for strong coffee, or seeing that the leaves have changed color. Building on the foundations laid down by early modern thinkers, some contemporary philosophers hold that the capacity to be in such states of mind is best explained by the having of inner mental representations with specific truth-apt contents. At the very least, it is accepted that mental representations form a subclass of intentional phenomena. Mental representations are distinct from ordinary representations. The latter are familiar public items of our everyday acquaintance. They include such things as drawings, maps, and natural language statements. Mental representations are modeled in weaker or stronger ways on these pedestrian entities and are thought to have similar properties and functions. They aim to say how things stand with the world. If they speak truly they can inform their users of the state of the world, successfully guiding their behavior and actions (or at least they can do this when they combine appropriately with conative states). Minimally, something is a mental representation if it presents some portion of the world as being a certain way; for example as ‘being hot’ or ‘being colored.’ The world may of course vary from the way it is presented as being: things are not always as they seem. Importantly, mental representations can be entirely nonconceptual and still be regarded as having content. A mental state will have content and be representational in character if it has correctness conditions that are specifiable, in principle. Conceptual and nonconceptual content are thus thought to be equally representational in nature; they are distinguished only in the way that they represent. Representing aspects of the world necessarily invokes correctness conditions. Although some philosophers have argued that certain types of
representations are immune to error – such as the necessary truths of logic or mathematics – the existence and status of such truths, and whether they take the form of genuine propositions, is a matter of controversy. It is generally accepted that naturally occurring representations that say how things stand with a changing world – for example, those of the perceptual sort – always entail a risk of error or misrepresentation. Explaining how this could be so is taken to be a primary condition of adequacy for naturalistic theories of mental content. Acknowledging this, it is almost universally agreed that a stronger condition on what it is to be a representation, is warranted. Something is a representation if and only if it presents some portion of the world as being in a certain truthevaluable way. As such, specific representations target aspects of the world, saying how things might be. Propositions expressed by well-formed sentences of natural language remain the clearest and least controversial models of contentful representations. Propositions must in some sense correspond to, or otherwise incorporate, the states of affairs that they represent. What this involves and what propositions are comes out differently on different analyses. For example, Russellian propositions are purely extensional; they include objects or individuals that are thought of as constituents of propositions themselves. Such propositions have the property of being true or false, but they do not do so in virtue of corresponding to anything beyond themselves. In contrast Fregeans hold that propositions have both intensional and extensional aspects. This idea is embodied in Frege’s famous distinction between sense and reference. Intensional content, or Fregean sense, can be understood as the way or manner in which the referent in question is presented to, or apprehended by, the subject. Famous examples include the different senses of the proper names Hesperus and Phosphorus. Although these names denote the same planet, Venus, they have different connotations; one labels it ‘the morning star’ and the other ‘the evening star.’ As such they have the potential to evoke different thoughts in their respondents. For Frege all genuinely referring names and whole propositions were thought
Mental Representation and Consciousness
to have intensional contents, which had to be understood in aspectual terms. Following Frege such contents and their analogues in contemporary theories of representation are commonly known as modes of presentation. Whereas Frege rejected standard correspondence theories of truth, most contemporary accounts do not. Many subscribe to some version or other of possible world semantics. They assume that the intentional content of any given representation – what it is about in extension – is some possible state of affairs. Such representations can also be understood in terms of specific representational targets – what is meant to be represented. Individual representations are thus made true by the obtaining of specific facts. Another important distinction is that of vehicles and contents. In case of language-based representations, linguistic signs serve as vehicles of meaning. They are meaningless, locatable spatiotemporal existents, even if the contents they sponsor are not. Nevertheless they are what enable discrete acts of mental representation; they make representation possible. Drawing directly on the analogy of sentences and their meanings, neural states or processes – or something identifiable with such – are imagined to be the basis of cognitive processing involving mental representation. There is much dispute about the exact form and properties of the vehicles of mental representation. Some, following empiricists, think of them as sensory and imagistic in character. Others endorse a more linguistically inspired conception, holding that the basic units of cognition are amodal symbols; elements of a ‘language of thought.’
Psychosemantic Theories of Content The semantic properties of mental representations, from which – some argue – all other more public representations derive their meaningful content, require special explanation. Psychosemantics is the attempt to produce a workable theory of content. In the prehistory of cognitive science only two sorts of naturalistic theories were advanced to explain representation: resemblance and causal theories. Both proved inadequate. Contemporary
21
thinkers agree that the vehicles of content need not resemble what they are meant to denote in order to represent. The contentful properties of mental representations do not depend or rely upon there being any kind of resemblance between what is represented and how it is represented. There need be no inner pictures in the head that match or copy outer scenes; nor would the existence of such exotica be sufficient to explain representational relations. Two coins from the same national mint will resemble one another (almost exactly), but they do not thereby represent one another. Being similar in such respects is not enough for representation. Causal theories are no better placed to provide a naturalistic account of representation. The crudest versions appeal to causal laws. These can be understood as trading on conditional statements of the following kind: if, in the right conditions, subject S perceives an object (or feature) of type X, then S’s mental machinery will produce a token of type ‘X’ (ceteris paribus). Theories of this kind suffer from a multitude of problems, but the most devastating is the misrepresentation problem. It is demonstrable that any causal account, which makes an appeal to strict causal laws, rules out the possibility of misrepresentation, and thereby the possibility of representation. The problem is that the tokening of symbols would be caused by many other things than those exclusively in the class of things to be represented. Yet if, ex hypothesi, it is only causal relations (as defined by strict laws) that determine how specific symbol types are meaningful (if it is only causal connections that fix their representational content, reference, and truth-conditions) – then such symbols must represent or stand for the entire class of things that would cause them. This rules out any possibility of misrepresentation. But since the possibility of misrepresentation is taken as an essential requirement for representing ‘tout court,’ it turns out that the imagined symbols represent nothing at all. The fatal consequence of crude causal theories of content is that they leave no room for error. All contemporary naturalistic theories of representational content acknowledge the need to account for misrepresentation. They recognize the need to go beyond purely causal-informational accounts if they are to explain mental representation.
22
Mental Representation and Consciousness
Some seek to account for the norms that ground content in counterfactual terms, by appealing to the asymmetric dependence of false tokenings on true ones. This is one way of rising to the challenge. But it leaves open questions as to why one class of things is propriety in this regard, and what establishes and maintains this. Any interesting psychosemantic theory must answer such questions without making an appeal to phenomena that are already infused with, or rely on, the existence of mental representations. Teleosemantic accounts have proved popular for this reason. Such theories look best placed to circumvent the problem of misrepresentation, while providing a substantive naturalistic theory of representational content. To achieve this they make appeal to the notion of organismic proper (or teleo) functions – those that are defined in terms of the ends they serve for organisms. These functions should not be confused with purely systemic functions of the sort that are defined by the role an item or device plays within a more complex system. Rather teleofunctions concern what a device or item is supposed to do as opposed to what it is disposed to do. In explicating teleofunctions the standard naturalist strategy has been to appeal to a principled notion of biological function. The aim is to account for the basis and source of the norms in question in a scientifically respectable manner. There are a number of ways of doing this, but by far the most popular has been to explain the basis of the norms etiologically – that is, by appeal to the historical conditions and evolutionary pressures under which the devices were formed and forged. Construed thus, proper functions are explained in terms of normal conditions of operation that tell how a function was historically performed on those (perhaps rare) occasions, when it was properly performed. The historical conditions in question are those in which a given biological response originally conferred the sort of benefits that brought about, or contributed to, the selection of its underlying mechanism. Appeal to historically normal conditions, therefore, enables the explanation of why a device, entity, or response proliferated, and this, in turn, enables us to say what it is supposed to be doing, even in those cases in which it fails to achieve its ends.
Unlike other representational theories of content, which tend to focus solely on the input, teleofunctional accounts seek to clarify the notion of representation both in terms of indicative relations between the representation and the represented, as well as, by crucially emphasizing the ‘use’ that is made of purported inner representations to achieve specific organismic ends or purposes. Such approaches are consumer-based.
Phenomenal Consciousness Consciousness is an umbrella term. Philosophers recognize many varieties associated with diverse criteria. Which forms are genuine and which are basic is a matter of intensive and ongoing debate. Philosophical taxonomies are meant either to descriptively capture core features of our pretheoretical understanding of consciousness, or to identify its true characteristics, as revealed through analysis. Still, there is no clear consensus on what the concept picks out. The situation reflects, and is exacerbated by the fact that we speak of consciousness in many different ways in ordinary parlance. Nevertheless it is widely agreed that consciousness has some prominent and interesting features that must be either explained or explained away. We say that a creature or organism is conscious if it is awake and sentient. This minimally implies that it has some degree of occurrent awareness, it does not entail that we can describe the character of such awareness. Such consciousness may be awareness of its surroundings or aspects thereof or it may take a more intransitive form. Either way this is generally thought to involve being in a state of mind with a characteristic feel – one in which there is something-that-it-is-like to be in it. So understood, consciousness is an all or nothing property: one either has it or one does not. Human beings, cats, octopi (apparently), and spiders (perhaps) are kinds of things commonly thought capable of possessing it while inanimate objects, such as chairs, are not. Consciousness takes specific forms. What-itis-like to be a human being varies considerably from what-it-is-like to be a dolphin, or more famously still, what-it-is-like to be a bat. Different types of conscious beings enjoy experiences with
Mental Representation and Consciousness
phenomenally or qualitatively different characters. Moreover, particular types of experiences have distinctive characteristics. Experiencing itchiness is quite different from experiencing anger. Seeing the peculiar greenness of an aloe vera plant differs from seeing the greenness of a Granny Smith apple. Apparently experiencing makes a difference. Encountering the unusual taste and smell of durian, for example, may evoke reveries or prompt actions. Yet experiences are only sometimes implicated in the guidance of behavior and action and its autonomous or rational control. It is easy to think of examples of complex activity involving sophisticated, but apparently habitual, automatic, or unreflective responses, which are not governed by agents, precisely because those agents are not conscious of what they are doing in ways that would make it possible for them to modify their behavior. For this reason some reserve the accolade of being conscious only for those beings that exhibit a certain degree of control over their actions or those that are capable of reporting or expressing how things appear to them. Here it seems unavoidable that to have such control a creature must be aware of specific features of its environment in more than a general and intransitive sense. In addition, some hold that awareness of this kind implicates at least some degree of selfawareness. In thinking about the phenomenal character of experience, analytic philosophers often claim that token mental states of specific kinds have distinctive qualitative properties of the sort just described. It is important not to confuse the claim that experiences have phenomenal characters with the claim that these are best understood as causally efficacious mental particulars or objects (often identified with subjectively accessible feels or qualia). Equally, acknowledging the existence of phenomenal characters does not entail that these are ineffable or logically private. Moreover, although it is common for phenomenal properties to be identified using abstract categories (such as ‘greenness’) it is likely that their character is too fine-grained (or analog) to be adequately identified in this way. While sentient beings can selectively attend to and experience certain features of their environment, it may be that they are,
23
nonetheless, only capable of experiencing the world in ways that are highly context-specific and circumstantial. Minimally, to say that a creature or being has an experience of a certain phenomenal character is to acknowledge that there is something-that-it-is-like for it to engage in certain activities. It is to observe that some creatures experience the world in characteristic ways. It is subjective, phenomenal consciousness, of the kind that equates to having a first-personal point of view, or perspective on the world, involving experiences with specific phenomenal characters, that is the primary concern of those who propose that consciousness can be understood in terms of representational facts or properties.
Representational Theories of Consciousness The idea that representation and consciousness are intimately connected made its most prominent debut with the advent of Cartesian philosophy in the seventeenth century. Descartes promoted an understanding of minds as being a special sort of mental substance. To have a mind is to have a coherent and unified individual perspective on reality. These unique points of view are imagined to be internally complex. It is possible to notice and attend to specific worldly features, such as the greenness of a particular apple, but this involves being able to see an apple as something more than just the sum of its presented features. To see an apple as something in which greenness, and other properties, might inhere is to see it as having a continued existence over time. To experience a world of objects and their features always occurs against a larger and more complex background in which such items are systematically related to other things. To have experience of the world, as opposed to merely having sentient capacities, is to experience it as structured. Although Descartes recognized the existence of perceptions, emotions, and sensations, to the extent that these were not under the control of the will – a decisive mark of the involvement of contentful thoughts – he relegated them to purely bodily phenomena, admitting of physiological explanation. For him these were not to be confused
24
Mental Representation and Consciousness
with interesting, bona fide cases of mental states. Only humans enjoy the latter, in his view. This is because we (humans, at least) do not simply get by in the world by responsive and reactive means using a battery of sensory modalities and mechanisms – being moved simply by colors, sounds, tastes or the like – rather we encounter and are conscious of the world and its aspects as being a certain way. It contains a range of properties and things, arranged thus and so. Or, at least, that is how it seems to us. How the world is, in actuality, may be quite different. The driving intuition behind this Cartesian insight is that all genuine conscious experiences are contentful; they necessarily involve having ideas – the ultimate basis for conceptual judgments. Plausibly, this is the central characteristic of the full blown, perceptual consciousness enjoyed by humans (though it remains an open question whether any other species have similar capacities). In promoting this idea Descartes is credited with having initiated the first cognitive revolution. Following in his footsteps, many of today’s philosophers and cognitive scientists also hold that the true phenomenal consciousness must have contentful features. Contemporary representational theories of consciousness endorse the basic Cartesian picture. The most ambitious versions hold that conscious experience simply equates to taking the world to be a certain way. Accordingly, what-it-is-like to be conscious boils down to having a coherent model of how things might be; one that logically and determinately excludes others. Thus, if some aspect of the world appears to be red then that same region of reality cannot also appear to be green, at the same time. This understanding of conscious experience is representational, because the way one takes things to be, can of course be false. The most ambitious representationalists not only claim that phenomenal consciousness is essentially representational in this respect; they also claim that the phenomenal character of experiences is nothing other than, or can be fully explained in terms of, the contents of tokened representations. In recent times this alleged link between phenomenal consciousness and mental representation has been less evident. Since the birth of cognitive science, until quite recently, there has been a
tendency to adopt an isolationist policy regarding these two aspects of mind. They have generally been treated as entirely separate topics of study. The tendency has been to focus on theories concerning unconscious mental states and subpersonal representations, and to ignore or avoid the question of consciousness altogether. In such a climate, demonstrating that representation and consciousness are essentially linked is a substantive project. Existing representational theories of consciousness come in a variety of forms and strengths. Some versions claim that representational facts exhaust all the experiential facts; that all experiential facts are representational facts. This is a relatively weak claim. Its proponents only commit to the view that phenomenal consciousness supervenes upon representational facts. The former need not reduce to the latter. Being in a state of mind with a representational content might be essential to being conscious, without it being the case that being conscious is nothing but representing. Taking the world to be a certain way might be one necessary factor amongst others (such as a mental state’s having particular functional properties). The phenomenal character of a given experience might be determined in complex ways. Pitched at this level of grain, representationalism can tolerate impurities in its account of what makes a creature (or its mental states) phenomenally conscious. This sort of proposal is typically cast in terms of covariance relations holding across possible worlds. The basic idea, which admits of considerable modal refinement, is that there is no possible world such that all the representational facts remain the same, yet exhibit a difference with respect to the phenomenal consciousness. For such a law to hold, it need not be the case that conscious properties per se reduce to representational properties, all that need be true is that facts about representation and consciousness travel together, always and everywhere. It may be true that there cannot be phenomenal character without representation, without it being the case that the two are more intimately related. A stronger thesis is that there can be no difference in phenomenal character without a corresponding difference in representational content; either because consciousness is just a kind of representational
Mental Representation and Consciousness
content or because it lawfully covaries with changes in such content. Those who believe that the phenomenal character of conscious mental states is exhausted, or entirely determined, by the representational properties of such states support such views. Whether representationalists advance fact-based or property-based variants, insofar as they only advance supervenience claims, they leave open a range of possibilities about what might best explain the metaphysical basis of the proposed systematic relationships. This might be explained in terms of causal dependencies, mereological part–whole relations, or token identities. The scope of such proposals is restricted in that even strong representationalists only propose that all experiential properties are representational and not vice versa. The claim is that all conscious experiences are, or are logically related to, representations, not that all representations are, or logically related to, conscious experiences. This is an important qualification because it seems clearly false that every representation does its work by invoking phenomenal properties. For example, sincerely asserting that ‘‘Snow is white’’ says that the world is such that snow is white (relative to a time, place, and language). The English sentence itself may have familiar phenomenal properties – indeed ones that we can attend to – but it does not represent the putative fact that ‘‘Snow is white’’ by presenting things to the speaker or hearer in a way that relies on it appearing that snow is white. The word ‘white’ does not denote whiteness by looking white or exhibiting the quality of whiteness. The English sentence, cited above, is a truthevaluable representation, but it is not of the mental sort. Still the point carries. For the same verdict ought to apply to items in the mental lexicon and the sentences composed from them (should any exist). For example, if we imagine that modes of presentation of some mental representations are entirely syntactic and linguaform – if they take the form of well-formed strings of amodal symbols – then such representations will clearly share the characteristic of natural language sentences of representing what they do by nonexperiential means. This does not show that no mental representations could have the relevant properties. But it does show that at best representationalism will be true only of mental representations of a select
25
kind. Representing the world as being a certain way is not sufficient for being an experience with a particular phenomenal character. Not every representational fact entails an experiential fact. For example, some believe that even the lowlevel activity in early stages of human visual processing involves the manipulation of contentful representations. But, if so, the content of such representations will be quite distinct from the contents of ordinary perceptual experience. The cells that fire in low-level visual processing are orientation-invariant, whereas, our experiential way of perceiving objects is not. This implies that, at best, only a subclass of representations exhibit or involve phenomenality. It also follows that genuine phenomenal consciousness might be the province of only a small class of organisms; those whose online perceptual processing requires rich, complex, and multilevel forms of representation. It has been argued, for example, that only the abstract representations allegedly involved in the intermediate stages of human visual processing would be of the right kind to possess experiential content. If so, perhaps as Descartes thought, only humans are truly conscious. Importantly, those who defend such views also hold that contentful experiences are logically distinct from (and typically ontologically prior to) perceptual beliefs and judgments. Seeing that the wall is of a certain illumination, with various hues and a certain distance away involves enjoying a nonlinguistic, nonconceptual, representational content. It does not involve making a belief-based judgment. As a result only a special class of representations will be suitable candidates for rendering true the supervenience claims of representationalists. Yet some proponents of this view draw further distinctions even within this category, restricting their claims to the specific types of perceptual experience found in particular modalities. Accordingly differences in the phenomenal character track differences in content within the visual, auditory, or tactical modalities. There is an obvious motivation for wanting some suitably strong version of representationalism to be true. If it could be convincingly demonstrated that the phenomenal features of consciousness are really nothing but the representational features of mental states, then this promises an attractive
26
Mental Representation and Consciousness
metaphysical economy. Two seemingly intractable problems in the philosophy of mind would in fact turn out to be a single problem: that of accounting for representational content in naturalistic terms. While naturalizing content is generally acknowledged to be difficult to achieve, it is also thought to be less challenging than providing a straight solution to the hard problem of consciousness. If strong representationalism were true then this would allow for a considerable reduction of effort: the naturalization of representational content, in a way consistent with physicalism or materialism, would be the only hard problem. Representationalism has some intuitive appeal. It trades on the insight that conscious experience is transparent or diaphanous. To the extent that experience has any qualitative aspects, we only become aware of these by focusing on features of what a given experience is about (or is apparently about). We typically see right through our experiences to the objects of our concern or interest. In most cases it is the character of what our experiences are about that is the focus of our attention and not the experiences themselves. We are conscious of the trees blowing in the wind or of the mountain in the distance, first and foremost. Only on rare occasions, if at all, is it possible to focus our attention on the properties of the way we experience things. And, arguably, even in those cases the content of what we introspect is borrowed from the basic content of the experience itself. Those who defend a strong transparency thesis maintain that even in cases in which we concentrate on features of our experiences, we have no way of specifying their phenomenal character, other than by describing the features of what the experiences are about. Despite its initial plausibility, the truth of the strong transparency thesis has been challenged of late and softer variants promoted. Notably, although the strong transparency thesis helps to motivate representationalism, the program could in principle survive its loss.
General Objections to Representationalism A standard objection to representationalism is that it fails as a general account of conscious
experience, because it fails to encompass experience of pain, feelings, or bodily sensations (such as tickles, itches, bouts of nausea) or undirected, diffuse moods such as depression and elation (those which have no precise focus or target). Its defenders have tried to show that such experiences can be accommodated if we think of them as telling us something about the state of the organism in question – that is, as indicating something about damaged body parts, internal or external, or about the state of the creature more generally. In effect, representationalists hope to demonstrate that all experience is ultimately perceptual in nature. For if it should turn out that there are nonperceptual experiences with distinctive phenomenal characters, it would undermine the general truth of their thesis. Others have directly attacked the supervenience claims of strong representationalists. A number of thought experiments have been fashioned in order to put these under intuitive pressure. The hope has been to demonstrate that it is logically possible that representational content and phenomenal character can come apart. One way of achieving this asks us to imagine that there could be a case of difference in phenomenal character without any difference in representational content. Thus we are asked to imagine two different subjects that are functionally equivalent in the performance of a given recognition-based task, say, the sorting of red socks from blue socks. It is assumed that in order to perform this task they must represent aspects of the world as being a certain way. But it is then further stipulated that one group of sock-sorters see red as blue, while the others see red as red. But this makes no difference to either group’s ability to perform the tasks. If such cases are possible, and not just imaginable, then the qualitative aspects of experiences might vary arbitrarily, relative to their representational contents. Another tactic is to construct cases in which there is a difference in representational content while the phenomenal character remains the same. Ned Block’s Inverted Earth attempts this. The colors of his imagined world are systematically reversed with respect to ours, such that there green things are red, blue things are yellow, and so on. The language spoken by Inverted Earthlings
Mental Representation and Consciousness
is also inverted; as a result, it perfectly mirrors that of Earthlings. Just like us they are inclined to voice that ‘Grass is green’ and ‘Roses are red.’ Bearing this in mind, we are then asked to imagine that after having undergone color-inversion operations to their visual systems some Earthlings are transported to Inverted Earth – all without their knowledge. On their arrival to the character of those they experiences would be imperceivably different to those they would have on Earth in similar circumstances, even though they would be everywhere misrepresenting the colors of this new environment. This would constitute a difference in representational content without a relevant difference in phenomenal character. If such imagined scenarios are to serve as genuine counterexamples, it must be assumed that our capacity to imagine them is metaphysically revealing and relevant. A standard response to the use of such thought experiments is simply to deny this. Naturalists in particular generally challenge the idea that such imagined cases have any philosophical bite. A stronger offensive tactic is therefore to identify internal tensions that render specific representationalist proposals incoherent or unworkable. In proposing a general metaphysical thesis about the nature of phenomenal consciousness, strong representationalism need not commit to any specific theory about the nature of content, such as externalism or internalism, nor need it commit to any particular psychosemantic theory of content, such as biosemantics or causal-informational theories. But evaluating the truth of any interesting version of representationalism ultimately requires assessing specific proposals in the light of such commitments.
Objections to Externalist Variants Some strong representationalists hold that when it comes to understanding phenomenal character, we should concern ourselves only with what it is that an organism is having an experience of. What is of interest is what the experience is and how what it is about is represented. Representationalists of this stripe are interested in intentional content – what it is that one is conscious of – and not the way in which one is conscious of it.
27
Such theorists endorse externalism about phenomenal character, if they also accept that intentional content cannot be individuated by internal factors alone; a view that has some support today. Thus if phenomenal character is nothing but (or strongly covaries with) intentional content and intentional content is wide or broad (in that it cannot be individuated solely by appeal to the internal properties of an organism) then phenomenal character will also share this feature. The standard externalist versions of representationalism hold that to represent is essentially a matter of tracking, or having the function to indicate specific worldly features. But which worldly features do experiences represent? These are identified with external, mind-independent, if subject-relative, ‘affordance-like’ properties – properties of the world that reliably (or reliably enough) elicit certain dispositions in specific organisms. Surface qualities or properties are what are represented. Advocates of this sort of view confront the problem of making good on objectivism about the so-called secondary properties. Consider the case of colors. Most philosophers are wary of the idea that there are any worldly features with which the colors we experience can be uniquely identified. It is generally acknowledged that our experience of color depends both on the character of our internal processes as well as our responses to specific external stimuli. To begin with, the colors and shades we see are in part determined by the ratio of activation across three light-sensitive cones in our retinas. These ratios of activation do not correspond in a one-to-one fashion to particular light stimulus configurations, but rather with a disjunction of the latter. Moreover, the distributed response of these cones initially effects changes in the retinal ganglion, and from there further processing occurs in two separate chromatic channels, known as the red-green and the blue-yellow. The interaction between these two channels underpins our chromatic experience in a way that makes the classification of hue perception a fourfold business. The problem for advocates of color externalism is that, if what constitutes our color experience constitutively depends on internal factors of this kind, then it looks as if there is not much prospect of identifying colors with any exclusively,
28
Mental Representation and Consciousness
mind-independent, external features of the world (such as particular wavelengths of light or spectral reflectance signatures or the like). If there are no interesting, mind-independent features of the world which can be cleanly identified with the various colors we experience then we must ask: Which properties of the environment are we tracking or representing in having color experiences? Unless there is a principled answer to this question externalism about experience is under threat. The problem generalizes, since a similar verdict looks apt to apply to other responseeliciting surface properties as well. A more serious problem facing those strong representationalists who endorse externalism concerns misrepresentation. Let us suppose that ‘phenomenal green’ is a mind-independent, response-eliciting property. We know that it can be represented truly in a variety of ways. I can represent its presence correctly even though my way of doing so does not depend on my representing its surface properties as being green per se – as when I state that the fourth book on the shelf behind me has a green spine (doing so from memory and without in any way imagining that color). But equally I might also be inclined to deny the presence of ‘phenomenal greenness’ even when I am, as a matter of fact, confronted with it. I might see something that is actually ‘phenomenal green,’ but which looks red to me: my mode of presentation may have a ‘red’ character. Perhaps this will be because the lighting conditions have been carefully adjusted so as to generate precisely this type of illusion. This would be a case of misrepresentation since I would be inclined to act as if (and would judge that) the portion of the world in question was in fact ‘red’ when it was really ‘green.’ This sort of situation is clearly possible (and not just imaginable like the Inverted Earth scenario). If so, it looks as if the relevant supervenience claim is false if one assumes that phenomenal character supervenes on wide intentional content. It seems possible to have experiences with phenomenal characters different from the intentional content of what is represented. In thinking about such cases it is important to realize that to represent successfully essentially requires that a correspondence holds between
what is represented and what does the representing. Following the linguistically inspired model of representation, specifying the content of truthconditional representation involves disquotation. Thus it is no accident that representational contents are defined by making direct reference to the worldly features or objects that representations aim to represent. But in some contexts this can be misleading, since to specify representational content in this way requires focusing squarely and exclusively on cases of successful representation. This raises important questions about how to specify the contents of representations in cases of misrepresentation; those in which there is a mismatch or, more precisely, a lack of correspondence between the state of the organism and the world. It is accepted that the way organisms respond to the range of nonveridical stimuli in instances of misrepresentation must be understood as strongly dependent upon, or as having been in some way fixed by the way they respond to a select propriety class of referents. It is because they respond appropriately when encountering tokens of those types that their representations are veridical. The precise analysis of this dependency relationship, and the prospects for explaining it naturalistically, vary according to different psychosemantic theories, as noted above. Nonetheless, whichever theory one favors, problems arise for those who want to defend the idea that phenomenal character equates to wide or broad content. For the phenomenal character of experience, and what is represented, appear not to always and everywhere keep in step. If anything, it looks as if phenomenal character is best identified with the nonintentional properties of certain modes of presentations. This might be understood in terms of representational character, as opposed to representational content. Phenomenal character should be understood as the way or manner in which things make their appearances rather than being identified with what is represented in particular cases (even if we assume that the latter places unavoidable and interesting constraints on the former). To recognize this one must acknowledge that acts of conscious experiencing with phenomenal characters have a dual structure, involving features of which one is aware and features in virtue of which such awareness is made possible.
Mental Representation and Consciousness
Objections to Internalist Variants Variants of strong representationalism that seek to reduce phenomenal character to (or identify it with) external, intentional content face serious problems. But externalism is not the only option for strong representationalists. Some have proposed that there may be an identity between phenomenal character and narrow content. Narrow content is a special sort of content recognized by internalists. The notion was first introduced in order to show how everyday psychology (which apparently only trades in wide content attributions) could nevertheless be at peace with methodological solipsism. Methodological solipsists hold that it is what agents have in mind (not external features of the world) that matter to what they do and for explanations of what they do. Recognition of the existence of narrow content was meant to show that it is possible, at some level of abstraction, to class behaviors as being psychologically identical, even when the propositional attitudes of the agents initiating those behaviors had different wide contents. This was meant to capture what is psychologically similar about physically identical duplicates when they are each operating in environments in which the wide contents of their thoughts vary, such as when one is on Earth and the other on Twin Earth. Twin Earth is a philosophical construct, a world exactly like this one in every detail, but in which the seas, lakes, and baths are filled with the substance twater. Twater has all the outward properties of water but it has the chemical make-up of XYZ (not H2O). Consequently, despite the fact that the thoughts of the two imagined agents are about different things, their behavior will be exactly the same. Internalists hope to explain this by proposing that what they have in common is the narrow content of their thoughts. Narrow content is a property of internal states; it is defined in terms of partial functions that map thoughts onto environments. This is the sort of content that the imagined twins can share even though the truth values of their thoughts may differ. It is not possible to give a positive, free-standing specification of narrow content. It is normally defined indirectly as the type of content that is independent of the way in which subjects are
29
related to their environment. It is emphatically not a kind of truth conditional content. On the other hand, it is not reducible to a mere vehicle of content, understood in terms of its syntactical or functional role. Narrow contents are at best truthapt in that they become truth-evaluable when and only when they are appropriately anchored to an environment. Hence the best way to characterize narrow content is to think of it as a partial function that needs to be contextualized. Ultimately its specification depends on making an appeal to the wide intentional content. Narrow content is thus inherently and radically inexpressible. This fact raises the concern that narrow contents (should they exist at all) are misnamed; perhaps they are not really any kind of content. For if narrow contents are best understood as functions (mathematically conceived) from contexts to truth conditions then they are not strictly speaking about anything. At best, they have a special kind of potential to be so. To equate phenomenal character with narrow content would entail that phenomenal character is at best a kind of nonveridical content. This avoids the problems associated with misrepresentation, but only because there is no question of narrow contents being true or false. Such a move would seem to suit certain widely held Cartesian intuitions. For unreformed Cartesians would want to identify phenomenal character with how things seem to us. And how things seem to us is not something about which we can be mistaken. I can be mistaken that the tomato before me is green, but not that it looks green to me. My phenomenal awareness (how it seems) can be at odds with how the world is. If the phenomenal character of experience is just a kind of narrow content then it would be a kind of content for which the question of misrepresentation does not arise. If there is any coherent type of representational content that has such a feature then it is narrow content. But narrow contents are not generally assumed to be consciously accessible and if the notion proves to be unstable then, once again, phenomenal character will be identified with the nonrepresentational properties of representational states (at best). For example if the notion of narrow content does not stand up to critical scrutiny then phenomenal
30
Mental Representation and Consciousness
character might be identified with noncontentful representational vehicles. But clearly any such straightforward identification on its own will not adequately explain why or how we sometimes represent the world in a phenomenologically salient way. To constitute a satisfying explanation of this kind the vehicles in question would have to have some rather special additional properties, but such properties may, at best, be logically related with certain kinds of representational acts without being representational or contentful properties per se.
Objections Concerning Subjectivity Another worry for representationalists is that more is needed to understand phenomenal consciousness than an account of phenomenal character in any case. Even if there were a workable representationalist story about the latter it would not suffice to explain the former. This is obviously so if, for example, we allow that it is possible to be in a state of mind with a specific phenomenal character without in any way being aware (or even being capable of being aware) of being in such a psychological condition. It is imaginable that certain creatures might undergo pain (in the sense that there is something-that-it-is-like for them to have painful experiences) even though they would be congenitally unable to access these qualitative feels for themselves. If so they would feel pain without being aware of the painful experiences. This seems more than merely theoretically possible. There is some phenomenological evidence that we sometimes undergo experiences with characteristic phenomenal feels, such as being in pain or anger, without noticing, attending or being aware of these at the time of their occurrence. If so, we can distinguish experiences of which organisms are unaware from those that are, to invoke some well-worn metaphors, brought before the mind’s eye or illuminated by the spotlight of attention. Some will object that the very idea of unexperienced experiences is incoherent. They will claim that to have an experience logically entails the existence of a subject (of some sort) – one that actively undergoes or is aware of the experience in question. Experiencing necessarily involves
bringing the subject in on the act. This is because experiencing is necessarily experiencing-for something or someone. Whether unexperienced experiences are possible or not, it is clear that experiences are ordinarily experienced. If so a proper account of everyday phenomenal consciousness requires something more than a mere account of phenomenal character; in addition it requires an account of the subjective nature of experiencing. A number of recent attempts have been made to account for the subjectivity of consciousness in representational terms. These come in the form of higher-order theories of representation of either perception-based or thought-based varieties. The core idea behind all such theories is that first-order mental states are in some sense rendered conscious when they become the objects of other mental states. This is thought to involve a kind of second-order self-monitoring that utilizes either meta-representations or re-representations. A root problem with such higher theories of representation is that they threaten to make the first-order phenomenal character of experiences superfluous in fixing or determining the what-itis-likeness of phenomenal experiences. This is because if the subjective aspect of experiencing is determined by second-order representational properties then there is always a risk of misrepresentation. Thus it is possible that what-it-is-like to experience, say, the color green will be fixed by higher order representational properties even in those cases in which the phenomenal character of the lower-order mental state is misrepresented. These will include situations in which one is, in fact, having an experience with a different phenomenal character (say, of red or blue) and those in which one’s mental states lack phenomenal character altogether. In such circumstances things might still appear to be green because the higherorder monitoring is doing all of the decisive work in determining how things seem to the subject. The problem emerges in duplex accounts of this kind because there is always the possibility that the low-order and higher-order states might be out of sync. This is because representation is an extrinsic, external relation and not a constitutive one. This difficulty can be overcome or avoided by supposing that our experiences are in some way
Mental Representation and Consciousness
intrinsically subjective. An intimate, nonpositional kind of awareness is quite unlike adopting a representational stance toward one’s mental states. Experiences might be Janus-facing – if so, they would not only exhibit world-direct intentionality, but their character is given to the experiential subject in some minimal, prereflective way. The idea that experience necessarily has the feature of givenness is a familiar one in the work of the great phenomenological thinkers and it can be found, arguably, even in Descartes and Kant. The closest representationlists have come to accommodating this has been to endorse the view that phenomenal consciousness involves the having of self-representing representations. But to make sense of this proposal requires positing the existence of representations of a unique kind, those with quite special (if not naturalistically impossible) properties. Accounting for such representations would be difficult, at best, within the compass of existing theories of psychosemantics.
Nonrepresentational Accounts Strong first-order representationalist proposals (of both the externalist and internalist varieties) look to be too strong. The way things seem or feel to an organism phenomenally does not appear to equate to any kind of representational content, narrow or wide. Weaker, impure versions of representationalism – those that cast their identity claims at the level of facts and not properties – are more plausible. Proponents of such accounts concede that not only do select representations have phenomenal features, but also that such features may be best understood in terms of nonrepresentational properties. As such weak representationalism allows that the phenomenal features that give experiences their special characters are not explained by their being a kind of representational content, but, at best, by their role in performing representationally related service. This claim is weak enough to be consistent with accounts of the nature of experiences that regard them as external, in a way quite different to the proposals of strong representationalists.
31
Emphasizing their ecological and interactive nature, enactivists claim that experiences are not inner events of any sort; rather they are constituted by the activity of organisms engaging with some environment or other. Subjective experiences that have specific phenomenal characters are identified with temporally extended worldly interactions. Brain activity may be necessary, but it is not sufficient for experience. The causal substrate that fixes the character of any token experience is temporally and spatially extended to include aspects of the brains, bodies, and environments of organisms. Ontologically speaking experiences just are extended spatiotemporal happenings – the reactions, actions, and interactions of embodied creatures situated in the world. As such they are a subclass of events; they are roughly dateable and locatable happenings. While such mental events necessarily involve active, living organisms, they do not occur wholly inside them. Experiences, understood as events or token occurrences, can also be characterized as physical. They are instances of lived, embodied activity, which can be designated and characterized using vocabulary that makes no use of phenomenal predicates. For this reason experiences can be the objects of scientific theorizing and experimentation. All of this is consistent with weak representationalism, but the more radical versions of enactivism go further, claiming that organisms act first and develop the capacity for thought later. Accordingly they hold that it is not necessary for creatures to think about or represent in order to act successfully in the world. If this is true it reduces the scope of even weak representationalism, since phenomenal consciousness would be more basic and fundamental than the capacity to represent the world as being a certain truthevaluable way. Experiencing aspects of the world might be thoroughly noncontentful; aspects of the world might not be presented as being in a certain contentful way to subjects at all, even though there would be something-it-is-like to engage with worldly offerings in phenomenologically salient ways. See also: Inner Speech and Consciousness; MetaAwareness; Visual Imagery and Consciousness.
32
Mental Representation and Consciousness
Suggested Readings Block N (1990) Inverted earth. Philosophical Perspectives 4: 53–79. Brook A (ed.) (2007) The Prehistory of Cognitive Science. Basingstoke: Palgrave. Byrne A (2001) Intentionalism defended. Philosophical Review 110: 199–240. Crane T (2001) Elements of Mind. Oxford: Oxford University Press. Dretske F (1995) Naturalizing the Mind. Cambridge, MA: MIT Press. Fodor JA (1987) Psychosemantics. Cambridge, MA: MIT Press. Gallagher S and Zahavi D (2007) The Phenomenological Mind. London: Routledge. Kind A (2003) What’s so transparent about transparency? Philosophical Studies 115: 225–244.
Kriegel U (2002) Phenomenal content. Erkenntnis 57: 175–198. Lycan W (1996) Consciousness and Experience. Cambridge, MA: MIT Press. Neander K (1998) The division of phenomenal labor: A problem for representationalist theories of consciousness. Philosophical Perspectives 12: 411–434. Noe¨ A (2004) Action in Perception. Cambridge, MA: MIT Press. Rowlands M (2001) The Nature of Consciousness. Cambridge, MA: Cambridge University Press. Seager W (2000) Theories of Consciousness. London: Routledge. Tye M (1996) Ten Problems of Consciousness: A Representational Theory of the Phenomenal Mind. Cambridge, MA: MIT Press. Zahavi D (2005) Subjectivity and Selfhood: Investigating the First-Person Perspective. Cambridge, MA: MIT Press.
Biographical Sketch
Daniel D. Hutto was born and schooled in New York, but finished his undergraduate degree as a study abroad student at St. Andrews, Scotland. His maternal roots are Scottish and many of his family still live in Inverness. He returned to New York to teach fourth grade in the Bronx for a year in order to fund his MPhil in logic and metaphysics, after which he carried on his doctoral work in York. He now lives in Hertfordshire with his wife and three boys, having joined the local university in 1993. He served as the head of philosophy from 1999 to 2005, and is currently the research leader for philosophy. He has published on wide range of philosophical topics in journals including: The Monist; Proceedings of the Aristotelian Society; Philosophy and Phenomenological Research and Mind and Language. He is the author of The Presence of Mind (1999), Beyond Physicalism (2000), Wittgenstein and the End of Philosophy (2006), and Folk Psychological Narratives (2008). He is also the editor of Narrative and Understanding Persons (2007) and coeditor of Current Issues in Idealism (1996) and Folk Psychology Re-Assessed (2007). A special yearbook issue of consciousness and emotion, entitled Radical Enactivism, which focuses on his philosophy of intentionality, phenomenology, and narrative, was published in 2006.
Meta-Awareness J M Chin, University of British Columbia, Vancouver, BC, Canada J W Schooler, University of California Santa Barbara, Santa Barbara, CA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Automaticity – The cognitive processing associated with highly practiced activities. Flow – A mental state marked by full immersion in a task, such as the feeling of being ‘in the zone.’ Meta-awareness – The process of directing attention toward the contents of consciousness, thereby gaining an appraisal of the contents of consciousness. Metacognition – Knowledge about one’s knowledge. Metaconsciousness – Another term for meta-awareness. Mindfulness – A state that is enhanced by a practice of meditation, in which one engages in watchful attention and is very present in the moment. Mind wandering – The experience of having thoughts that are unrelated and often counterproductive to the task and hand. Recovered memories – Memories, corresponding to what would seem to be highly memorable experiences such as trauma or sexual abuse, that are perceived to have been forgotten and subsequently recalled. Temporal dissociation – A temporary disconnect between consciousness and meta-awareness in which one goes for some period of time without noticing the contents of their thought. Translation dissociation – A disconnect between consciousness and metaawareness, in which meta-awareness, in reappraising consciousness has misrepresented the original experience. Tuning out – Mind wandering that an individual is aware of engaging in.
Verbal overshadowing – The phenomenon of verbal report interfering with a related task. For instance, describing a face and subsequently having poorer recognition for that face. Zoning out – Mind wandering without awareness.
Introduction ‘Meta-awareness,’ a term often used interchangeably with metaconsciousness, is the state of deliberatively attending to the contents of conscious experience. Frequently when researchers speak of consciousness, they distinguish between two general states: unconscious, in which if information is processed at all, it is processed without any concomitant experience, and conscious, in which individuals experience what ever is occupying their minds. In this context, meta-awareness can be thought of as third level of consciousness in which consciousness is turned upon itself in order to re-represent the contents of experience. In other words, meta-awareness is one’s explicit appraisal of the current contents of consciousness. The distinction between unconscious, conscious, and metaconscious processes can be illustrated with the example of mind wandering while reading. Consider the all too familiar case of reading along and then suddenly realizing that despite your best intentions, although your eyes have continued to move across the page, your mind was fundamentally elsewhere. In this example, the pattern recognition procedures that allowed you decode the individual words correspond to an unconscious processes. Conscious experience would primarily correspond to specific contents of mind wandering musings, although
33
34
Meta-Awareness
presumably some superficially processed aspects of the reading (e.g., the sounds of some of the words) would also enter consciousness. Initially in this example, meta-awareness would be absent until you notice that you are mind wandering. This abrupt realization (almost like waking up) represents the dawning of metaconsciousness, in which you take stock of what you are thinking about and realize that it has nothing to do with what you are reading. Once meta-awareness of mind wandering is achieved, then you are able to redirect your attention to the narrative and can search for the last sentence you remember actually attending to, and try again. Several important features of meta-awareness emerge from the mind wandering while reading example. First, as conscious and unconscious processes are presumed to carry on relatively continuously throughout waking life, metaconsciousness is intermittent. Only periodically does one take stock of the contents of thought, which is how it is possible for one to mind wander during reading despite knowing that it is impossible to comprehend text while simultaneously thinking about something entirely unrelated to what one is reading. Second, when one notices that they are mind wandering while reading, there is often an abrupt experience of ‘coming to.’ It is almost as if one has just awoken even though one was not previously asleep. The abruptness of metaconscious understanding illustrates the qualitative difference between a mental state that is associated with metaconsciousness and one for which metaconsciousness is lacking. Finally, it is important to emphasize that when one notices that the mind has wandered, this fact temporarily becomes the dominant element in consciousness, as one thinks to themselves ‘‘shoot I just spaced out again, let’s see: where was I?’’ Thus, although metaconsciousness represents a distinct category of thought, it may be nothing more than a particular kind of subject matter of consciousness, namely when the subject of consciousness becomes an acknowledgment of whatever it is that consciousness had just been focusing on. In short, metaconsciousness need not represent a unique mental state, nor need it entail specific dedicated brain regions. It may simply be a topic area (albeit a very important one) of consciousness.
History Meta-awareness can be thought of as a subtype of a larger category of mental phenomenon known as ‘metacognition.’ Metacognition corresponds to our knowledge about what we know. The term was introduced by John Flavell in the context of studying children’s capacities for monitoring their cognitions. Flavell distinguished two general classes of metacognition. The first was metacognitive knowledge, or in other words understanding how thought operates in the world. Belief that rehearsal improves memory falls under this category of metacognitive knowledge. The second class was metacognitive experiences, which was more focused on the feelings associated with metacognition. The feeling that a sought for word is on the ‘tip of one’s tongue’ falls under this heading. A key moment in the development of the construct of meta-awareness came with Thomas Nelson and Louis Narens’ introduction of a basic model of the control and regulation of cognition in a paper published in 1990. This early model proposed two levels of cognition, the object-level and the meta-level. The object-level deals with cognitions about external objects, while the metalevel deals with cognitions about object-level cognitions. Moreover, information was said to flow between these two levels. Monitoring occurs when the object-level is informed by information from the meta-level. Control, on the other hand, occurs when information from the meta-level modifies the object-level. These applications of metacognitive knowledge provided a shift from work focusing on an understanding of conscious thought, toward work on meta-awareness as a specific representation of consciousness. The distinction between monitoring and control remains prevalent in current work on meta-awareness.
Meta-Awareness and Monitoring Many aspects of daily life require routine monitoring and adjustment. When one speaks, one needs to adjust their volume to the ambient noise in the room. When one reads, one needs to adjust the rate at which they move their eyes in accordance with comprehension. In many cases, this
Meta-Awareness
type of monitoring and adjustment can carry on behind the scenes without any explicit awareness. However, the tacit monitoring system is limited in the types of things it can monitor. It can, for example, help one to keep one’s volume appropriate under normal circumstances, but if one is wearing headphones one has to explicitly attend to one’s volume when speaking, otherwise one will be shouting. Similarly, tacit monitoring processes can control eye movements sufficient to recognize the word, but recognizing that one is not attending to what the words are saying requires a more sophisticated monitoring process. Meta-awareness provides this more sophisticated form of monitoring. Comparisons can be drawn between the meta-aware monitoring system and the pilot of an airplane. The autopilot system (tacit monitoring) can efficiently pilot a plane under most circumstances, making minor adjustments to keep the plane on track under most conditions. However, if something major occurs, the pilot (metaawareness) is needed to make major corrections and decisions. The meta-aware monitoring system has a lot more resources available as it draws from several different systems, but is also more resource-taxing and can potentially interfere with carrying out concurrent tasks. Thus, while it is necessary to invoke meta-awareness periodically to make sure that things are on track, the common absence of meta-awareness is adaptive because it frees up resources that can be applied to the task at hand.
Dissociations between MetaAwareness and Consciousness Because meta-awareness is re-representation of the contents of consciousness and not consciousness itself, it is possible that meta-awareness can in some cases be an imperfect or poorly timed translation of conscious experience. In other words, people may fail to take stock of their conscious thoughts, or may do so inaccurately. These dissociations between meta-awareness and consciousness can have important and far-reaching consequences. In 2002, Schooler elaborated on the concept of metaconsciousness by proposing two specific types of dissociations between experiential consciousness
35
and metaconsciousness: temporal dissociations, which are discovered experiences that once had eluded meta-awareness, and translation dissociations, which encompass occasions when meta-awareness does not accurately reappraise conscious experience. Both types of dissociations are explained in greater detail in the following sections.
Temporal Dissociations Conscious experience can often occur in the absence of meta-awareness. When triggered, metaawareness can lead to a reappraisal of elements of conscious thought that once eluded meta-awareness. This experience of discovering an experience that one was previously not meta-aware of is known as a temporal dissociation. Mind Wandering Mind wandering without noticing it is a quintessential example of a temporal dissociation metaawareness. The pervasive phenomenon of mind wandering occurs when attention is decoupled from the task an individual intended it be directed toward. In many situations, mind wandering may be quite adaptive or at least, harmless. For example, when one is walking to work, it may be helpful to think about what one needs to do that day, rather then devoting all attention to the nondemanding task of walking down the sidewalk. However, in other situations, for example, when one is driving in difficult traffic or reading an important paper, mind wandering is counterproductive. The fact that individuals mind wander even when engaged in tasks that they recognize as being undermined by mind wandering illustrates how easy it is to temporarily lose track of what is going on in one’s mind, that is, to have a temporal dissociation of meta-awareness. In recent years, a number of laboratory studies have investigated the process and impact of mind wandering without meta-awareness. Under laboratory conditions, two different approaches have been used to sample mind wandering. The first approach, the probe-caught method, samples the experience of the individual at varying time intervals as they perform a cognitive task. The second
36
Meta-Awareness
approach, the self-caught method, requires the individual to respond with a button push whenever they catch their own mind wandering. Probe- and self-caught measures of mind wandering yield different information on the occurrence and awareness of mind wandering because they systematically sample the different aspects involved in off-task experiences. The probe-caught technique provides evidence of how readily the mind turns inward, and can be used to study the onset of decoupling, or the speed of drift within attention. On the other hand, the self-caught method requires the individual to recognize that his or her mind is wandering, and so illustrates the engagement of meta-awareness of their own mind wandering. Evidence of the value of distinguishing between probe-caught and selfcaught mind wandering comes from the findings that the two measures are differentially associated with task performance. Interestingly, it is the probecaught mind-wandering episodes that tend to be maximally associated with detriments in performance, including both reading comprehension and memory. The more modest consequences of selfcaught mind-wandering episodes suggest that when individuals are meta-aware of mind wandering, they are more effective in circumventing its costs either by more effectively dividing attention or by more efficiently recovering information that was missed during the lapse. A second way in which meta-awareness during mind wandering has been assessed is simply to ask people following a probe whether or not they had previously been aware of the fact that they were mind wandering. Strikingly this simple procedure reveals consistent differences (in keeping with the self-caught probe-caught distinction introduced above). For example, mind wandering episodes that are characterized as having occurred without meta-awareness (termed ‘zone-outs’) are typically correlated with performance detriments, whereas mind wandering episodes that occur with metaawareness (termed ‘tune-outs’) tend to be less problematic. Neurocognitive measures have revealed a similar story. For example, errors on a simple vigilance task were found to be correlated with zone-outs but not with tune-outs. Similarly, in a functional magnetic resonance imaging (fMRI) study of mind wandering, it was found that the difference in brain activation between on-task and
off-task performance was markedly greater when individuals reported being off-task and unaware, relative to off-task and aware. Automaticity During mind wandering, it can often seem like the task at hand (e.g., reading, in some cases) has been put on autopilot. There are some tasks, however, that can more easily be put on automatic, and these tasks are also understood better under the umbrella of meta-awareness. Automatic behaviors are often considered to be unconscious, a designation that can prove problematic. For instance, driving an automobile can become automatic, especially if one is attempting to do something like hold a conversation while doing so. The person will then often find that he has arrived at his destination with little memory of the actual drive. The driving, however, was not unconscious as the driver was certainly experiencing the road at some level. Meta-awareness allows psychologists to posit that the driver was conscious of the driving, but not meta-aware of these behaviors. Unwanted Thoughts As noted earlier, sometimes the mind can wander even when the individual is explicitly told to reign it in. But, could individuals experience this same gap in meta-awareness when the cost is revisiting a terrible thought or memory? Psychologists, such as Daniel Wegner, have often wondered about why it is so difficult to suppress unwanted thoughts. Meta-awareness shines new light on this troubling issue. Some unwanted thought theorists have hypothesized that these unwanted thoughts lie in the unconscious (or preconscious) mind. An unconscious monitoring system is said to patrol these thoughts and purposefully avoid them, but when this system gets tired or overwhelmed, these thoughts can surface. Research supporting this theory finds that unwanted thoughts are more likely when a person is under a high cognitive load, thus occupying the monitoring system. Meta-awareness provides another level at which to understand the prevalence of unwanted thoughts. This account suggests unwanted thoughts may simply lie in conscious thoughts, ones that
Meta-Awareness
occupy people’s minds but may not penetrate meta-awareness all the time. The monitor, in this account, would then be patrolling conscious thoughts looking for evidence of the unwanted. This formulation of unwanted thoughts can potentially tell researchers more about how unwanted thoughts are banished and where they go. Recovered Memories In a similar vein to this research on unwanted thoughts, meta-awareness also provides a useful framework for accounting for recovered memories of sexual abuse. These memories that individuals were previously unaware of, but come streaming back as if from nowhere, are difficult to understand through many traditional psychological theories. Further, there is a good deal of controversy over the truthfulness of some of these memories, an important issue given the often traumatic and sexual nature of some of these memories. Although there are good reasons to believe that recovered memories can be fictitious (particularly when they are recovered in therapy), many of these memories (at least those occurring outside of therapy) can be corroborated and thus treated as memories of actual events. Scrutiny of these corroborated recovered memories demonstrates that they are often consistent with the current conceptualization of meta-awareness. One way in which the notion of meta-awareness can help us understand recovered memories is with respect to people’s estimations that prior to the recovery, the memory had been unrecalled. The characterization of a memory as having been previously forgotten is itself a metacognitive judgment. One is making an appraisal of what one thinks one previously knew. However, if individuals often lack meta-awareness of the contents of their minds, then it is in principle possible that individuals who report recovered memories could in fact have known and thought about the experiences before, but simply failed to note this fact. Several lines of evidence are consistent with this interpretation of at least some recovered memories. First, a number of documented cases of recovered memories have involved individuals who are known to have talked about their experience during the period in which they believed
37
themselves to have been amnesic. Second, individuals with memories that are recovered out of therapy have been found to be particularly susceptible to failures in metacognitive judgments regarding previous episodes of recollection. In other words, they tend to be poor at determining what information they have previously recalled. Third, individuals with recovered memories of sexual abuse tend to be poor at noticing when they are having unwanted thoughts. Together these findings suggest that reports of recovered memories of abuse may at least sometimes be the consequence of a deficit in metaawareness, in which individuals lived for a period of time occasionally recalling their abuse, but failed to explicitly notice that they had done so. Meta-Awareness and Affect Temporal dissociations need not only deal with lapses of meta-awareness of thoughts, but also of feelings. For example, as the old children’s song goes, ‘‘If you’re happy and you know it, clap your hands.’’ This line certainly implies that it is possible to be happy, but to not have realized it yet. And indeed, the current understanding of meta-awareness suggests it is indeed possible to experience an affective state without having realized it yet. The experience of flow illustrates the dissociation between experience and meta-awareness of pleasure. One of the most effective ways of assessing the occurrence of pleasure in everyday life is the experience sampling technique in which participants are equipped with a pocket computer that intermittently probes them regarding what they are doing and how much they are enjoying it. Using this methodology, research has found that many of most pleasurable moments occur when individuals are in a state of flow. The flow state occurs when one is deeply absorbed in a task that is both highly challenging yet also accomplishable. What is so striking about research on the flow states is the fact that it indicates that individuals’ most positive experiences occur when they are not thinking about themselves, but are rather deeply absorbed in the activity itself. Indeed the flow state is so absorbing that individuals do not have the attentional resources to explicitly notice that they are happy at the time.
38
Meta-Awareness
As it seems that experience and meta-awareness of positive and negative affect can often become dissociated, then it stands that inducing metaawareness of affect can change the entire experience of an feeling-laden event. Recent research has shown that the induction of meta-awareness does alter the nature of such events. In one such study, subjects were instructed to continually rate their level of happiness while listening to hedonically ambiguous music. Researchers found that subjects who did rate their happiness throughout the study reported less postmusic happiness than subjects who did not continually rate their happiness throughout the study. Results from this study indicate that inducing meta-awareness of emotion can inextricably change the experience itself. Further, research from flow literature also strongly suggests that introducing meta-awareness during a flow state would interrupt the state and any positive feelings that go along with it.
Translation Dissociations When reappraising the contents of consciousness, there is a chance that the reappraisal might not be perfectly veridical to the actual experience. These experiences are classified as translation dissociations because there has been a break down of the translation of the experience to meta-awareness. The likelihood of a translation dissociation is particularly great under three sets of circumstances. First, if the experience is essentially nonverbal and a person attempts to verbally reflect on it; then there is an increased chance that he or she will get it wrong. Second, a person could be especially motivated to misrepresent an experience. And third, a lay theory about how an experience should be could lead to a reinterpretation that is unfaithful to the actual experience. Attempting to verbalize a nonverbal experience can lead to a misrepresentation of that experience, an effect that is often labeled as ‘verbal overshadowing.’ The hallmark in finding the verbal overshadowing literature deals with the verbalization of faces. Faces are known to be represented holistically in the mind, a quality that makes them difficult to verbalize completely. Numerous studies have shown that subjects instructed to verbally describe a face are poorer
at recognizing that face later, as opposed to subjects who simply viewed the face without instructions to verbalize it. Verbal overshadowing is likely due to the recoding of the visual image into words, a modality that lacks the intricacies and nuances to properly describe the complex, holistic nature of a face. In this situation, the meta-aware reappraisal of an experience interferes with the task at hand. The verbal overshadowing effect is not specific to faces, and generalizes to other areas of perceptual memory. Several studies have shown that many visual stimuli that defy words are vulnerable to verbal overshadowing effects. The detrimental effects of verbalization include stimuli such as color and shapes. Beyond visual memories, verbalization can also interfere with other perceptual modalities, such as audition and taste. One such study demonstrated that untrained wine enthusiasts had poorer memory for the wine they tasted if they attempted to describe wine-tasting experience. These wine drinkers were ostensibly well accustomed to tasting wine, thus possessing the ability to detect various nuances present in the wine. They did not, however, have the vocabulary or expertise to properly describe the wine, leading to a reappraisal that did not do a good job of describing their perceptual experience. On the other hand, trained wine writers showed an improvement of memory when verbally describing the wines, likely due to their ability to accurately and fully verbalize their wine-tasting experience. A meta-awareness of an experience can both aid and obstruct a task depending on one’s ability to take stock of and reanalyze that experience. Constructing a meta-awareness of a nonverbal experience applies not only to the domain of perceptual memory, but to other areas of nonverbal cognition. In the domain of judgment and decision making, verbally describing or rating the qualities of the available choices to be made can lead to substandard outcomes. This detrimental effect of verbalization has been reliably demonstrated in contexts where the choices to be made are affect laden, such as taste and visual appeal. For instance, researchers performed a study veiled as an inquiry into consumer judgments of strawberry jams. They asked some participants (verbalizers) to taste the jams and then list their reasons for liking or not liking the jams, as well as analyzing their reasons.
Meta-Awareness
Control participants tasted the jams, but did not list or analyze their thoughts about the jams. These researchers found that participants who did not list and analyze their reasons made judgments that were more similar to that of expert jam raters (from consumer report magazines) as compared with verbalizers. Other research has shown that analyzing reasons can also promote choices that yield decreased postchoice satisfaction. The verbal overshadowing effect and other evidence of the sometimes problematic nature of meta-awareness have been shown to be reliable phenomena in the psychological literature. Still, it is important to note that verbal reflection is often helpful. This is so when the experience is easily translated into words. Research has shown that logical problem-solving is aided by verbalization, as are situations when the verbalizer has the expertise or training to create accurate verbalizations. Recall that wine drinkers trained to accurately describe wine show a better memory for wines they verbalize. Regardless, the influence of verbal reflection on memory, judgment, and decision making illustrates the importance of a full, rich, and most importantly, accurate meta-awareness of an experience. While sometimes an experience is difficult to describe, other times individuals may simply be motivated to misrepresent the experience to themselves. It has been shown that homophobic individuals may not want to recognize when they are aroused in response to depictions of homosexual acts. In other words, they may consciously experience arousal, but not be meta-aware of these feelings due to a strong motivation to suppress such information. Even if the motivation to develop an accurate meta-awareness exists and verbal reflection is not a problem, other barriers exist to cause translation dissociations. One such barrier is a faulty theory about what a meta-awareness should contain. That is, people may have a faulty theory about what they should be feeling or thinking in a particular situation, which in turn colors their appraisal of their actual experience. A compelling example of this comes from people’s reports of their experience of catching a ball. Most people believe that as they watch a ball, their eyes first rise and then go down following the trajectory of the ball. Indeed,
39
this is the case when one watches someone else catch a ball. However, when people catch a ball themselves, they actually maintain the ball at precisely the same visual angle. Nevertheless, when people who just caught a ball are asked what they experienced, they rely on their theory of experience rather than on what they actually did.
The Double-Edged Nature of Meta-Awareness The various dissociations between meta-awareness and experiential consciousness serve to show that meta-awareness can have both beneficial and detrimental effects. Meta-awareness allows humans the ability to monitor and control their thoughts, which in turn make goal-driven behavior possible. To illustrate, imagine a pilot attempting to land a plane: Without the ability to take stock of his thoughts he would be unable to stave off, or even recognize a potentially disastrous bout of mind wandering during this important process. Phenomena such as verbal overshadowing, however, show that meta-awareness can be incongruent with success at certain tasks. These times when it is perhaps better to not be meta-aware are summed up well by the old children’s story of the centipede and his beautiful dance. As the story goes, there was a centipede who, using all his 100 legs, did a wonderful dance that all of the other creatures were jealous of, none more so than the tortoise. The tortoise devised a devious plan to derail the centipede, by sending him a letter asking how the dance was performed. Did he move his 28th leg before the 39th? And was that followed by the 12th and 72nd at the same time? The centipede began to think about what exactly he did and was never able to dance again. This classic story illustrates well the danger of paralysis through analysis, or in other words, the danger of becoming meta-aware (i.e., dissecting the dance) of a conscious experience (i.e., the dance itself). Still, the field of mindfulness may shed light on how to achieve analysis without the accompanying paralysis. Research on mindfulness meditation training demonstrates a process that may focus on the benefits of meta-awareness, while avoiding its pitfalls. Like other forms of meditation, mindfulness
40
Meta-Awareness
mediation deals with becoming aware of consciousness. Unlike other types of mediation, it does not involve focusing on a stimulus, but instead strives for a broad observation of several facets of experience. Mindfulness-based interventions have been shown to help decrease stress, anxiety, and lead to several positive outcomes. These results indicate that there may be different sorts of meta-awarenesses to be experienced, and one that involves being broadly mindful may avoid some of the pitfalls of other varieties.
intuitive mindful awareness that develops through contemplative practices such as meditation. As research continues on the nature of these different flavors of meta-awareness, students and researchers can expect a deeper knowledge of how to harness this fundamentally human ability. See also: The Control of Mnemonic Awareness; Philosophical Accounts of Self-Awareness and Introspection; Self: The Unity of Self, Self-Consistency.
Suggested Readings Conclusion Meta-awareness is likely unique to humans. It represents the ability to step outside one’s own thoughts and reflect upon them. This function of meta-awareness endows a high degree of flexibility, such as the ability to monitor and control cognition. While meta-awareness confers incredibly flexibility, it does not always work as one might wish. For one, meta-awareness is limited and costly to sustain. Such can be seen in temporal dissociations such as mind wandering. Attention drifts during these lapses in meta-awareness allow individuals to get off track, and perhaps more importantly fail to notice they are off track. A lack of meta-awareness may even be confused with the feeling of having never known, such as with the problem of recovered memories of sexual abuse. Meta-awareness is also, by nature, imperfect. When consciousness is reanalyzed and reappraised, mistakes can be made – seldom is the time when the copy is as robust as the original. Translation dissociations such as verbal overshadowing illustrate this pitfall of meta-awareness. In such cases a verbal recoding of a perceptual experience, like the image of a face or the bouquet of a wine, are not as good as the original. This mismatch between experience and reappraisal can cause problems with memory as well as judgments and decisions. Translation dissociations can also occur if the motivation to accurately reflect on a memory is not present, or if theories about how an experience should be are faulty. Although the two-edged nature of metaawareness may never be completely avoidable, some forms of meta-awareness may be more beneficial than others. For example, explicit analytic reflection may be more disruptive than the more
Csikszentmihalyi M and LeFevre J (1989) Optimal experience in work and leisure. Journal of Personality and Social Psychology 56(5): 815–822. Flavell JH (1979) Metacognition and cognitive monitoring: A new area of cognitive-developmental inquiry. American Psychologist 34: 906–911. Lambie JA and Marcel AJ (2002) Consciousness and the varieties of emotion experience: A theoretical framework. Psychological Review 109: 219–259. Mason MF, Norton MI, Van Horn JD, Wegner DM, Grafton ST, and Macrae CN (2007) Wandering minds: The default network and stimulus independent thought. Science 315: 393–395. Nelson TO (1996) Consciousness and metacognition. American Psychologist 51: 102–116. Norman DA and Shallice T (1986) Attention to action: Willed and automatic control of behaviour. In: Davidson RJ, Schwartz GE, and Shapiro D (eds.) Consciousness and Self-Regulation: Advances in Research and Theory. New York: Plenum Press. Schooler JW (2002) Re-representing consciousness: Dissociations between consciousness and meta-consciousness. Trends in Cognitive Science 6: 339–344. Schooler JW, Ariely D, and Loewenstein G (2003) The pursuit and monitoring of happiness can be self-defeating. In: Carrillo J and Brocas I (eds.) Psychology and Economics, pp. 41–70. Oxford, GB: Oxford University Press. Schooler J and Schreiber CA (2004) Experience, metaconsciousness, and the paradox of introspection. Journal of Consciousness Studies 11(7–8): 17–39. Smallwood J, Beach E, Schooler JW, and Handy TC (2008) Going AWOL in the brain: Mind wandering reduces cortical analysis of external events. Journal of Cognitive Neuroscience 20: 458–469. Smallwood J and Schooler JW (2006) The restless mind. Psychological Bulletin 132: 946–958. Teasdale JD, Moore RG, Hayhurst H, Pope M, Williams S, and Segal Z (2002) Metacognitive awareness and prevention of relapse in depression: Empirical evidence. Journal of Consulting and Clinical Psychology 70: 275–287. Winkielman PW and Schooler JW (in press) Unconscious, conscious, and metaconscious in social cognition. In: Strack F and Fo¨rster J (eds.) Social Cognition: The Basis of Human Interaction. Philadelphia: Psychology Press.
Meta-Awareness
41
Biographical Sketch
Jonathan W. Schooler is a professor of psychology in the University of California at Santa Barbara. He pursues research on consciousness, memory, the relationship between language and thought, problem-solving, and decision-making. A cum laude graduate of New York’s Hamilton College, Dr Schooler earned his PhD in psychology at the University of Washington in 1987. He joined the faculty of the University of Pittsburgh as an assistant professor, and in 2004 he accepted and held the position of a full professor and the Canada Research Chair in Social Cognitive Science until 2007 when he accepted his current position. A fellow of the Association for Psychological Science, Dr Schooler has been the recipient of three Akumal Scholar Awards from the Positive Psychology Network, an Osher Fellowship given by the Exploratorium Science Museum, and a Lilly Foundation Teaching Fellowship. His work has been supported, among others, by the National Institute of Mental Health, the Office of Educational Research, and Canada’s Social Sciences and Humanities Research Council. He currently is on the editorial boards of Consciousness and Cognition and Social Cognitive and Affective Neuroscience. Dr Schooler is the author or coauthor of more than 100 papers published in scientific journals and the editor (with J.C. Cohen) of Scientific Approaches to Consciousness.
Jason M. Chin is a PhD student at the University of British Columbia studying social psychology. He completed his undergraduate work at the University of Virginia, where he majored in economics and psychology, graduating cum laude. At the University of Virginia, Mr Chin researched the psychological underpinnings of aversion to risk, suggesting that it stems from a tendency to overweight future negative emotional reactions. He earned his MA in psychology in 2005 from his present university for his work on verbal overshadowing, and he currently studies motivational and informational models explaining prosocial behavior. Mr. Chin’s research has been supported by the University of British Columbia and the Izaak Killam Memorial Trust. A member of the Society for Personality and Social Psychology, he has presented his research at several conferences and is a coauthor of two scholarly publications. Mr. Chin has also served as a reviewer for several psychological publications.
The Mind–Body Problem M Rowlands, University of Miami, Coral Gables, FL, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Consciousness (phenomenal) – What it is like to have or undergo an experience; the ways things seem to feel to you when you have an experience. Dualism – The claim that the mental phenomena are nonphysical. This can be advanced as a claim about token mental phenomena – the result typically being a form of ‘substance dualism.’ Or it can be advanced as a claim about types of mental phenomena – resulting in what is known as ‘property dualism.’ Functionalism – The claim that mental phenomena are defined by their functional role, that is, by their place in a systematic network of causally related mental phenomena. Identity theory – The claim that mental phenomena are identical with (i.e., one and the same thing as) physical phenomena. This claim can take two forms. According to the ‘type-identity theory,’ types of mental phenomena are identical with types of physical phenomena. According to the ‘token identity theory,’ mental phenomena tokens are identical with physical phenomena tokens. Intentionality – The aboutness of mental mental phenomena; the directedness of mental phenomena toward objects apparently distinct from them. Supervenience – A one-way relation of dependence or determination that obtains between properties. A property F is supervenient on a property G when, necessarily, any two objects that are identical with regard to G must also be identical with regard to F. Tokens – Individual, particular instances of kinds. Types – General kinds or categories.
Introduction: Mind–Body Problems The mind–body problem has been traditionally understood as the problem of understanding the relation between the mind and the body. However, since the nature of the body is usually assumed to be uncontroversial, this problem has found itself ruminating largely on the nature of the mind. The mind–body problem is, in effect, the problem of understanding the nature of the mind. However, since the conclusion of at least some of these ruminations is the rejection of the view of the mind as a substance distinct from mental states and processes, the problem can less tendentiously be formulated as follows: ‘what is the nature of mental phenomena?’ Upon analysis, however, even this question fragments into two distinct questions, and this fragmentation follows from the nature of mental phenomena. The concept of a mental phenomenon can be given a preliminary specification by way of ostension. Mental phenomena include beliefs, desires, thoughts, feelings, hopes, fears, expectations, pains, emotions, experiences, and so on. It is common to divide up these phenomena into two categories: sensations and propositional attitudes. Sensations are mental items defined by ‘what it is like’ to have them or (equivalently) by their possessing phenomenal consciousness. Pain is the paradigm example of a sensation: to be in pain is for things to feel a certain way to the subject of the pain. Propositional attitudes, on the other hand, are mental items defined by their possession of propositional content. Beliefs, desires, and thoughts are paradigm examples of such states. Beliefs are attributed to a subject using an embedded ‘that’clause. Jones believes that p, where p stands for some or other proposition. For example, she believes that the cat is on the mat. The sentence ‘the cat is on the mat’ has a particular content, or expresses a particular proposition. The content of Jones’ belief is identical with the proposition
43
44
The Mind–Body Problem
expressed by the sentence that we use to attribute the sentence to Jones. The same is true for thoughts, desires, hopes, fears, expectations, anticipations, and so on. This is why such items are referred to as propositional attitudes. The distinction between sensations and propositional attitudes is, therefore, a distinction between mental items that are defined by their possession of phenomenal consciousness and those items defined by their possession of propositional or semantic content. While this distinction is common, it is not entirely clear that it is either adequate or illuminating. With regard to the issue of adequacy, for example, it seems far from clear that the distinction exhausts the category of mental phenomena. Experiences, for example, seem to occupy a curiously ambiguous position – being characterizable both by their propositional content and what it is like to have them. Some suggest that this ambiguity can be accommodated by recognizing that the content possessed by experiences is of a different kind from that possessed by propositional attitudes: the latter content is conceptual, the former merely nonconceptual. However, this suggestion faces significant difficulties. For the distinction to be illuminating we would have to understand in what the difference between possession of phenomenal consciousness and propositional content consists. However, this difference is far from transparent. Often, the distinction is glossed in terms of the idea that propositional attitudes are about other things, whereas sensations are not. Or, more plausibly, propositional attitudes are about other things in a way that sensations are not. However, this too is controversial. An increasingly popular position today, for example, is a form of representationalism or inseparatism that holds that phenomenal consciousness consists in a certain type of (representational) content. While this view is most commonly applied to the content of experiences, attempts have been made to apply it to sensations also. This article is not the place to address these issues. These sorts of controversies, and these failures of transparency, are endemic in discussions of the mind–body problem. Here, we can simply note them and work around them in so far as this is possible. In this vein, we should note that mental phenomena seem to be characterized by two sorts
of property. First, many mental states – arguably all, but at the very least many – are defined by the fact that ‘there is something that it is like’ to have or undergo them. This, I shall henceforth refer to simply as consciousness. This is most obviously true of sensations and experiences. The propositional attitudes, on the other hand, can exist in nonconscious – or perhaps even unconscious – form. Nevertheless, it is arguable that for a state to be mental it must at least be the sort of thing that can, in appropriate circumstances, become conscious. Second, many mental states – arguably all, but at the very least many – are defined by the fact that they are about other things in such a way that they possess content. This feature of mental states – their aboutness – is traditionally referred to as their intentionality. If we ignore the rather vexed issue of which states are principally characterized by which – consciousness or intentionality – and the even more vexed issue of the relation between the two – then the claim that consciousness and intentionality are defining features of mental items seems a relatively safe one. Therefore, I shall label consciousness and intentionality as the essences of the mental. The mind–body problem is the problem of answering this question: ‘what is the nature of mental phenomena?’ Now, however, we see why this question fragments into two: it can be understood as a question about mental items, or as one about the essences of mental items. That is, the question must be replaced with two: 1. What is the nature of mental items? 2. What is the nature of the defining essences of the mental – consciousness and intentionality? I shall refer to the first question as the item problem and the second as the essence problem. The item problem can, in turn, be divided into two because the concept of a mental item can be divided into two. Thus, the question can be understood as a question about mental tokens or one concerning mental types. Mental types are kinds; mental tokens are (usually understood as) concrete dated, unrepeatable instances of those kinds. Pain, in general, would be a type of mental item. The event of pain you suffer at 14.05 p.m. on 9 August 2007, and which lasted for precisely 15 s, would be a token of that type. In the literature,
The Mind–Body Problem
tokens are often referred to as events, whereas types are referred to as properties. The latter terminology is unfortunate, since it invites confusion of types and essences – consciousness and intentionality are also often referred to as properties possessed by mental items. To avoid this, I shall eschew talk of properties, and distinguish clearly between mental types and essences. Pain is a mental type. What it is like to be in pain is a mental essence.
The Item Problem As we have seen, the item problem is really two problems: (1a) What is the nature of token mental items? (1b) What is the nature of mental item types? (1a) Mental Tokens. If we ignore some of the more idiosyncratic proposals thrown up by history, answers to (1a) divide into two primary sorts. The first is that mental tokens are identical with physical tokens. This view is known as the token-identity theory. This theory is one version of materialism: the view that mental phenomena are ultimately physical. The other is that mental tokens are nonphysical events. This view is known as dualism. According to dualism, token mental events are nonphysical. But events are often thought of as changes of an object with respect to a property or properties at a time or through a period of time. Therefore, the claim that mental events are nonphysical is typically understood as the claim that they consist in changes in a nonphysical object or substance. The result is what is known as substance dualism. According to substance dualism, token mental events are identical with events occurring in an immaterial substance. This view is still principally associated with Rene´ Descartes. Descartes’ dualism – Cartesian dualism – was shaped by a particular conception of the defining essence of the mental – for Descartes this essence is thought – and his insistence that this defining essence could not be realized in any physical structure or mechanism. The essence of physical things is extension. Mental items are, therefore, nonextended. One may question whether
45
Descartes was correct in his conception of the essence both of the mental and physical. But any form of substance dualism is going to regard mental and physical tokens as being of essentially different sorts. And so it is possible to eschew talk of substances and simply formulate dualism in terms of the nonidentity of mental and physical tokens. That is, dualism can be understood as the view that token mental events, states, or processes are distinct from any physical events, states, or processes. We might call this a dualism of particulars. Both substance dualism and this dualism of particulars are attempts to answer (1a). Problems with dualism (of both kinds) typically divide into two related sorts. The first concerns understanding exactly what the mind is. In its typical formulations, dualism is a negative thesis: token mental events are not identical with token physical events; mental substances are not identical with physical substances. This sort of formulation is often supplemented by additional claims dualists make about what the mind does not have: is not extended, has no physical properties, is not subject to physical laws of nature. Negative definitions such as this are, by themselves, obviously inadequate. First, they do not even suffice to fix the content of the dualist thesis: it is not, in general, possible to explain or identify what an item is by providing a list of everything that item is not. Second, purely negative characterizations invite the charge that dualism is not even in the business of providing a theory of the mind but is, rather, a theoretical hole just waiting for a theory to be put in its place. That is, the worry is that dualism makes no genuine attempt to say what the mind is, how it is constructed, and what laws or principles govern its operation. Of course, dualism is not restricted to providing a purely negative account of the mind. Then, however, the worry is that the sorts of positive things the dualist is able to say about the mind are insufficient to distinguish dualism from competing views. The materialist, for example, might accept with equanimity Descartes’ assertion that the essence of the mental is thinking. For thinking is something that the mind does – it expresses a functional concept – and this, it is, so the materialist would argue, as amenable to materialist analysis just as much as it is to dualist explanation.
46
The Mind–Body Problem
The second sort of problem afflicting dualism concerns how to reinsert the mind back into the natural order once it has been identified as being so different from it. It is a commonplace, for example, that mind and world can affect each other. This is often understood in causal terms. Things occurring in the world can cause things to occur in the mind, and vice versa. However, what is the basis of this causal interaction? The danger is that dualism makes mental events so different from physical ones that it becomes impossible to understand how they can causally affect one another. This is, in fact, the traditional objection to Descartes’ dualism. According to Descartes, the mind has no extension: it does not occupy space. But our ordinary concept of causation seems bound up with the concept of space, in particular, with the idea of spatial contiguity. It may well be that our ordinary concept of causation is naı¨ve. But, if so, it is incumbent on the dualist to supply an alternative in terms of which mental–physical causal interaction is intelligible. This traditional objection to Descartes’ dualism is, in fact, an instance of a wider worry: if the mind is as different from anything physical as the dualist maintains, then how do we explain its manifest interconnections – causal and developmental – with the physical world? These worries might lead one to suppose that materialism – in the form of the token-identity theory – wins by default. Most arguments for materialism rest, in one way or another, on the manifest connections between mental and physical phenomena. For example, in developing his anomalous monism – an influential version of the token identity theory – Donald Davidson restricted his argument to mental events that causally interacted with physical events. Davidson was convinced that he could demonstrate the identity of all such token mental events with token physical events, but remained neutral on the status of any mental event that did not so interact. However, the token identity also faces its problems: one concerning the idea of identity, the other concerning the idea of the physical. Firstly, it is not clear why mental event-tokens would have to be the sorts of things that that can be identical with physical events tokens. To see why, consider another category of events – economic events. When I pay for something in a shop, where does
the event of paying take place? Suppose, for example, I am paying by credit card. Does the event of paying occur at the moment of hand-to-handtransfer of the card? Or does it incorporate the movement of my wallet from my pocket to my hand? Or, does it incorporate the movement of the teller to the machine? Or does it occur during subsequent bank-to-bank transfers of funds? It is not clear that there is any principled way of answering these questions: economic events may simply be spatially ill-defined. If this is true, then it is difficult to understand the sense in which economic events could be identifiable with physical events, if we assume these are spatially well-defined. The problem for the token-identity theory then, is this: the theory implicitly presupposes a fairly restrictive conception of mental events – that they are not at all like economic events, for example. This presupposition has not been subjected to the appropriate amount of critical scrutiny and, to this extent, is unjustified. Faced with this sort of problem, the materialist might be tempted to abandon the idea of mental–physical token-identity in favor of some sort of mental–physical composition thesis: while mental events are not identical with physical events, they are composed of physical events (with the added proviso that a composite event is not identical with the sum of its parts). This response, however, does not seem to avoid the problem. If physical events are spatially determinate then no matter how we combine physical events, if the composition is mereological we will still end up with a spatially determinate composite event and. And if mental events are not spatially determinate, then they cannot be composed of a spatially determinate sum. In other words, this compositional version of materialism has not jettisoned the conception of mental events that led to the objection to the token-identity theory. The other option for the materialist is to abandon the idea that physical events are spatially determinate. This is not the same as assuming that physical events are spatially indeterminate, it is simply to remain noncommittal on the issue. By adopting this strategy, the identity theory is no longer vitiated by an unjustified assumption concerning the nature of mental events: the theory can now remain noncommittal about the spatial character of both mental and physical events.
The Mind–Body Problem
However, this strategy raises problems of its own. In particular, it is not clear how much substance there remains to the idea of identity (or for that matter composition). In particular, the theory now faces the formidable problem of understanding the sense in which one potentially spatially indeterminate event can be identical with another equally spatially indeterminate event. The second problem concerns what it is for something to be physical. There is actually a cluster of worries here, but perhaps the most significant one was identified by Bertrand Russell as long ago as 1927. It is reasonable to suppose that physics – the basic science of matter – will tell us what it is for something to be physical. Indeed, if physics does not tell us this, it is difficult to see what would. However, Russell notes that the knowledge of matter supplied by physics is ‘abstract,’ ‘purely formal,’ ‘merely a matter of structure’. He writes: Physics is mathematical not because we know so much about the physical world but because we know so little; it is only its mathematical properties that we can discover. For the rest, our knowledge is negative . . . The physical world is only known as regards certain abstract features of its space-time structure – features which, because of their abstractness, do not suffice to show whether the physical world is, or is not, different in intrinsic character from the world of mind. (1948: 240)
Physics proceeds by the identification of an abstract mathematical formulism that will detail the relations between things. However, intuitively, not definitely by any means, but this is the way we tend to think about things – there are not just relations. When things enter into relations, they do so because of the intrinsic natures of the things that are related. Relations are ontologically dependent on things, and things have natures that explain the relations those things enter into. However, of such natures – the natures that ground the relations between things – physics, Russell argues, can tell us nothing. The moral, for present purposes, is this: if you think that what makes something physical is its underlying nature, a nature that provides the basis of the relations that physical things enter into, relations described by physics, then physics gives us no idea what it is to be physical. The dualist’s failure to tell us anything
47
significant about the mental is, therefore, arguably mirrored by the materialist’s failure to tell us anything significant about the physical. And if this is so, any purported victory of materialism over dualism has a distinctly hollow ring. (1b) Mental Types. Let us overlook these problems, assume we do understand what it is to be physical, and continue with the development of a materialist conception of mental phenomena. It is fairly clear that a token-identity theory will not, by itself, do all the required work. The tokenidentity theory, in itself, is a very weak doctrine. To begin with, the identity of mental and physical tokens is compatible with a dualism of mental and physical types: the claim that every mental eventtoken is identical with a physical event-token is compatible with the claim that mental types are distinct from and irreducible to physical types. This creates a problem for the materialist. It would be possible for two individuals to instantiate exactly the same physical types but mentally have nothing in common at all. For example, two molecule-for-molecule duplicates might share no mental types at all. Each mental token they undergo is identical with a physical token, but the mental types that these tokens instantiate are, in each case, different. The condition of token-identity has been satisfied, but, from a materialist perspective, there clearly seems to be something missing. What is missing is the idea that the way the world is physically determines the way it is in every other respect. If we do not at least secure this claim there is little substance left to the idea of materialism. And to secure this claim we must tighten the relation between mental and physical types: we must make the former in some way dependent on the latter. One way of doing this is to suppose that mental types straightforwardly reduce to physical types. The type-identity theory is a common way of understanding this reductive thesis. According to this theory, mental types are identical with physical types. For example, pain – pain as a type, not a specific instance of pain, but pain in general – might be identified with C-fiber firing. Pain, so the idea goes, is identical with C-fiber firing in much the same way that water is identical with H2O; in much the same way that heat is identical with molecular motion, or lightning is identical with an
48
The Mind–Body Problem
electrical discharge to earth from a cloud of ionized water particles. This type-identity theory is thought by many to be fatally susceptible to what is known as the problem of variable realization. As Kripke has emphasized, identity is a necessary relation in this sense: If X is, in fact, identical with Y, then necessarily X is identical with Y. X could not have been nonidentical with Y. For example, if water is, in fact, H2O, then water could not have been anything other than H2O. We might think that it could. Why could not water be identical with some entirely different chemical substance? But to entertain this possibility is simply to describe a situation in which there is a substance that looks, tastes, feels, etc., exactly like water but really is something else. This thesis of the necessity of identity creates problems for the type-identity theory. For it certainly seems possible that mental types can, in different creatures, or even in the same creatures at different times, be realized by different physical types. Imagine, for example, a Martian who felt pain in exactly the same way that we did, but where this feeling was realized not by c-fiber firing but by some entirely different sort of neural state. What is decisive is determining whether the Martian is in pain is the way things feel to it when it undergoes the event. If things feel the requisite way it is in pain is pain irrespective of the state of its neural machinery – or whether it even has something we would recognize as neural machinery. This argument does not depend on the actual existence of Martians or other extraterrestrials. Their genuine possibility is enough. If, as the principle of the necessity of identity claims, identity is a necessary relation, then the mere possibility that pain might, in different creatures, be realized by different neural states or structures, is enough to preclude the typeidentification of pain with C-fiber firing. These sorts of reasons – the possibility of variable realization coupled with the necessity of identity – have led to widespread, if not universal, rejection of the type-identity theory. And if the type-identity theory is, indeed, untenable we need another way of formulating the idea that mental types are dependent on physical types. The usual response, here, is to appeal to the concept of supervenience. The idea of supervenience is, in essence, a simple one: it is a one-way relation of dependence.
With regard to the relation between mental and physical types, the idea of supervenience can be expressed thus: mental types supervene on physical types if and only if any two creatures that are the same with regard to the physical types they instantiate must also be identical with regard to the mental types they instantiate. Equivalently, if a creature changes with regard to the mental types if instantiates, then it must also change with regard to the physical types it instantiates. The dependence is one way. It does not follow, that if any two creatures identical with regard to the mental types they instantiate must also be identical with regard to the physical types they instantiate, or that a creature that changes with regard to the physical types it instantiates must also change with regard to the mental types it instantiates. It is the fact that the concept of supervenience expresses only a one-way relation of dependence that distinguishes it from the type-identity relation. The concept of supervenience, while in essence a simple one, is notoriously difficult to pin down, and many journal and book pages have been devoted to a delicate juggling of the scope and strength of modal operators – flagged in the above definition by the term ‘must’ – required to get the concept ‘just right.’ We do not have the space here to pursue these issues. Instead, let us pursue a more pressing issue. What reason, if any, do we have for supposing that this supervenience thesis is true? What reason, that is, do we have for supposing that mental types supervene on physical ones? It is precisely here that the importance of the view known as functionalism can be properly appreciated. According to functionalism, mental phenomena are identified according to their causal or functional role. Consider an analogy. A carburetor is a physical object located somewhere in the innards of a car’s engine (or older cars anyway – fuel injection systems have replaced them in more recent models). What is a carburetor? Or, more precisely, what makes something a carburetor? The answer is that a carburetor is defined by what it does. Roughly, it is something that takes in fuel from the fuel inlet manifold, takes in air from the air inlet manifold, mixes the two in an appropriate ratio, and sends the resulting mixture on to the combustion chamber. It is fulfilling this role that makes something a carburetor, and
The Mind–Body Problem
anything that fulfils this role in a car thereby counts as a carburetor. Most carburetors tend to look pretty similar. But this is at best a contingent fact, because it does not matter what a carburetor looks like as long as it fills this role. The details of its physical structure and implementation are of secondary importance compared to the role it fills, for it is filling this role that makes something a carburetor, and not the details of its physical structure or implementation. Of course, not every physical thing is capable of playing the role of a carburetor. A lump of Jell-O inserted into your car engine would have a hard time mixing fuel and air – or doing anything else for that matter. A lump of Jell-O is simply not the right sort of thing for fulfilling the functional role of a carburetor. So, the details of how the functional role is physically implemented are not irrelevant. But as long as you have a suitable physical structure – one that is capable of fulfilling the role of a carburetor, then it does not matter what it is as long as it, in fact, fulfills this role. If it does, then it is a carburetor. Functionalism takes a similar view of the nature of mental properties. That is, such properties are defined by what they do – by their functional role. What is it that mental phenomena do? Fundamentally, they relate to each other, to perception and to behavior in various complex, but in principle analyzable, ways. Take a belief, for example, the belief that it is raining. This is a belief that is typically caused by perception of certain environmental conditions, rain being the most obvious. Of course, perception of other environmental conditions might also produce the belief; for example, someone, unbeknownst to you, using a hosepipe outside your window. But rain is the most typical cause of the belief. The belief can, in turn, go on to produce certain sorts of behavior. Because of your belief you might, for example, carry an umbrella with you when you leave the house. However, the belief has these sorts of ramifications for your behavior not in isolation but only in combination with other mental states. You will carry the umbrella because you believe it is raining, but only if you also want to stay dry. Your desire to stay dry is necessary for your behavior too. And this belief-desire combination will produce your behavior only if you also believe that the umbrella will
49
keep you dry; only if you believe that it is not too windy to use an umbrella; indeed, only if you believe that what you have picked up is an umbrella, and so on. What emerges is a complex network of mental states, perception, and behavior. According to functionalism, each mental state is defined by its place in this network: by the relations in which it stands to perception, to other mental states, and to behavior. To specify the place of a mental state in this network is, according to functionalism, to define that state. Of course, any such definition would be grotesquely long. But, its practicalities aside, the strength of functionalism consists in giving us a general vision of what mental phenomena are. The vision is of mental phenomena forming a vast causal system – a system of interrelated causal connections – where each mental property is individuated by way of its place in this system. Functionalism provides a clear explanation for why mental types would supervene on physical ones. A mental type is defined by its functional role, and for any such role there is a physical type that fills it. Given the necessity of identity and the possibility of alternative physical types filling that role, we cannot identify the mental type with the physical type. But the filling, by a physical type, of the role that defines a mental type is sufficient for that mental type to supervene on the physical type. During the second half of the twentieth century, attempts to answer the item question gradually coalesced around a position that is, by now, sufficiently widely accepted to be regarded as the default position. This position is made up of three claims: (1) identity of mental and physical event-tokens, (2) supervenience of mental types on physical types, where this supervenience is underwritten by (3) a functionalist account of mental types. That this combination provides the default position on the item question is, of course, no guarantee that it is correct.
Consciousness and the Essence Problem Recent work on the mind–body problem has tended to focus on the essence – rather than the item – problem. And the apparatus designed to solve the item problem – token-identity, supervenience, and functionalism – allows us to make only
50
The Mind–Body Problem
limited headway on the essence problem. The essence problem can be developed for both putative defining essences of the mental: consciousness and intentionality. Due to constraints of space, however, I shall discuss only the former. Many suspect that the existence of phenomenal consciousness – what it is like to undergo an experience – provides a serious, perhaps even intractable problem for materialist accounts of the mind. The basis of this suspicion is to be found in an intuition, one eloquently expressed by Colin McGinn: How is it possible for conscious states to depend on brain states? How can Technicolor phenomenology arise from soggy grey matter? What makes the bodily organ we call the brain so radically different from other bodily organs, say the kidneys – the body parts without a trace of consciousness? How could the aggregation of millions of individually insentient neurones generate subjective awareness? We know that brains are the de facto causal basis of consciousness, but we have, it seems, no understanding whatsoever of how this can be so. It strikes us as miraculous, eerie, even faintly comic. Somehow, we feel, the water of the brain is turned into the wine of consciousness, but we draw a total blank on the nature of this conversion. (1991: 1)
I shall refer to this as the intuition. Probably the safest way of characterizing the arguments of those who think that consciousness is materialistically problematic is as a series of developments, explications, refinements, and defences of this intuition. If these developments, explications, refinements, and defences are successful, then collectively they add up to a serious obstruction to the project of accommodating consciousness into a materialist world view. The intuition has been refined, elaborated, and supported by several well-known thought experiments: imaginative scenarios whose purpose is to make the intuition particularly concrete and graphic. Here are two of the most famous: 1. Some bats navigate their environment by means of echo location: they emit high-frequency noises and use the pattern of echoes to build up a cognitive map of their environment. Thomas Nagel invites us to imagine we know everything there is to know about the neural mechanisms that allow the bat to do this. Still, Nagel intuits, we would not know what it is like to be a bat. This
intuition he develops into a general argument against materialism, an argument that turns on the idea of subjectivity. 2. Frank Jackson invites us to consider the case of Mary. Mary lives in an entirely monochromatic environment – never seeing any colors except black, white, and shades of grey. However, despite this impediment, she has become the world’s leading neuroscientist knowing, in particular, everything there is to know about the neurophysiology of color vision. That is, she knows everything there is to know about the neural processes involved in visual information processing, the physics of optical processes, the physical makeup of objects in the environment, etc., What she does not know, however, is what it is like to see color. Thus, Jackson argues, when she leaves her room for the first time and sees a bright red object, she learns something new – what it is like to see red. Now she knows something she did not know before, and this new knowledge could not have been constructed or derived from her previous knowledge. Jackson thinks this spells the demise of materialism. Much work has been done examining and evaluating these thought experiments. However, the thought experiments are predicated on the intuition, in the sense that the former are ways of making the latter graphic. Therefore, I shall here focus on the intuition itself. There are, broadly speaking three different stances one can adopt with regard to the intuition: (1) one can reject it, (2) one can accept it but restrict its consequences, or (3) one can accept it. I shall begin with (2), and then discuss (1) as a way of objecting to (2). Finally, I shall make a few brief remarks about (3).
The Epistemological Interpretation: The Explanatory Gap One thing that the intuition might be taken to show is the existence of an explanatory gap between mental and physical items. Perhaps the most comprehensive development of this idea is due to McGinn. According to McGinn, the explanatory gap is grounded in a conceptual poverty that is, in turn, grounded in a poverty of
The Mind–Body Problem
faculties. The explanatory gap arises because we do not have the requisite concepts to apply to the natural order – concepts that would allow us to see how it produces or constitutes consciousness. But this, in turn, is derived from our lack of the appropriate faculties – concept-forming capacities – that would allow us to acquire the requisite concepts. McGinn’s position is characterized by: (1) ontological naturalism: consciousness is a natural feature of the world, and (2) epistemic irreducibility: there is no – and probably can be no – explanation of consciousness available to us. McGinn’s development of his ‘mysterian’ position can be regarded as comprising three interwoven strands. The first strand is concerned with differences in the ways we know about consciousness and the natural world (including the brain). We know about consciousness through introspection. In this, consciousness is unique; our knowledge of the rest of the world is grounded in perception. The second strand develops the idea that consciousness has a nonspatial character. Because our knowledge of consciousness is grounded in introspection, consciousness presents itself to us as nonspatial. And, in this, consciousness is again unique. Thus an idiosyncratic mode of access to consciousness yields an idiosyncratic feature of consciousness. Because of this idiosyncratic feature of consciousness, we will encounter major problems trying to incorporate consciousness into the natural order. The third strand revolves around inherent limitations to our cognitive capacities. Our failure to incorporate consciousness into the natural order is not the disaster many have supposed. To the extent there is a disaster, it is an epistemic, not ontological one. The problems ultimately stem from natural limitations on our cognitive capacities, limitations which make the problem of consciousness insoluble for us, but not for a creature with the appropriate cognitive faculties.
Objections to the Explanatory Gap In his commitment to ontological naturalism coupled with epistemic irreducibility, McGinn’s position is a version of stance (2) with regard to the intuition: acceptance of the intuition combined
51
with restriction on its limitations. Specifically, the intuition is argued to have only epistemic, rather than ontic, consequences. Stance (1) – the rejection of the intuition – can best be pursued by way of critique of McGinn’s position. This position involves two logically distinct claims: 1. An explanation of consciousness must proceed by way of identification of a mechanism. If an explanation of consciousness required only correlations between neural and conscious states, there would be no deep problem of consciousness. Further, if this underlying mechanism is to explain consciousness, it must do so by eliciting in us a certain kind of insight: 2. The neural mechanism that explains consciousness must allow us to see how consciousness is produced by the brain. Accordingly, a genuine explanation of consciousness works only to the extent that it allays ‘the feeling of mystery that attends our contemplation of the brain–mind link’. Both (1) and (2) are not unassailable. One possible objection to (2) is that it involves a conflation of the concept of explanatory adequacy with what we might call epistemic satisfaction. Some explanations produce in us a feeling of epistemic satisfaction: a Eureka! feeling ‘Now I understand!’ or in more Wittgensteinian mode, ‘Now I can go on!’ The molecular explanation of the macroproperties of graphite provides a good example of an explanation likely to elicit this sort of feeling. Graphite consists in layers of carbon arranged into hexagonal rings. The atoms in each layer are covalently bonded to three neighboring atoms at an angle of 120 to each other. Within each layer, the covalent forces binding each atom to its neighbor are relatively strong. However, the layers themselves are bound together only by the very weak van der Waals forces. As a result, adjacent layers can slide over each other – resulting in the soft, flaky, nature of graphite, its ability to mark paper, act as a lubricant, etc. A focus on explanations of this sort might tempt us into thinking that the adequacy of an explanation is to be judged by whether it elicits in us a feeling of epistemic satisfaction. And this assumption is
52
The Mind–Body Problem
questionable. Consider, for example, the molecular explanation of solidity in terms of a rigid lattice structure of atoms held together by ionic binding. How, one might reasonably ask, can a solid object be made up mostly of empty space? How could such an item, for example, retain its volume? An obvious response is to explain away any lack of epistemic satisfaction in terms of our empirical ignorance – specifically, of relevant atomic or quantum level facts and laws. For example, we might explain the disposition of solids to retain their volume in terms of the characteristics of the specifically ionic bonding that seems to be responsible for this ability. Ionic bonding involves electron transfer of electrons, rather then merely their sharing, and so ionic bonds are very strong. But this merely pushes the problem back a step. Why should bonds that involve transfer of electrons be any stronger than bonds which merely involve their sharing? What reasons are there for supposing that this explanation will be any more epistemically satisfying than the original? We can push the explanation back further, and explain the salient characteristics of ionic bonding in terms of wave interaction, superposition, and so on. Perhaps, once we acquaint ourselves with the relevant laws of wave dynamics, then everything else will fall into place? But, once again, the same question will arise. Why must explanations cast at this level be any more epistemically satisfying than the original molecular explanation? Is it obvious, for example, why, waves should obey the laws of wave dynamics? More generally, why should the world, at a fundamental level, be an epistemically satisfying place? The dialectic, here, is tricky because McGinn will, of course, argue precisely that the world is not an epistemically satisfying place, at least not for us; and this is the basis of his mysterian position. The present point, however, is that a lack of epistemic satisfaction need, in itself, be no impediment to recognizing that something is an explanation of a given phenomenon, and an adequate one at that. We can accept that a wave dynamical account of a phenomenon such as solidity is both true and an explanation even if it does not produce in us – in any of us – the sort of feeling occasioned by the molecular explanation of the macroproperties of graphite. If this is correct, then explanatory adequacy is not a function of epistemic satisfaction:
explanatory adequacy does not consist in a specific inner process. Some explanations – ones that we recognize as adequate – possess a sort of inchoate, protoversion, of epistemic satisfaction: protoepistemic satisfaction. At the core of this concept is the notion of analogy. Many of our best theories have their origin in provocative initial analogy; one that may be seriously flawed, but subsequently proved to be a fruitful vehicle of understanding. Consider, again, the molecular explanation of solidity. While this may not occasion the sort of epistemic satisfaction elicited by other explanations, it does produce a certain form of enlightenment carried, to a considerable extent, by the relations between properties of the reduced domain and those of the reducing. Thus, suppose we accept that a given solid is composed of a lattice structure of atoms tightly bound together, each oscillating around a fixed point. We can then, with relative ease, accept that the addition of energy to this structure might increase the frequency of this oscillation. And then, also, that the addition of sufficient energy might increase the oscillatory frequency to such an extent that the bonds break down. And the addition of further energy might increase this breakdown further. So, if we accept that solids are made up of a rigid lattice structure of oscillating atoms, then we can also see that the difference between this sort of structure and one where the bonds are more diffuse is something like, somewhat analogous to, the difference between a solid and a liquid. And, in virtue of this sort of rough analogy the molecular explanation of solidity possesses a certain protoepistemic satisfaction. While it is plausible to suppose that any explanation we recognize as an explanation must elicit some or other psychological states in us, the precise nature of these states may vary considerably from one explanation to another – varying from, at one extreme, the full-blown ‘Eureka!’ feeling to, at the other, a nebulous, imprecise, and analogybased form of protoepistemic satisfaction. The latter form of understanding can then be reinforced by the sorts of social pressures characteristic of a scientific education. Consider, now, claim (1). This is the claim that mere correlation of neural and conscious states is not sufficient for an explanation of consciousness: that
The Mind–Body Problem
would require identification of a mechanism. The distinction between mechanisms and correlations is, however, a questionable one. Specifically, the sort of enlightenment provided by mechanisms consists in the breaking down a correlation into a structured series of smaller correlations, where each of the smaller correlations is more readily intelligible than the original one. Mechanistic explanation is not something radically different from or opposed to the identification of correlations. On the contrary, mechanistic explanation is a specific form of correlation-based explanation. It may be that a correlation between two items can be rendered intelligible by the uncovering of an underlying mechanism. But this is not to replace the correlation with something fundamentally different; it is to break down, and thus explain, the correlation by means of further correlations. With these points in mind, the best case that can be made for reductive naturalism, and hence against the intuition, involves three claims: 3. There is no fundamental opposition between mechanistic explanation and the identification of correlations. 4. The explanatory adequacy of correlation-based explanation does not require that it elicit in us epistemic satisfaction in any full-blooded sense. To these principles, we can add a third: 5. There is not an explanation of consciousness. Rather, there are many such explanations – as many as there are features of consciousness that require explanation. Even in the case of properties such as solidity, there is not necessarily, any such thing as the explanation of solidity. Rather, there seem to be at least two. There is an explanation of the disposition to resist deformation and an explanation of the disposition to retain volume. Since not all rigid structures retain volume then an explanation of the former is not, in itself, and explanation of the latter. We might expect this general point to be reiterated in the case of consciousness. The concept of consciousness almost certainly fragments, upon analysis, into several distinct concepts, including phenomenality, subjectivity, nonrelationality, and so on. If this is so, then it is likely that separate explanations will be required for each of them.
53
With (3)–(5) in mind, consider the much maligned claim of Crick and Koch to have explained consciousness in terms of 40 Hz oscillations in the sensory cortex. Taken in itself, such a claim is, of course, laughable. However, what 40 Hz oscillations might be able to play a role in explaining is not consciousness as such, but one of its features: its gestalt character or, as we might put it, its all-atonceness. Conscious experience is not presented serially – like, for example, a description of that experience in the form of a sentence. It is presented all at once. Part – though presumably not all – of explaining this feature of consciousness almost certainly involves explaining the brain’s capacity for binding information together into a unified whole. And this is precisely what the identification of a single oscillatory frequency might enable us to understand. It would do this not in the sense of providing us with full-blooded epistemic satisfaction with regard to the production of consciousness. Rather, it may yield a form of protoepistemic satisfaction with regard to one aspect of consciousness. That is, we can see that the gestalt character of experience is something like, somewhat analogous to, disparate information that has been bound together in various ways. Consequently, we can understand, in a somewhat nebulous manner, that changes in the quantity and types of information that are bound together at any given time might systematically vary with changes in the content of the visual gestalt. In short, the best case that can be made against the intuition, certainly in the form presupposed by McGinn, involves arguing (a) that McGinn is committed to principles (1) and (2), but that (b) these principles should be rejected in favor of principles (3), (4), and (5), and then arguing that (c) principles (3), (4), and (5) are precisely the sort of principles that drive, in an admittedly nonreflective manner, current scientific research on consciousness.
The Ontological Interpretation: Consciousness as Outside the Natural Order If one does not accept the arguments against intuition, then it is possible to regard this intuition as
54
The Mind–Body Problem
indicative of something more than merely an epistemic hiatus. That is, one can take intuition as showing that consciousness is not a natural feature of the world. This can be understood in two very different ways. Chalmers understands the problem of consciousness as one stemming ultimately from the ontological poverty of the sciences of consciousness (although this is reflected in an associated conceptual poverty also). To rectify this, Chalmers advocates what he calls naturalistic dualism. He allows that there is an explanation of consciousness we can understand. But to get this explanation we have to be willing to expand our catalogue of basic substances. So, while there is an ontological gap between consciousness and the basic substances currently countenanced by science – and in this sense consciousness is not part of the natural world as this is currently understood – this gap can be bridged by suitable adjustment of what we take the basic substances to be. The ontological gap between consciousness and the natural world is one that can be bridged through appropriate adjustment of the sciences of consciousness. I have argued for a more radical sort of ontological gap between consciousness and the physical world. Consciousness is not a region of reality to which our access is idiosyncratic but rather it exists only in the accessing itself. There is no region of reality to which subjective phenomena belong; they simply belong to our accessing of regions of reality that are, in themselves, perfectly objective. This stems from the hybrid nature of consciousness: it can be both act and object of experience. Consciousness can be both that upon which awareness is directed (i.e., inner sense is possible) and the directing of awareness (the act of inner sensing is numerically distinct from the states or facts that it reveals to the subject). And what it is like to undergo an experience, I argued, is something that attaches to consciousness as act not object. What it is like to have an experience is not something of which we are aware in the having of that experience but, rather, something in virtue of which we are aware of distinct, and nonphenomenal, objects. This view of consciousness has Kantian roots – consciousness is a condition of possibility of objects being presented to a subject under a mode of
presentation, and in this sense is a transcendental feature of the world. This ‘transcendentalist’ view of consciousness, when pushed, has a striking consequence: consciousness is real but nowhere at all. In this, the position shares McGinn’s emphasis on space as the problematic feature that undermines reductive explanations of consciousness. But, unlike McGinn’s form of mysterianism, it also entails that there can be no explanation of consciousness at all – even if our conceptual repertoire were Godlike.
Summary The mind–body problem is really two problems. The item problem concerns the nature of mental items: are they or are they not physical? The essence problem concerns the nature of the defining essences of mental phenomena – consciousness and intentionality: can they or can they not be explained in physical terms. With regard to the item problem consensus gradually seems to be coalescing on a combination of (1) mental–physical identity at the level of tokens, (2) mental–physical supervenience at the level of types, where this supervenience is underwritten by (3) a functionalist account of the nature of mental properties. There is no such consensus with regard to the essence problem however. The problems of explaining the nature of both consciousness and intentionality are among the hottest topics in contemporary philosophy of mind. While there is much agreement about the nature of the problems, substantive solutions are still very much up for grabs. See also: History of Philosophical Theories of Consciousness; Intentionality and Consciousness; Mental Representation and Consciousness; Self: Body Awareness and Self-Awareness.
Suggested Readings Chalmers D (1996) The Conscious Mind: In Search of a Fundamental Theory. Oxford: Oxford University Press. Crick F and Koch C (1994) The Astonishing Hypothesis. New York: Scribner. Davidson D (1970) Mental events. In: Foster L and Swanson J (eds.) Experience and Theory, pp. 79–101. London: Duckworth. Jackson F (1982) Epiphenomenal qualia. Philosophical Quarterly 32: 127–132.
The Mind–Body Problem Jackson F (1986) What Mary didn’t know. Journal of Philosophy 83: 291–295. Kim J (1984) Concepts of supervenience. Philosophy and Phenomenological Research 65: 153–176. Kripke S (1980) Naming and Necessity. Cambridge, MA: Harvard University Press. Leder D (1985) Troubles with token identity. Philosophical Studies 47: 79–94. McGinn C (1991) The Problem of Consciousness. Oxford: Blackwell. Nagel T (1974) What is it like to be a bat? Philosophical Review 83: 435–450.
55
Rowlands M (2001) The Nature of Consciousness. Cambridge: Cambridge University Press. Russell B (1927) The Analysis of Matter. London: Kegan Paul, Trench, Trubner. Russell B (1948) Human Knowledge; Its Scope and Limits. London: George Allen & Unwin. Strawson G (2003) Real materialism. In: Anthony L and Hornstein N (eds.) Chomsky and His Critics, pp. 49–88. Oxford: Blackwell. Tye M (1995) Ten Problems of Consciousness. Cambridge, MA: MIT Press.
Biographical Sketch
Mark Rowlands is a professor of philosophy at the University of Miami. He is the author of a dozen books, translated into 15 languages. These include The Body in Mind: Understanding Cognitive Processes (Cambridge University Press 1999), The Nature of Consciousness (Cambridge University Press 2001), and Body Language: Representation in Action (MIT Press 2006). His autobiography, The Philosopher and the Wolf, was published by Granta in 2008.
Mind Wandering and Other Lapses J Smallwood, University of California, Santa Barbara, CA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Action slips – A class of error in which the individual fails to perform a well-practiced task because of a lapse in attention to the task in hand. Also known as an attentional lapse. Current concerns – The salient personal concerns that an individual may have at any one moment in time. Generally these are considered issues which extend from medium to long-term time frames (e.g., days, weeks, and months). Default mode network – A term coined to describe a constellation of cortical and subcortical structures which are unusually active when individuals rest in a scanner. One current hypothesis on the default network is that it provides the internal content of mind-wandering episodes. Ironic processing theory – A theory developed by Dan Wegner to explain the regular wanderings of consciousness. According to ironic processing theory the mind wanders because of the manner that control is instantiated in the brain. One system – the intentional monitoring system is concerned with effortful control of attention, a second – the ironic monitor – checks whether the intentional monitor is currently on task. Meta-awareness/consciousness – An intermittent state in which we periodically take stock of the contents of consciousness by reflecting on what we are thinking about and comparing this to what we thought we were doing. Mindless reading – The temporary experience of reading a narrative without making sense of the narrative. Mind wandering – An intermittent state when attention shifts from the task of the moment and is often accompanied by
thoughts related to unrelated personal concerns. The accompanying thoughts are often referred to as task-unrelated thought or stimulus-independent thoughts. Task-unrelated/stimulus-independent thoughts – The experience of thoughts which are generally internally generated and are unconstrained by the task in hand.
Introduction One of the distinguishing features of the human mind is the ability to transcend the here and now using imagination. No phenomenon, however, demonstrates the emphasis that our cognitive system places on imaginative processes better than the experience of mind wandering during reading. The strong intuition that our mind has wandered from the text to some other topic of thought, not only underscores the power of the mind to transcend the here and now, it suggests that our imagination has ability to capture attention and so temporarily derail the local goal of reading. This article provides a basic summary of the current state of research into the experience of mind wandering. It describes the challenges that must be overcome when studying the wandering mind and present a simple information-processing model within which to make sense of research. In order to do this, this article considers the results of cognitive, behavioral, and neuroimaging studies which reflect the current state of knowledge of how and why mind wandering occurs in the waking brain. Among the topics in this volume mind wandering is one of the more enigmatic – it is, for example, probably one of the few topics covered that will be as easy to explain to a layperson as to a cognitive scientist. Such a contradiction arises because, before the recent surge in interest over 57
58
Mind Wandering and Other Lapses
the last 4 or 5 years, empirical work in mind wandering was sporadic – despite the obvious importance of mind wandering to our species, cognitive scientists simply did not study it. To understand why research on mind wandering has not matched the resonance of the phenomenon to the human condition, it is necessary to understand why states of task-unrelated thinking are problematic to study. The Collins Concise English Dictionary defines the verb ‘to wander’ as ‘to move or travel about in, or through (a place) without any definite purpose or destination.’ One component of ‘wandering,’ therefore, is that it implies movement between different places. In order to apply this notion of movement or travel to the mind, it is first necessary to define the places to which the attentional system could travel. On the one hand, the standard influences in attention to which cognitive psychology is familiar are those which are necessary to complete the task. Usually tasks employed in the laboratory emphasize information contained in the external environment and so psychology has a detailed understanding of how attention is deployed on sensory information. When we engage in mind wandering, the focus of attention is withdrawn from the external environment and instead is often directed to information which is derived internally. When we catch ourselves thinking about something other than the topic of what we are reading, we are engaging in what cognitive psychologists refer to as ‘task-unrelated’ or ‘stimulusindependent thought’ – labels which capture the notion that these thoughts are not related to the immediate external environment or what we are doing. One aspect of mind wandering which sets it aside from other domains of psychological inquiry is that the content of the episodes themselves is often the product of imaginative processes. One barrier to studying mind wandering, therefore, is the need to develop tools for capturing covert attentional shifts which only indirectly available to scientific enquiry. To date, the most direct method for studying mind wandering is by asking individuals to self-report the contents of consciousness as they perform a task. A second aspect of ‘wandering’ is that it implies movement or ‘travel’ and so the wandering mind is one that is in transition. When we mind wander
during reading, focal attention shifts from the narrative toward our own thoughts. It is this state of transition or flux which motivated William James to use the term ‘the stream of consciousness.’ Studying the ‘wandering’ mind, therefore, involves studying how attention shows a tendency to ebb and flow between different mental states. A second difficulty in studying mind wandering is that it requires an experimental design which acknowledges the ebb and flow in attention.
A Simple Information-Processing Model of Mind Wandering The first step in understanding mind wandering is developing a simple framework in which to understand the current knowledge of mind wandering. A simple way to conceptualize mind wandering and other mental states is by considering their relation to the flow of information through an attentional system, such as that presented in Figure 1. This model is based on one that I have been developing over a number of years in conjunction with Jonathan Schooler and others. Figure 1 has three different panels and each describes how a particular mental state relates to different configurations of informational flow through the attention. Different Mental States as Different Information Flow The first aspect to note from this model is that information can flow from two basic sources. One source of information to which we can attend arises from the external environment. Over the last 40 years psychology has developed a thorough account of how brain detects, codes, and responds to such sensory information. The second source of information to which we can attend arises from the internal environment. The brain has the ability to store representations of past events, such as memories. When we use our imagination to consider information from the internal milieu, we use these internal representations to free our mind from the constraints imposed by the outside world. Both sources of information – the internal and the external – form the contents of consciousness when they gain access to working memory.
Mind Wandering and Other Lapses
59
External information
External information
Working memory
Panel (A) Attention is primarily focused on incoming sensory information.
External information
Working memory
Internal representations
Internal representations
Panel (B) Attention is coupled to both incoming sensory information and information arising from the internal representations.
External focus to attention
Working memory
Internal representations
Panel (C) Attention is focused on information arising from the internal representations. Internal focus to attention
Figure 1 Schematic account of the flow of information through an attentional system and the associated mental states. Panel (A) presents the attentional coupling which accompanies the processing of information which arises from the external environment. Panel (B) the state of coupled attention associated with processing information from both internal and external information in order to perform a task. Panel (C) presents the state of decoupled processing when the mind wanders.
A general assumption in psychology is that working memory resources are limited and so, different sources of information (such as those derived from internal and external environment) compete for access to a general attentional resource usually equitable with concepts such as awareness. Information that gains access to this general-purpose mental resource becomes the focus of attention. So that when the flow of information into working memory is primarily from the internal environment, attention will be focused on this source of information and less attention will be directed externally. The model is a variant on what is called a global workspace model which has been developed by a number of different theorists, including Bernard Baars and Stanislas Dehaene. Global workspace models have provided influential accounts of many different conscious phenomena covered in this volume. Figure 1 illustrates three states of information flow and the associated mental states. On the
right-hand panel attention is focused exclusively on the external environment. The flow of information is represented by the arrows, and in the right-hand panel they indicate that information is cycling between working memory and the external environment. This mental state would involve attention giving priority to incoming sensory information, such as when we are startled by a loud noise. In the central panel, information is flowing from both internal and external environment into working memory. To reflect that both internal and external information is contributing to focal attention, arrows indicate that information is traveling from both internal and external sources into working memory. When we are engaged in a task with an extended narrative such as reading or watching a film, we use both internal and external sources of information to make sense of what is happening. Finally, in the left-hand panel, information flows from internal representations into working memory and so attention is decoupled
60
Mind Wandering and Other Lapses
from both the task and the environment. In the literature, a focus on internal information processing has been described in a number of different terms – daydreaming, task-unrelated thought, and stimulus-independent thought. All of these labels emphasize a mental state in which attention is giving priority to internal information. Transitions in Mental States as Changes in the Flow of Information As noted in the ‘Introduction’ section, a second key feature of mind wandering is that it involves a state of attentional transition. It is easy to see how the dynamic aspects of mind wandering are captured by the model presented in Figure 1. When we are engaged in normal reading, information flows from both external (e.g., the text) and internal (the representation of the narrative) sources into working memory (Panel B). When the mind wanders from the narrative, the flow of information from the text is suppressed in favor of task-irrelevant information from the internal milieu (Panel C). In terms of Figure 1, therefore, when the mind wanders during reading, the direction of information flow through the attentional system shifts from Panel B to C.
Why the Mind Wanders While the model presented in Figure 1 explains what is entailed in different mental states it does not explain why the mind should wander in the first instance. Despite being a compelling and frequent aspect of our mental life, explaining why the mind wanders is in fact one of the harder aspects of the phenomenon to accommodate into models of cognitive science. In order to see why the relation between mind wandering and intentionality is controversial requires that we examine the position that a wandering mind occupies in a standard hierarchical model of attention. In such a model of attention, basic information processes which are more or less directly related to the processing of stimulus information are seen as occupying a relatively low-level position in the hierarchy. An example of a low-level attentional process is the detection
of luminance or color information by the visual system. Low-level processes are generally described as influencing the control of attention through a bottom-up process. Low-level cognitive processes are often contrasted with higher-level cognitive processes which operate on more abstract information. Such abstract information is usually the product of computations performed on low-level sensory information, or alternatively is retrieved from memory systems. The influence of higher-order cognition is described as top-down. Examples of higher-order cognition would involve planning (anticipating future events) or metacognition (reflecting upon particular thoughts). In standard models of attention, higher-level cognitive structures provide the opportunity for the top-down control of behavior. Top-down control allow individuals to strategically modulate the manner in which they respond in the current environment in order to achieve a goal. When mind wandering is conceived of in terms of the top-down/bottom-up dichotomy it is clear that the phenomenon is somewhat paradoxical. On the one hand, the complex information-processing requirements to introspect in a goal-motivated manner suggests a top-down process, however, the strong sense that we do not intend our minds to wander suggests an absence of top-down control. The puzzling nature of control during mind wandering is satirized by comedian Steven Wright when he jokes, ‘‘I was trying to daydream but my mind kept wandering.’’ Ironic Processes and the Wandering Mind One explanation for why the mind wanders is suggested by the theory of ironic processes developed by Dan Wegner. Wegner argues that the experience of mind wandering occurs because of the manner with which control is instantiated by the brain. According to Wegner, mental control is instantiated by the cooperation of two systems which combine to produce the control we have over our mental life. The first system – termed the ‘intentional monitor’ – operates in much the same way that top-down control was described in the previous section. The intentional operator allows us to strategically constrain our attention to a particular topic or activity, and is associated
Mind Wandering and Other Lapses
with the palpable sense of effort when we focus our own thoughts. Wegner’s insight was that a system such as the intentional operator could become preoccupied when engaged in a demanding activity. In order to keep the intentional system on track, Wegner suggested that we have a second monitoring system – termed the ‘ironic monitor’ – which is responsible for ensuring that the intentional monitor does not get side tracked. According to Wegner, it is the interactions of these two systems that lead the mind to wander. One implication of the theory of ironic processes is that if the intentional monitor was engaged on a demanding task, the ironic monitor would make a greater contribution to the contents of consciousness. To do so, Wegner and others developed what is known as the thought suppression or ‘white bear’ paradigm. In these studies, participants are asked to suppress thoughts of a particular topic that is not present in the immediate environment – often a ‘white bear’ – and are either asked to perform a demanding secondary task or not. Participants are asked to indicate (often by ringing a bell) whenever they experience a target thought. Results of this paradigm are quite striking – generally when participants are asked to suppress the target thoughts and are placed under load they report the most intrusions of the unwanted thought. This ironic increase in unwanted thoughts is consistent with Wegners’ model because it depends upon (1) the ironic monitor being directed to the unwanted thought by instruction to suppress the thought and (2) the intentional monitor being occupied by the demanding secondary task. The thought suppression paradigm, and related studies provide important support for the notion that the experience of mind wandering is in fact compelled by the manner in which cognitive control in instantiated by the mind. Awareness and the Recognition of a Wandering Mind A second possible explanation for why mind wandering occurs has been proposed by Jonathan Schooler. He argues that in addition to the normal division of conscious processes into explicit and implicit, we are also intermittently able to represent the contents of awareness – a concept he has
61
referred to as meta-awareness or metaconsciousness. Mind wandering during reading illustrates a situation when we are aware of the content of awareness (the off topic thoughts) and yet are unaware that we are currently engaging in thoughts which potentially undermine our ability to comprehend what we are reading. According to Schooler, therefore, mind wandering occurs because of the difficulties we have in continually representing the contents of consciousness. On the flip side, catching ourselves mind wandering occurs because we suddenly take stock, or rerepresent the contents of awareness, and realize that for some time we were engaging in thoughts which were unrelated to the task of reading. One clear advantage of the notion of meta-awareness is that it captures the compelling sense that in many mind-wandering episodes we fail to recognize for several moments that we were off task. According to Schooler, therefore, the ebb and flow of our minds’ wandering occurs in part because we are unable – or unwilling – to continually represent the content of our own experiences.
Three Routes to a Wandering Mind So far this article has considered a simple information-processing account of information flow which allows mind wandering and other mental states to be clearly conceptualized and has described two of the most compelling explanations for why the mind should wander in the first place. In the next section, this article will consider studies which have examined the influences that lead the mind to wander. Figure 2 presents a schematic diagram of different attentional configurations which could be conceivably be associated with greater likelihood of engaging in mind wandering. Each of these different configurations can be broadly considered as a different route that leads the mind to wander. The first route to a wandering mind is when the external environment does not exert a sufficient influence on working memory to necessitate that task-irrelevant representations are suppressed. The lack of a demanding external environment is represented in Figure 2 by the relative absence of downward arrows in the left-hand panel. The second
62
Mind Wandering and Other Lapses External information
Working memory
Internal representations
Route 1: Insufficient need to constrain attention to the current task.
External information
External information
Working memory
Working memory
Internal representations
Route 2: Increased accessibility in a set of internal mental representations.
Internal representations
Route 3: Insufficient working memory capacity to successfully constrain attention to the task.
Figure 2 The three routes to a wandering mind. Route one – the external task does not require continuous monitoring and idle working memory resources are available to be co-opted into mind wandering. Route two – an increase in the accessibility of a salient internal mental representation attracts resources away from the outside world. Route three – failures in attentional control allows distractions from the internal or external environment access to working memory.
route to the wandering mind occurs when our attention is drawn to a particularly salient thought. The increase in salience is represented by the single black arrow in the central panel. The third and final route to a wandering mind is when the individual lacks the necessary attentional constraint in order to maintain attention on the task. The inability to constrain attentional resources is represented by the shading in the rectangle in the right-hand panel. In the next three sections, this article will consider experimental evidence that each route can lead to a greater incidence of a wandering mind.
Route One – The Influence of the Task Environment A tourist, particularly one who is British, can find driving around a foreign city one of the more stressful aspects of their holiday – not only must
they attend to the hectic alien traffic system, such as driving on the right-hand side of the road, they must also determine the correct route to take them to their destination. By contrast, the commuter who makes the journey home from work has a different experience. Unlike the tourist, a commuter is familiar with the traffic systems, and because they have a detailed knowledge of the spatial layout of their hometown they need give few thoughts to the task of navigation. As a consequence the commuter has the time and resources to explicitly consider information unrelated to the current circumstance or may instead simply let their minds wander. The tourist does not have such a luxury – their attention is likely to be entirely wrapped up in the task of getting to their destination. This anecdote illustrates one of the well-documented routes to a wandering mind – our minds wander most frequently when the current task does not require that we constrain our attention to task-relevant material. One way to investigate the effect of task environment is to examine the effect that the rate of events occurring in the task has on the amount of mind wandering. Work conducted in the 1960s by John Antrobus and Jerome Singer demonstrated a clear linear relationship between the frequency of mind wandering and the rate of signal presentation during sustained attention. In these studies, participants were asked to perform a signal-detection task and were periodically asked to report the amount of mind wandering they were engaging in. In these studies, the rate of presentation was varied, and the results showed that slower presentation rates tended to lead to greater mind wandering than did fast rates. The suppression of mind wandering by frequent events is perhaps why when driving in a city we seem to maintain better attention on what we are doing than when driving on an empty highway were incidences of mind wandering seem much more frequent. In the anecdote described at the start of this section, it is the need to continuously monitor the task that distinguishes the commuter from the tourist. As both commuter and tourist are driving the same route they are engaged in the same task – it is the locals’ familiarity with both the traffic system and their knowledge of the route which allows them the opportunity to let their
Mind Wandering and Other Lapses
minds wander. Research by John Teasdale elegantly demonstrated the fact that tasks requiring continuous monitoring tend to suppress the mind’s tendency to wander. In a series of studies, participants were shown a series of digits and were asked to either say the number out loud immediately, or to hold it in mind for a short period (c. 2 s). In both task participants were also asked to report the amount of mind wandering they were engaged in. Even though these two tasks are trivially easy, the simple fact that individuals had to hold the digits in working memory for a short period significantly reduced the amount of mind wandering experienced. The simple involvement of working memory, however, is not the whole story with respect to mind wandering. A quick review of the last time we were absorbed in a good book or a film indicates that often our mind wanders ‘least’ when the primary task interests or intrigues us. Instead, in the context of tasks such as reading, our experience is more likely to be driven by features of the task which interest us rather than those which are simply demanding to process. Leonard Giambra and others compared the influence of interest and difficulty in reading. In this study participants read a number of texts – preselected for either interest or difficulty during which mind wandering was measured. Their results demonstrated that mind wandering during reading was predicted by interest in text but not difficulty. Presumably, participants find it more absorbing to read an interesting story and as a result they are better able to maintain continuous focus on what they are doing. This finding concurs with our own everyday experience that getting lost in a good novel or film is possibly the best way to temporarily reduce the mind proclivity to wander. Route Two – The Influence of Current Concerns The second route to a wandering mind is when our thoughts get attracted to a particularly salient topic. A review of our mind-wandering experiences suggest that we spend a great proportion of time thinking about things which are generally personally relevant to us as an individual. Eric Klinger suggested the term ‘current concerns’
63
to reflect that much mind wandering generally has this personally relevant focus. Several studies have explored how the salience of particular concerns are related to the frequency of mind wandering. In the 1960s during the Vietnam War, Antrobus and others recruited a cohort of individuals to perform a laboratory task in which mind wandering was measured. While the participants were waiting to perform the task, they either heard a fake news broadcast that China had entered the Vietnam War or a control broadcast. During the subsequent attentional task, the erroneous information that China had entered the Vietnam War led to an increase in the frequency of mind wandering and an increase in signal-detection errors. At debriefing, participants reported experiencing reasonably distressing thoughts associated with their concern about the future. This study clearly indicates that the second route to the wandering mind is when a particularly salient concern overshadows performance on the task in hand. An alternate way to investigate whether mind wandering is associated with particularly salient cognition is to compare populations which differ on the salience of their current concerns. One population that tends to have particularly salient concerns are depressed individuals. Since the 1960s it has been well known that a significant contributing factor to depressed mood are current levels of stress such as daily hassles. Recently, in conjunction with a number of my collaborators I explored whether mind wandering really is associated with particularly pressing concerns. To do so we recruited a cohort of participants and in addition to completing a measure of depression asked them to perform a word-learning task in the laboratory during which we measured mind wandering. While they performed the cognitive tasks we also measured autonomic function (heart rate and the skin conductance level). Results indicated a number of interesting things. First, the ability to remember the words presented was significantly impaired when individuals were off task. Second, the size of the skin conductance response was significantly higher during periods when mind wandering occurred. Finally, participants who were particularly unhappy experienced greater increases in physiological arousal (particularly their heart rate) when their minds wandered.
64
Mind Wandering and Other Lapses
These data are consistent with the notion that mind-wandering episodes are focused on salient and so arousing information because of the associated physiological responses and so supports the importance of an individuals ‘current concerns’ in leading the mind to wander. Route Three – A Lack of Constraint When we notice that we are mind wandering, as we all frequently do, this act is no mean feat. Among other things catching our minds wandering requires that, first, we can reflect upon the contents of consciousness and, second, assess whether we are still on course to achieve our current goal. Moreover once a mind-wandering episode is identified, the correctly motivated individual will set aside that particular thought and return their attention to what they should have been doing. The ability to catch one’s own mind wandering and correct this attentional problem requires sophisticated metacognitive or attentional control. Unlike the previous two routes, the third route to a wandering mind is through difficulties in metacognitive or attentional control which could lead to a general increase in distractibility and so an increase in mind wandering. Recent studies have confirmed that the ability to control attention impacts upon the ability to control mind wandering. Michael Kane and others explored the relationship between the ability to control attention – often referred to as working memory capacity – and the tendency to engage in mind wandering in every day life. To do so Kane and other recruited a large number of students who performed a number of laboratory tasks measuring their ability to maintain and control information in working memory. After this the participants were given pagers which they carried for a number of days which periodically beeped at which the participants were asked to provide detailed accounts of the content of their thoughts and the context within which it occurred. The results documented an interesting relationship between mind wandering and working memory capacity. Under challenging circumstances individuals who had a higher working memory capacity tended to mind wander less, than at other times. Individuals with a lower working memory
capacity, however, tended to mind wander regardless of whether the task was easy or difficult, suggesting that the poorer attentional control prevented these individuals from suppressing mind wandering if and when the situation demanded they do so. Thus, the ability to control the contents of working memory plays an important role in controlling the wandering mind.
The Consequences of the Absent Mind The previous section of this article considered three different routes that lead the mind to wander. The mind’s tendency to wander, however, can lead to disastrous consequences for the integrity of the task in hand. This research covers a wide range of tasks – from simple signal-detection tasks through to more complex tasks such as reading – and so provides an illustration of the far reaching consequences that mind wandering and the associated lapses can have on our daily lives.
Mindless Reading Throughout this article we have used the example of mind wandering during reading as a simple example of the phenomenon. Reading is a complex task in which the reader must continuously integrate from what they are reading into a larger model of the narrative. Common experience informs us that it is a relatively common experience to engage in mind wandering during reading, and when it occurs we often have to reread whole sections of the text as we have no clear idea of what happened in the narrative. This leads to the clear contention that mind wandering during reading could have implications for what is being read. To explore exactly what the consequences of mind wandering during reading were, we examined the occurrence of off task thoughts while participants read a detective novel. We chose a detective novel because it was an interesting story but one in which the reader must develop a detailed model of the narrative in order to solve the crime. We reasoned that if mind wandering prevents participants from making sense of what they were
Mind Wandering and Other Lapses
reading it would ultimately prevent them from building a model of the story and so would prevent them from solving the crime. To examine this question, we measured mind wandering at both random intervals in the story and at points when the author delivered a clue which if correctly attended would help the reader solve the crime. At the end of the task participants were asked to answer several questions – one of which required that they had solved a critical aspect of the mystery. Consistent with expectations, analysis indicated that the ability of the individual to solve the crime depended on whether they were on task at the critical parts of the novel. Interestingly, only reports of mind wandering without awareness prevented the reader from solving the crime. By contrast, general mind-wandering propensity (e.g., being off task at random intervals in the text) did not determine whether the individuals solved the crime. The absence of a relation between overall mind-wandering propensity and the participants’ ability to solve the crime ruled out alternative explanations such as attentional control as being the main contributor to the study’s results. This study clearly indicates that the effects of mind wandering during reading is to interfere with the ability of the individual to create a model of what is being read and that one implication of the failure is that individuals are unable to make the inferences necessary for reading. Given that the majority of education proceeds through reading, one implication of this finding is that mind wandering could well be an under-recognized influence on the educational achievement of both children and adults alike. Action Slips The consequences of mind wandering are not limited to tasks that are cognitively demanding it can also lead to error in tasks which at first glance seem trivially easy. Often task in such safety critical situations are largely automated and rely on humans as a watch keeper rather than as the direct controller of the action. While this automation has largely removed direct human error, an indirect consequence of the increased automation is that it provides the ideal environment to encourage the wandering mind. In some contexts, the simple and
65
routine nature of the task makes mind wandering one of the few ways that humans can still contribute to error. Train and automobile driving, provide clear examples of long monotonous tasks in which even momentary inattention to the task in hand can have calamitous consequences. Errors which occur in seemingly simple tasks are called action slips. Action slips occur whenever a participant makes an error in a routine action sequence that under normal circumstances can be easily performed. Action slips, or, skill-based slips were first studied in the 1980s by the psychologists Jim Reason and Donald Broadbent. From the perspective of an external observer action slips arise from the inattention of the individual rather than because of the difficulty associated with the task. As such action slips arise in the same mysterious manner as does mind wandering – the event is determined by an internal change and not an external event. Initially, this research employed diary studies to assess the frequency that lapses occurred with reasonable frequency in everyday life, although they were hampered by their inability to study the phenomenon experimentally. In the late 1990s, the study of action slips was moved into the cognitive laboratory by the work of Ian Robertson, Tom Manly, and others. In order to mimic the relatively undemanding nature of many everyday tasks, these researchers examined the performance of individuals on a simple routine task such as withholding a response to a familiar stimulus (often the digit ‘3’) presented in a long sequence of nontargets. This task is deceptive – the vigilance component of the task is minimal as the stimulus is easy to detect. The requirement to withhold a response to an infrequent target, however, requires a surprisingly demanding form of attentional control. Because the target stimulus occurs very infrequently, it is tempting for the individual to perform the task in a rapid stimulus driven manner, or in more familiar terms on ‘automatic pilot.’ In these circumstances when a target arrives the individual has only a short interval to intervene before the automatic tendency to respond to the target takes over and an error ensues. Even though the detection of a target is trivially easy in the SART, one needs to attend to one’s performance in a more or less continuous manner in order to perform the task effectively. This
66
Mind Wandering and Other Lapses
emphasis on one’s own behavior has led to the task being described as the Sustained Attention to Response Task and is often shortened to the acronym the SART. The SART mimics the requirement that occurs in many safety conscious industries because it confronts the individual with a dull task environment and requires that they maintain almost exclusive attention on their performance. It is clear that the concept of action slips is closely linked to the more subjective aspects of mind wandering – both represent situations in which attention is not fully constrained on what is being done. While there is an obvious similarity between the notion of action slips and mind wandering, just how closely related are these different states? This is the question that Dan Smilek and others set out to answer. They recruited a sample of individuals who completed separate measures of action slips and mind wandering in everyday life and also completed the SART. The results illustrated an interesting dissociation – both measures of action slips and mind wandering in everyday life made unique contributions to performance on the SART. Measures of mind wandering were associated with how fast participants responded to nontargets, while the likelihood of action slips was strongly associated with the amount of errors made on the task. This research clearly documents that mind wandering and action slips are closely related yet experimentally dissociable aspects of conscious experience and leaves open the intriguing question about the aspects of our mental life that lead mind-wandering episodes to become an error.
Imaging the Wandering Mind As with many areas of psychology, the detailed measurements provided by neuroimaging has helped reveal many fascinating insights into how mind wandering occurs in the waking brain. In fact a key problem associated with mind-wandering research is that the processes which occur are private and so invisible to direct observation and in such circumstances the ability of neuroimaging to reveal the processes which take place in the
waking brain is a very useful skill. In this article, we focus on three neuroimaging studies that reveal three key aspects of mind wandering: (1) the extent of cortical processing on the task it is performing while the mind wanders, (2) the neural substrates that produce the introspective content during mind wandering, and (3) the processes which are used to keep the wandering mind in check. The Absent Task One advantage of neuroimaging is that it makes it possible to ascertain the depth of processing that an individual is engaged by measuring the response of the brain to events in the task. These evoked responses are referred to as event-related potentials (ERP) and are calculated by measuring the electrical signals generated by the brain using electrodes placed on the scalp. These signals are averaged across different conditions and can be analyzed to estimate the amount of processing that is being deployed to a given stimulus under different task conditions. In a simple account of mind wandering, we assume that when the mind wanders participants are less aware of the task at hand than when they are focused on what they are doing. In collaboration with Todd Handy and others, we tested whether periods of mind wandering really were associated with less detailed processing of the task than periods of task focus. To do so we recruited a number of volunteers who performed the SART and measured how their brains responded to events in the task when they were mind wandering. We were interested in a component of the event-related potential, known as the P3 which occurs c. 300 ms after stimulus onset and is conceptualized as indicating the amount of effort that is being deployed on the task. We reasoned that if participants really were not giving the task their full attention during mind wandering they should show a smaller P3 in periods when they were off task. The results documented that both reports of mind-wandering and errors on the SART were associated with a reduction in the amplitude of the p3 component of the ERP. Interestingly, further analysis indicated that reports of mind wandering without awareness tended to
Mind Wandering and Other Lapses
be most closely linked to errors, suggesting that failing to recognize that our mind has wandered could help explain why lapses occur.
67
default network structures in the practiced blocks of the task, and their association with daydreaming tendency support that assumption that the default network plays a role in mind wandering.
The Default Network and the Wandering Mind
Constraining the Wandering Mind
One of the most controversial findings in neuroimaging over the last several decades is that when participants rest in a scanner, a large number of cortical and subcortical regions are more active than during many demanding experimental tasks. One influential view of this result – described as the default mode hypothesis by Marcus Raichle and others – is that the activity in these structures in periods when participants were ostensibly resting in the scanner reflects a psychological baseline that participants adopt when their task does not require their undivided attention. One crucial feature of the default mode hypothesis is that it predicts that many of the cortical and subcortical structures may play a role in delivering the introspective content during states of mind wandering. Recent research conducted by Malia Mason has provided important support for the assumption that the default network does play a role in the introspective content when the mind wanders. In a longitudinal study, Mason and others trained participants to perform the same version of a working memory task in three half hour sessions over consecutive days. On the fourth session, participants performed either the same well-practiced version, or a matched novel version. During this session the authors employed thought sampling to estimate how much mind wandering was occurring – participants reported more off task thoughts in practiced than novel blocks. On a fifth session, participants performed both novel and practiced sessions in a functional magnetic resonance imaging (fMRI) scanner. Finally, participants completed a measure of their tendency to daydream. The results indicated that during scanning, a number of structures in the default network were more active in the practiced than nonpracticed blocks. Moreover, the size of activity in several default structures was positively associated with the individual propensity for daydreaming measured via questionnaire. Taken together, the relative activation of
Given the studies described in this article so far, it seems somewhat miraculous that we are able to engage in any task without continually losing track of what we are doing, and yet despite our minds tendency to wander, as a rule humans are able to successfully navigate their way through many complex tasks environments. It is clear, therefore, that evolution has provided humans with the apparatus to control our minds tendency to wander – but how do we do this? One possibility is that the mind employs the same neural structures to constrain mind wandering as it does to resolve conflicts which arise when we perform other tasks. Models of cognitive control suggest that two neural structures are involved in conflict management – aspects of the lateral prefrontal cortex which are thought to be involved in the strategic regulation of effort to a given task in a sustained manner and the anterior cingulate which is thought to be involved in identifying and responding to transient changes in the conflict when the need arises. To explore whether the same conflict management processes are involved in controlling mind wandering, Jason Mitchell and colleagues asked participants to perform a thought suppression task in an fMRI scanner. Participants were asked to either suppress or report the experience of thoughts regarding a white bear. The results indicated that relative to the report condition, periods of thought suppression did indeed mobilize lateral prefrontal structures indicating that the brain employed the same control structures to suppress the experience of unwanted thoughts. Similarly, following the experience of unwanted thoughts the anterior cingulate was more active indicating that this structure was implicated in the response to failures in mental control. Taken together, this result clearly indicates that one reason that we are able to refrain from mind wandering in general is because we are able to employ existing executive control strategies to regulate our mental lives.
68
Mind Wandering and Other Lapses
Future Directions in Mind-Wandering Research It is an exciting time to be a researcher studying mind wandering. Recent theoretical and experimental developments have shed important light on why the mind wanders and the consequences that can occur when it does. Meanwhile, advance in neuroscience are beginning to reveal in an unprecedented fashion the neural structures that support and control the introspective world that plays such a pivotal role in our mental life. In the future, we are likely to gain important insights into this remarkable feature of the human mind. One area of future research is likely to be the potential functional basis that mind wandering could play in our daily lives. One important human skill is the ability to solve tasks which require an indirect or counter intuitive solution. Often these sorts of cognitive processes are referred to as insight problems. On the face of it, mind wandering shares a number of basic similarities with insight problem solving. Both insight problem solution and prospective memory share with mind wandering the requirement that the mind is not to tightly constrained to the current task. The relationship between mind wandering and creativity has been relatively overlooked in much of the work on mind wandering, as most of the work has focused on the negative aspects of this concept – as typified by the title of this article. Alternatively, mind wandering could share many features with what researchers refer to as prospective memory – our mind’s ability to remember future goals. Studying whether mind wandering is of functional significance to the individual could be important in the future in detailing the function of an experience which makes up such a large component of our waking life. Another important question facing mind-wandering researchers is the development and validation of an indirect marker a so called mind wandering meter. While this question was almost impossible to imagine several decades ago, in practical terms technological advances in neuroimaging techniques have made it a viable possibility. Such a discovery would revolutionize the study of mind wandering because it would make it possible to identify the experience of mind wandering without interrupting the experience. If successful, this would provide the opportunity for
investigators to begin to identify in an online manner the occurence of the wandering mind. Because this measure would be entirely independent of the selfreports of individuals themselves, the extent to which theory predicted fluctuations in this index of mind wandering would provide the strongest source of data to date to approach question such as why the mind wanders. The lasting message of research on mind wandering, however, is that it reminds us of the essential flux which is such a fundamental property of our mental lives. It is this ebb and flow of attention, so characteristic of processes such as mind wandering, which motivated William James to use the term ‘the stream of consciousness’ as a metaphor. Rather than assuming that the lack of control over the phenomenon is a limitation, recent success in understanding the theoretical and neural aspects of mind wandering reminds us of the importance that the detailed empirical observation of naturally occurring phenomenon plays in our more general understanding of the mind. It is only by adopting a less constrictive experimental approach than cognitive experimental psychology have favored in the past, that we can shed light on the ebb and flow of attention that leads to states like mind wandering. See also: Attention: Selective Attention and Consciousness; Automaticity and Consciousness; Neglect and Balint’s Syndrome; Philosophical Accounts of Self-Awareness and Introspection.
Suggested Readings Reviews Antrobus JS (1999) Toward a neurocognitive processing model of imaginal thought. In: Salovey P and Singer J (eds.) At Play in the Fields of Consciousness: Essay in Honor of Jerome L. Singer, pp. 1–28. Hillsdale, NJ: Erlbaum. Klinger EC (1999) Thought flow: Properties and mechanisms underlying shifts in content. In: Singer JA and Salovey P (eds.) At Play in the Fields of Consciousness: Essays in the Honour of Jerome L Singer, pp. 29–50. Mahwah, NJ: Lawrence Erlbaum. Smallwood J and Schooler JW (2006) The restless mind. Psychological Bulletin 132(6): 946–958. Smallwood J, Fishman DF, and Schooler JW (2007) Counting the cost of the absent mind. Psychonomic Bulletin & Review 14: 230–236. Robertson IH and Garavan H (2004) Vigilant attention. In: Gazzaniga MS (ed.) The Cognitive Neurosciences, 3rd edn., pp. 563–578. MIT Press.
Mind Wandering and Other Lapses Wegner DM (1997) Why the mind wanders. In: Cohen JD and Schooler JW (eds.) Scientific Approaches to Consciousness, pp. 295–315. Mahwah, NJ: Erlbaum. Empirical Papers Cheyene JA, Carriere JSA, and Smilek D (2006) Absentmindedness: Lapses of conscious awareness and everyday cognitive failures. Consciousness and Cognition 3: 578–592. Kane MJ, Brown LH, McVay JC, et al. (2007) For whom the mind wanders, and when: An experience-sampling study of working memory and executive control in daily life. Psychological Science 18(7): 614–621. Mason MF, Norton MI, Van JD, et al. (2007) Wandering minds: The default network and stimulus independent thought. Science 315: 393–395. Mitchell JP, Heatherton TF, Kelley WM, et al. (2007) Separating the transient and sustained aspects of
69
cognitive control during thought suppression. Psychological Science 18(4): 292–297. Raichle, et al. (2001) A default mode for brain function. Proceeding of the National Academy of Sciences 98(2): 676–682. Smallwood J, O’Connor RC, Sudberry MV, et al. (2007) Mind wandering & dysphoria. Cognition & Emotion 21(4): 816–842. Smallwood J, Beech E, Schooler JW, et al. (2008) Going AWOL in the brain: Mind wandering reduces the cortical analysis of the task environment. Journal of Cognitive Neuroscience 20: 458–469. Teasdale JD, Proctor L, Lloyd CA, Baddeley AD, et al. (1993) Working memory and stimulus-independent-thought: Effects of memory load and presentation rate. European Journal of Psychology 5: 417–433.
Biographical Sketch
Jonathan Smallwood has had a long-term research interest in the processes which contribute to the experience of thoughts which are not derived from the immediate environment. He began studying this mind wandering in 1996 when he studied for his PhD at the University of Strathclyde in Glasgow, Scotland. After completing his thesis he worked as a research associate in several universities in the Glasgow area, before moving to the University of British Columbia, Vancouver, Canada to complete a postdoctorial with Jonathan Schooler. Currently he lives in the USA and works as a research scientist at the University of Santa Barbara, California. To date he has published more than 30 scholarly articles, 20 of which are on the topic of mind wandering.
Neglect and Balint’s Syndrome P D L Howe, Brigham and Women’s Hospital, Harvard Medical School, Cambridge, MA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Anosognosia – A neurological disorder, usually caused by brain injury, characterized by the sufferers denying they have a specific impairment, such as hemiplegia. Balint’s syndrome – A neurological disorder characterized by simultanagnosia and optic ataxia, typically caused by bilateral damage to the posterior parietal cortex. Extinction – A neurological disorder where the ability to attend to stimuli in a particular region of the visual field is impaired by the presence of stimuli outside that region. Optic ataxia – The inability to guide the hand toward an object using visual information that cannot be explained by motor or sensory deficits. Simultanagnosia – The inability to perceive more than one object at a time. Somatoparaphrenia – The belief that part of one’s body belongs to another person.
Introduction Neglect, also known as unilateral neglect, hemispatial neglect, hemineglect, neglect syndrome, or spatial neglect, is a disabling brain disorder, usually caused by a brain lesion. Patients with neglect typically have difficulty paying attention to contralesional space. Thus, if the lesion causing the neglect occurs in the right cerebral cortex, the patient may find it difficult to attend to an object situated to the left of their current point of fixation. Neglect is most often caused by damage to the right cerebral cortex, but can be caused by damage to the left cerebral cortex, though the resultant neglect is usually less severe. Because its symptoms can vary from individual to individual and because it can be caused by a lesion in any
of several different locations in the brain, it may not be a single disorder but rather a family of several related disorders. It is especially likely to be caused by lesions near the junction of the temporal lobe and the inferior parietal lobule, but can also be caused by lesions in the frontal cortex, in the white matter, or in subcortical areas such as the thalamus and basal ganglia. Neglect is necessarily oriented relative to a specific frame of reference. For example, to say that a patient has left-sided neglect one must first specify what frame of reference is used to define the term ‘left.’ Neglect oriented relative to the patient is referred to as egocentric. Egocentric neglect is most often orientated relative to the patient’s point of fixation. For example, a patient with left-sided egocentric neglect will typically ignore objects to the left of their point of fixation. However, egocentric neglect may also be orientated relative to a part of the patient’s body. For example, a patient may ignore objects to the left of their left hand. Even when oriented primarily relative to the patient’s point of fixation, the degree of egocentric neglect may be influenced by the orientation of parts of the patient’s body, such as the head and trunk. For example, an object that is to the left relative to the point of fixation, but to the right relative to the head and to the right relative to the trunk, may be less neglected than an object that is to the left in all three coordinate systems. Although left-sided and right-sided neglect are the most common, other forms of egocentric neglect are possible. For example, a patient may have upper visual hemifield neglect, and so ignore objects that are situated above their point of fixation. Neglect is not always oriented relative to the patient and may instead be oriented relative to the object that is being viewed, in which case it is referred to as allocentric. Unlike a patient with left-side egocentric neglect, a patient with leftsided allocentric neglect may be able to attend to
71
72
Neglect and Balint’s Syndrome
objects on their left, but will ignore the left half of each object, regardless of where the object is located. Curiously, if the object has a well-defined left side, a patient with left-sided allocentric neglect may continue to ignore the left side of the object even when the object is rotated by 180 , so that its left side is now situated on the right. For example, when viewing a face with a small blemish on the left side, a patient with leftsided allocentric neglect may continue to ignore the side of the face with the blemish even when the face is rotated by 180 . It seems that when presented with a face, the patient mentally rotates the face until it is orientated in the standard fashion (i.e., the eyes above the mouth) and then ignores the left side. Although neglect can be so extreme that the patient fails to notice large objects on the neglected side, it need not be total. Instead, it may manifest itself only as a tendency not to respond to stimuli in the neglected region. If neglect is not total, the addition of stimuli to the nonneglected side may further decrease the ability to attend to stimuli on the neglected side, a phenomenon known as extinction. As extinction and neglect can occur independently of each other, they may be distinct disorders. For example, a patient may exhibit left-sided extinction without exhibiting left-sided neglect. Such a patient would have no difficulties attending to an isolated object presented on their left. However, introducing objects on their right, might make it hard (or even impossible) for the patient to continue to attend to the object on the left. Conversely, a patient might exhibit neglect without exhibiting any extinction, so the patient’s ability to attend to an object on the neglected side would not be influenced by whether or not objects are presented on the nonneglected side.
Forms of Neglect Neglect most commonly occurs in the visual domain, in which case it may result in some or all of the following symptoms: Patients may shave or apply makeup to only one side of their face. They may eat from only one side of a plate. When moving in a wheelchair, they may bump into objects on the
neglected side. If asked to bisect a line that crosses the visual midline, they may be biased in the nonneglected direction. When copying a picture, they may have a tendency to copy only the nonneglected side. When presented with an image, they may look mainly (or exclusively) at the nonneglected side. Rapid eye movements (REM) to the nonneglected side may occur in REM sleep. If asked to circle all occurrences of a specific letter, they may concentrate on the occurrences that appear on the nonneglected side, circling them repeatedly to the near, or even total, exclusion of those that appear on the neglected side. Neglect can occur in other sensory domains such as the auditory, olfactory, or somatosensory domains. If it occurs in the somatosensory domain the patient will ignore a region of her body and may even deny ownership of neglected limbs, sometimes believing that they belong to someone else (somatoparaphrenia). Often the neglected area is also paralyzed. In such cases, the patient may be unaware of or deny the paralysis (anosognosia). Neglect can also affect motor responses. Although the patient may have no physical impairments, the patient might have difficulty initiating movement or their movement may be slow. Although comparatively rare, a patient may exhibit representational neglect and ignore a portion of an imagined scene. For example, when patients with left-sided representational neglect were asked by Bisiach and Luzzatti to imagine viewing the Piazza del Duomo in Milan standing next to the cathedral in the piazza, the patients often failed to mention streets or places on the left side of the piazza. However, when asked to imagine looking directly at the cathedral, so that they imagined viewing the same scene as before, but from the opposite direction, the patients then recalled the objects and places they had previously failed to mention, since these objects were then situated on the right side of the piazza, relative to their new viewpoint.
Processing of Neglected Stimuli Neglect is not caused by a disruption of the visual system per se, but by a disruption of the cortical system that deploys attention. The only reason
Neglect and Balint’s Syndrome
why a patient with neglect fails to consciously perceive stimuli located in their neglected region is because the patient cannot attend to them. If patients are prevented from seeing because of an abnormality in their visual system such as in their eyes, optic nerves, lateral geniculate nuclei, or visual cortex, they are said to exhibit blindness, not neglect. To diagnose a patient as having neglect, one must first rule out any other reason why the patient might not be able to see. Some stimuli can be perceived even when they are not attended, so can be readily seen by a patient with neglect. For example, if a bright spot of light is presented in complete isolation on a uniform black background, most neglect patients will be able to detect it, regardless of where it is located, especially if it is flashed repeatedly. There is evidence that quite sophisticated processing can occur in the neglected region. For example, if a word is presented on the neglected side, even when it is not consciously seen, it may cause the patient to respond more quickly to similar words presented on the nonneglected side. Similarly, if a patient is simultaneously presented pictures of two different houses and asked which would be better to live in, the patient might reliably choose the house that is not on fire, even though the flames appear only in the neglected hemifield, so are not consciously perceived. When asked to explain their choice, the patient will not be able to do so and will often confabulate. Other studies have asked patients to compare two simultaneously presented pictures and report whether they are the same or different. Patients could do this task even when one or both of the pictures were presented to the neglected hemifield.
Balint’s Syndrome Balint’s syndrome is a brain disorder, closely related to neglect, first reported by Reszo Balint in 1909. Whereas neglect is caused by unilateral damage, Balint’s syndrome is caused by bilateral damage, typically to the posterior parietal cortex. Unlike a patient with neglect, who is unable to attend to objects in a particular region of the visual field, a patient with Balint’s syndrome will be able to attend to an isolated object, regardless of where
73
it is located. However, the patient will find it difficult to point to the object (optic ataxia) or to perceive more than one object at a time (simultanagnosia). While such patients can often perceive the features in a scene, they have difficulty determining which features belong to which object, an issue known as the binding problem. As a result, they have a tendency to conjoin features that belong to different objects and so perceive illusory conjunctions. For example, if a scene contains only red vertical bars and blue horizontal bars, a patient with Balint’s syndrome might perceive a blue vertical bar.
Conclusion Studies on neglect indicate that there is interhemispherical competition between the cortical circuits that control the deployment of attention. For example, damage to the right cerebral cortex may allow the attentional circuits in the left cerebral cortex to dominate. As the left cerebral cortex is responsible for processing the right visual hemifield, attention is directed more often (or even exclusively) to the right visual hemifield, causing the patient to ignore objects that occur to the left of the point of fixation. Similarly, damage to the left cerebral cortex may cause the patient to ignore objects that occur to the right of the point of fixation. If both cerebral hemispheres are damaged, then neither can dominate, so a patient with Balint’s syndrome can perceive an isolated object regardless of where it is located. However, due to the damage to their attentional circuits, the patient may be able to perceive only one object at a time. Several studies have shown that an object located in the neglected region can be represented even when not consciously seen. How this occurs is controversial. A possible explanation is that the attention required for conscious awareness might be different from that required to form object representations, and in neglect, only the first type of attention is inhibited. See also: Attention: Change Blindness and Inattentional Blindness; Psychopathology and Consciousness.
74
Neglect and Balint’s Syndrome
Suggested Readings Bartolomeo P (2007) Visual neglect. Current Opinion Neurology 20: 381–386. Bartolomeo P, Thiebaut de Schotten M, and Doricchi F (2007) Left unilateral neglect as a disconnection syndrome. Cerebral Cortex 17: 2479–2490. Buxbaum LJ (2006) On the right (and left) track: Twenty years of progress in studying hemispatial neglect. Cognitive Neuropsychology 23: 184–201.
Driver J and Vuilleumier P (2001) Perceptual awareness and its loss in unilateral neglect and extinction. Cognition 79: 39–88. Husain M and Nachev P (2007) Space and the parietal cortex. Trends in Cognitive Science 11: 30–36. Milner AD, Karnath H-O, and Vallar G (2002) The Cognitive and Neural Bases of Spatial Neglect. Oxford: Oxford University Press.
Biographical Sketch
Piers Howe graduated as an exhibitioner from Oxford University in 1998, with a masters in physics. Winning a Presidential University graduate fellowship, he obtained his PhD from Boston University in 2003, under the guidance of Stephen Grossberg. His PhD thesis was titled ‘Cortical mechanisms of depth and lightness perception: Neural models and psychophysical experiments.’ He then worked as a Helen Hay Whitney postdoctoral fellow with Margaret Livingstone at Harvard Medical School, before moving on to Brigham and Women’s Hospital to work as a research fellow with Todd Horowitz and Jeremy Wolfe. His research has involved a variety of techniques including computational modeling, macaque neurophysiology, human fMRI, and human behavioral experiments. He has published articles on lightness perception, motion perception, depth perception, and visibility. More recently his focus has shifted to visual attention and to devising computational techniques for determining brain connectivity from fMRI data. He is a member of the Harvard fMRI Center for Neurodegeneration and Repair. He has taught a psychology course at the University of Massachusetts (Boston) and at a cognitive modeling course at Boston University. He lives in Boston, Massachusetts.
The Neural Basis of Perceptual Awareness J-D Haynes, Charite´ – Universita¨tsmedizin Berlin, Berlin, Germany ã 2009 Elsevier Inc. All rights reserved.
Glossary Binocular rivalry – When conflicting stimuli are presented to the two eyes, conscious perception can alternate spontaneously between the input to the left and the right eyes. Choice probability – The accuracy with which an organism’s perceptual choice in a decision task can be predicted from a neural signal. Global workspace theory – A theory that postulates the neural process underlying conscious awareness is a global distribution of information throughout the brain. Isomorphism – Literally ‘identity in structure’ typically refers to the notion that similarities between perceptual experiences are reflected in similarities of the underlying neural signals. Microconsciousness – The theory that perceptual awareness depends only on suitable representations in sensory brain regions and not on additional activity in prefrontal or parietal cortex. Perceptual threshold – When the intensity of a barely visible stimulus is gradually increased, there is no abrupt transition between ‘unseen’ and ‘seen’ but a gradual transition with an intermediate intensity range where the stimulus is sometimes seen and sometimes not. Reversible figures – Typically this refers to visual shapes that can be seen in different ways and can give rise to different geometric or semantic interpretations.
Introduction The rich qualitative properties of perception have made it a key focus of consciousness research for philosophers, neuroscientists, and psychologists
alike. The ‘qualia’ or ‘raw feels’ of sensory experience such as the redness of red, the timbre of an instrument, or the scent of a specific flower are the most vivid aspects of consciousness. In contrast, our abstract thoughts (such as the feeling of understanding a sentence) appear to have much weaker experiential qualities. Research on the neural correlates of perceptual consciousness has mainly focused on visual perception, which has been studied like no other field of neuroscience and especially like no other sensory modality. The visual system belongs to the best understood and most researched parts of the brain. A number of key concepts in neuroscience originated from the field of vision such as the concept of receptive field or the role of neural synchronization. The abundance of research in this field has yielded detailed mathematical models that make detailed quantitative predictions about a number of visual phenomena. Because of the rich body of research on visual perception, this article focuses primarily on the neuroscience of perceptual awareness in the visual modality. Many of these findings can be transferred to other modalities where phenomena often have direct counterparts in the visual system.
Crossing the Threshold to Awareness A starting point for an investigation into the mechanisms of conscious visual experience could be to compare cases where stimuli are clearly visible to cases where stimuli fail to reach awareness. For example, if the intensity of a weak, invisible stimulus is gradually increased, it will at some point be strong enough to reach awareness. The intensity where the transition from ‘unseen’ to ‘seen’ occurs is called the perceptual threshold. To compare neural processing with and without awareness, one could conduct a simple experiment 75
76
The Neural Basis of Perceptual Awareness
that contrasts neural responses to stimuli that are either above or below the threshold of perception. Any corresponding differences in brain activity could reflect the effect of awareness on neural processing. However, the comparison between seen and unseen stimuli would be confounded because in general the physical intensity of visible stimuli is stronger than the intensity of invisible stimuli. It would thus remain unclear whether any observed neural effects are indeed due to the difference in awareness or due to differences in the physical characteristics of the stimuli. There are however ways to compare ‘seen’ and ‘unseen’ stimuli while at the same time avoiding stimulus confounds. When the intensity of a simple stimulus is gradually increased, there is no abrupt transition between ‘unseen’ and ‘seen’ at the perceptual threshold. Instead there is a range of intensities where the stimulus is sometimes seen and sometimes not, thus yielding a certain percentage of ‘seen’ and ‘unseen’ responses. Because there is no sharp transition from unseen to seen, the perceptual threshold is usually defined probabilistically as a specific proportion of ‘seen’ judgments (or alternatively a specific proportion of correct discriminations between stimulus present and stimulus absent). The gradual transition from ‘unseen’ to ‘seen’ can be used to separate physical stimulus properties from awareness in two different ways. The first approach is based on the shape of the threshold function that relates increases in physical stimulus intensity to the monotonously increasing proportion of ‘seen’ responses. The threshold function is s-shaped, thus visibility increases slowly for low and high physical intensities, but strongly for intermediate intensities. Because the s-shape means that visibility undergoes a nonlinear change in a range where physical stimulus properties change linearly, this allows one to separate the physics from perception by identifying brain regions that exhibit an increase in response amplitude that matches the s-shaped threshold function. Using this approach several studies have shown tight links between threshold functions for stimulus intensity and signals in early visual cortex. Threshold functions for the identification of more complex features such as objects are also closely linked to response profiles of cells in brain regions specialized in object recognition.
There is a second, even more powerful, way to study visual awareness using the gradual, probabilistic nature of perceptual thresholds. Stimuli that only reach awareness on a certain proportion of trials are very useful because they allow one to directly compare conscious and unconscious trials for the same physical stimulus parameter. For example, trial-by-trial fluctuations in perception of simple pattern stimuli are reflected by corresponding changes in activity already in primary visual cortex. Thus, already the earliest regions of the cortical visual system can closely reflect conscious visual perception of simple stimulus features. Some studies indicate that already at early stages of processing, the effects of consciousness can be stronger than the effects of physical stimulus characteristics. The differences between processing of stimuli that do or do not reach awareness is also manifest at much higher brain regions, including regions in the prefrontal cortex that are involved in top–down control of processing and in behavioral report. A broader conceptual framework for understanding what happens in the brain when humans are viewing stimuli around the threshold to awareness is offered by perceptual decision making where subjects are required to perform simple detection and discrimination tasks. In a detection task a subject is asked to judge whether they believe to have seen a stimulus on an individual trial or not. Discrimination tasks come in two basic variants. Either the observer is shown one stimulus and has to judge which of several potential alternative stimuli it was. Or the observer is shown several stimuli and has to judge which one is which. Discrimination tasks do not probe for awareness of a stimulus but for awareness of a difference between stimuli. Performance in perceptual decision making is often accounted for by a sequence of simple information processing steps. In a first step, the presentation of a stimulus evokes a neural process encoding ‘sensory evidence’ about the stimulus presented. A second step consists of a ‘decision variable’ that is derived from the sensory evidence. The decision variable collapses all available sensory information in a way that provides for an efficient decision given the current behavioral goals. Finally, the values of the decision variable are mapped to a set of ‘judgments.’ For categorical
The Neural Basis of Perceptual Awareness
judgments such as ‘stimulus A’ versus ‘stimulus B,’ these signals reflecting the outcome of the decision are necessarily dichotomous and are typically directly related to specific motor commands that are used to indicate the judgment. Several studies have used a special type of motion stimuli to unravel the sequence of steps involved in perceptual decision making, so-called ‘random dot kinematograms.’ These are a blend of a ‘signal’ and a ‘noise’ stimulus. The signal stimulus is composed of a field of random dots coherently moving in one direction plus a noise stimulus consisting of randomly moving dots. The task of the subjects is to detect the drift direction of the coherently moving stimuli. The more the stimulus consists of signal dots and the less of noise dots the better the drift direction can be seen, thus yielding a threshold for perceptual motion detection. These stimuli allow one to calculate a so-called ‘choice probability’ that describes the accuracy with which the perceptual choice for ‘unseen’ and ‘seen’ is predicted by the activity of neurons in a specific area. A number of single-cell recordings in animals performing choice tasks with such stimuli has revealed that signals in regions of the brain specialized for motion processing partly predict the outcome of an animal’s decision – and hence presumably their perception. However, the prediction from single cells in these regions is far from perfect and choice probabilities are barely above chance. This suggests that the perceptual decision is either encoded in brain areas further downstream, or that it is encoded in pools of neurons. Perceptual decision making models can partially explain human perceptual choices at the threshold to awareness. But additional assumptions need to be made to account for what a person consciously sees in such tasks. One question is whether the perceptual experience is more closely reflected by the sensory evidence or by the decision variable. If only a single, simple stimulus is presented, the decision variable can be equated with the sensory evidence. But tasks where multiple stimuli have to be compared require a decision variable that computes a comparison, hence a relational property. In these cases the decision variable cannot directly reflect what we see but it reflects differences between things we see. A different question is how the decision making process
77
relates to a person’s subjective confidence in the accuracy of their decision. Confidence in a decision is frequently taken to be a good indicator of awareness based on the notion that if we are conscious of something we know that we see it and can be confident about our judgments. But most perceptual decision making models do not treat unseen and seen conditions as qualitatively different cases where visible stimuli undergo a different processing stream than invisible stimuli. Thus, they for example do not capture the finding that visible and invisible stimuli undergo different depths of processing in the brain.
Visual Competition: Masking and Rivalry A different way to cross the threshold between ‘unseen’ and ‘seen’ without changing the intensity of a stimulus is to render its perception difficult by introducing additional, competing stimuli. A large number of experimental approaches follow this logic. For example, in motion-induced blindness a target stimulus can pop in and out of awareness when it is presented in the vicinity of a moving set of dots. In these cases it appears as if the target fails to win the competition for awareness against the highly salient moving stimulus. A related phenomenon is flash suppression where a brief flash in the vicinity of a target (presented either to the same or the opposite eye) can strongly reduce its visibility. Flash suppression has been shown to affect processing already very early in the visual system.
Visual Masking One of the most prominent experimental procedures for manipulating awareness is visual masking, where the visibility of a target stimulus is decreased by presenting it in close spatial and temporal proximity to a so-called ‘mask.’ For example, if a target image alone is presented for brief periods it can normally be perceived quite effortlessly. However if the brief target is immediately followed by a second image consisting of an arrangement of random lines and patterns, its visibility is strongly reduced. This phenomenon is
78
The Neural Basis of Perceptual Awareness
known as backward masking and the mask image is known as a pattern mask. Visual masking has frequently been used to study the neural correlates of consciousness. Depending on conditions, masking can lead to decreases of brain activity at various stages of the visual system. But despite being invisible, masked targets can undergo a considerable degree of processing in the visual system. Object recognition is disrupted by pattern masking only at later stages of the visual system such as the lateral occipital complex (LOC) which is specialized for recognition of objects. There are a number of different approaches to visual masking. Masks can be effective when presented either before or after the target stimuli and can be spatially either overlapping or nonoverlapping. There are also several related phenomena, including object substitution masking and crowding. A special case is so-called metacontrast masking, where masks do not spatially overlap with the targets but share common contours with them. Metacontrast masks are most effective when there is a small time delay between the target and the mask. With very brief delays or long delays the visibility is high, thus the relationship between the target visibility and the delay follows a characteristic u-shaped function. For this reason, metacontrast masking is particularly interesting because the interaction between the target and the mask has to bridge not only space but also has to seemingly operate backwards in time. The u-shaped function can be used as a signature to identify the neural locations of masking effects in the brain. Although neural activity in V1 can be disrupted by metacontrast masking, this occurs only at later stages of processing, whereas an initial transient response to the target remains unaffected. Information about masked orientation stimuli can still be available at the level of primary visual cortex, suggesting a dissociation between awareness and encoding of features in V1. If targets and masks are presented to different eyes, the masking effects are only seen in higher areas, again suggesting that the contents of consciousness are encoded beyond monocular processing in primary visual cortex. The strongest effects of metacontrast masking occur in higher levels of the visual system and in regions of the prefrontal cortex involved in high-level executive control.
Reversible Figures and Binocular Rivalry In masking, the competition between the target and the mask is resolved in favor of the mask and the target remains invisible. But when the visual system is confronted with other types of competing or rivaling stimuli, it can sometimes oscillate between different interpretations of the visual input. One type of such multistable stimuli are reversible figures. These are typically 2D shapes that can be interpreted in different ways, such as the famous Necker cube that can be viewed either as a cube seen from the top or from below depending on which square is interpreted as being in front. Many such reversible figures are known, including stimuli that alternate between different directions of motion (e.g., stroboscopic alternating motion and 3D-structure from motion stimuli) or different semantic interpretation (e.g., a rabbit or a duck). Reversible figures are particularly interesting because they open the possibility to study changes in the contents of consciousness without any corresponding change in physical stimulation, thus allowing disentangling conscious perception from sensory factors. Extensive research on reversible figures has shown that perceptual changes coincide mainly with changes in activity patterns of those brain regions that are specialized for the specific contents of perception. When a face becomes visible during reversible perception, the so-called fusiform face area is activated, a brain region that is specialized for processing of faces. When the perceived direction of motion of similar displays changes, the motion-processing region MT is activated. Choice probabilities for perceptual reports of ambiguous 3D structure from motion stimuli are very high even for individual cells in motionprocessing regions of the brain. This suggests that changing the contents of consciousness involves mainly processes in regions specialized for these contents. Similarly, the transition periods between different percepts seem to involve high-level in prefrontal cortex, suggesting that the stability of our conscious perception might be regulated in high-level executive control regions. A phenomenon similar to reversible figures is binocular rivalry, where a competition arises not between two interpretations of the same figure but
The Neural Basis of Perceptual Awareness
between a conflicting stimulation of the two eyes. When conflicting stimuli are presented to the two eyes, visual perception cannot fuse them. Instead, perception alternates between seeing the stimuli presented to the left and to the right eye. There has been a long controversy regarding the exact locus of such rivalry in the brain. One dominating view is that rivalry is due to a conflict between monocular populations of cells in the early stages of the visual system. Information about the eye-of-origin is largely lost beyond primary visual cortex, and truly monocular cells can only be found in V1 and earlier stages of the visual system. In this monocular account of rivalry, the stimulus presented to one eye is suppressed from awareness because the input to this eye is attenuated at a monocular level of processing. This theory was supported by several behavioral findings. For example, the sensitivity to input from the suppressed eye is reduced, suggesting that processing of the suppressed eye is reduced. Similar support for a monocular selection was that if the input is exchanged between the dominant and suppressed eyes, perception in most cases tends to follow the eye, not the stimulus. However, an alternative pattern-based view was subsequently suggested. This was based on findings when mixed images are presented to the two eyes that allow for perceptual grouping between the two eyes. To create such stimuli one starts with two images, say a face and a tree. The next step is to mix the two images by exchanging coherent subregions in one image by the corresponding regions in the other image. This results in two complementary images, each showing sections of a face in some regions and sections of a tree in other regions. The question is: If perception during rivalry is dominated by the input from only one eye at a time, one would expect perception to alternate between the patchy image presented to the left eye and the complementary patchy image presented to the right eye. Instead there is a tendency to see coherent percepts that combine input from both eyes into a meaningful figure. This suggests that perception during rivalry depends not only on the eye of input and thus the monocular account of rivalry cannot be the full truth. This pattern-based account of rivalry was supported by single-cell recording studies that suggested that rivalry only affects 20% of cells in regions V1,
79
the latest stage with substantial monocular information. Instead, the main effects of rivalry seemed to be restricted to higher levels of the visual system. In regions of the temporal lobe that are specialized in high-level object recognition, rivalry affects 90% of cells. Similarly, rivalry in humans has been shown to affect high-level regions of the visual system. For example, when rivalrous perception alternates between faces and houses, corresponding increases can be seen in functional magnetic resonance imaging (fMRI) signals measured from regions of cortex specialized for processing faces and houses. However, in the following years studies on humans have gathered evidence that rivalry might affect earlier stages of neural processing than previously believed. It has been repeatedly shown that rivalry affects fMRI signals in primary visual cortex and even has effects on subcortical processing in the lateral geniculate nucleus. The current view on the mechanisms of binocular rivalry combines the monocular and the pattern-based explanations and postulates a selection for awareness at multiple stages of the visual system. Thus, the access to awareness can be regulated at many levels of processing ranging from very early subcortical regions to prefrontal cortex.
Encoding the Contents of Consciousness Reversible figures and binocular rivalry have long dominated experimental research on visual awareness because they allow dissociating changes in conscious perception from mere stimulation factors. When the contents of consciousness change without corresponding changes in the intensity or level of awareness of conscious perception, this might help isolate where the contents of our consciousness are stored in the brain. The rationale is that a brain region that encodes the contents of our conscious perception will need to change its activity when the contents of perception change. However, at a closer look this approach has a severe flaw. It does not allow one to disentangle brain regions that specifically encode the ‘contents’ of consciousness from brain regions that are ‘unspecifically’ involved in switching between different contents.
80
The Neural Basis of Perceptual Awareness
Besides involvement during perceptual transitions, more needs to be demonstrated to show that a brain region encodes an aspect of conscious perception. It needs to be shown that the neural responses in this brain region change in a ‘contentspecific’ manner. The contents of consciousness can be described along a number of different dimensions. Neuropsychological data from patients with brain lesions show that different dimensions of perceptual space are encoded in a number of different brain regions. For example, there is a double dissociation between regions encoding for color and motion because awareness of both can be disrupted independently following lesions either to colorselective brain regions in the fusiform gyrus versus lesions to the motion-processing region MT. Studies on agnostic patients have revealed that the perception of complex objects can also fail independently of the perception of the simple features they are composed of. This independent drop-out of specific contents of consciousness following specific brain lesions suggests that different aspects of awareness are encoded in separable cortical regions. Lesions can provide valuable information about brain regions encoding different aspects of awareness. But they still leave several questions open. For example, lesions to primary visual cortex cause an almost complete loss of conscious visual perception. But this does not mean that all contents of our perception are encoded in V1. Cells in primary visual cortex encode flicker at much faster rates than can be perceptually resolved. If V1 really were encoding the contents of conscious perception, then one would expect that the temporal resolution of perception would match the temporal resolution of signals in V1, which is clearly not the case. Similarly, when a stimulus is flashed into just one of the two eyes, primary visual cortex still encodes the eye-of-origin. But perceptually a subject can typically not tell which eye was stimulated, demonstrating again that V1 encodes information that does not reach awareness. Thus, these dissociations raise doubt whether signals in primary visual cortex encode contents of visual consciousness or rather serve a role of relaying information into the cortical visual system without participating in awareness directly. Such dissociations between encoding of features in V1 and the properties of visual awareness can be
taken even a step further by examining precisely how much ‘information’ is encoded in a neural response and how this compares to the perceptual information available for awareness. Unperceived flicker and unconscious eye-of-origin are cases where a neural process has more information about a stimulus feature than is represented in consciousness. The complementary case would be if neural signals had less information about a stimulus than is encoded in perceptual consciousness. Take as an example color perception. For a neural population to encode color percepts, it must respond with at least one different state to each identifiable color hue. If there were fewer neural states than possible color percepts, then the neural population would not be powerful enough to encode all the possible perceptually distinguishable shades of color. This suggests a useful test: To find out whether a neural population encodes perceived color hue, one can try and decode the perceived hue from signals in that neural population. If there is indeed at least one neural signal that corresponds to each perceptual state, it should be possible to fully decode color perception from this neural signal. This argument directly addresses the content-specificity of neural correlates of consciousness. It can be used to rule out that a brain area that is involved in awareness is merely realizing unspecific enabling factors of awareness, rather than encoding specific perceptual contents. More generally, the contents of conscious perception can be described along a large number of dimensions. For each point in the visual field, we can define its brightness, its color hue and saturation, its contrast, its speed and direction of motion, depth, and many more. Furthermore, simple features can jointly form geometric shapes and even meaningful objects. The contents of consciousness along a number of different dimensions can be jointly viewed as a complex ‘perceptual space.’ Using an information-theoretic framework, one can directly search for the population of neurons that encodes each specific subdimension of this perceptual space. Each subdimension of perceptual space can be viewed as a data structure that is encoded in some brain region by some parameter of brain activity. There are several techniques available for assessing the information encoded in neural populations.
The Neural Basis of Perceptual Awareness
One approach is to record from ‘single cells’ and to relate their encoding of a stimulus with the perceptual information available to a human or animal. But this neglects the information contained in distributed networks of neurons. A more powerful approach is to take into account the full information contained in ‘populations of neurons’ within a brain region. For example, to understand how motion percepts are encoded in the motion-processing region MT, one has to take into account not only the activity of single neurons, but of the entire neural population. This is because even a cell that is not tuned exactly to the feature of interest can still carry information about this feature. In fact, information-theoretic analyses have shown that neurons carry only very little information about features where they show the strongest responses. This is because at the peak of the tuning curve the neural responses do not differ very much for different features. Most information is contained in the side bands of the tuning function, where the neuron changes its response rapidly with changes in physical features. Using such decoding approaches it has been shown that single neurons in regions of the medial temporal cortex that are involved in recognition and memory carry information about specific visual contents, such as thoughts about specific individuals. Despite the distributed nature of neural representation, certain cells can exhibit an incredible sparseness, meaning that each visual object is encoded by only a few cells. Such cells respond say to a picture of a specific person, but not to pictures of other people. Information-theoretic approaches to the encoding of contents of consciousness are very powerful. But, depending on the theoretical perspective, they could be considered too powerful, because they can extract information also from ‘unusual’ brain signals such as for example from the side bands of orientation tuning curves or from deactivations of brain activity. Some theories about awareness in contrast postulate that the contents of consciousness are encoded in the brain in the form of explicit rather than implicit representations. An explicit representation directly signals the presence of a particular feature and does not need further processing to be read out. For example, the encoding of a face in a face-selective cell constitutes an explicit representation whereas the representation of a face in the retina constitutes an implicit representation
81
because it requires additional processing to tell that a face is present in the spatial pattern of retinal signals. To access the information encoded in entire neural populations, it is necessary to record from multiple cells simultaneously using so-called multielectrode grids. Such recordings can only be done in animals and occasionally in humans with implanted electrodes for diagnosis of epilepsy. A noninvasive alternative approach is offered by decoding techniques for electroencephalography (EEG), magnetoencephalography (MEG), and fMRI signals. These noninvasive techniques cannot resolve individual cells; the resolution is not even sufficient to resolve individual cortical columns (the basic units of information storage in the human brain where cells encoding similar features are clustered together across a span of approximately 0.5 mm of cortex). But these noninvasive techniques can nonetheless provide a handle on information encoded at a fine-grained scale in the visual cortex. This is possible due to small fluctuations in the topography of cortical maps that produce interference patterns that can be picked up with a standard fMRI measurement grid. Using information-theoretic decoding techniques, it is possible to access the information contained in these interference patterns, for example, about a specific type of visual feature, and to compare it to the information available to a human subject who is currently perceiving the same feature. Such decoding-based neuroimaging has been used to reveal which brain regions encode which visual features and how such information relates to the information encoded in visual awareness. This has revealed that orientation-information that does not reach awareness can nonetheless be encoded at the level of primary visual cortex, another proof of a dissociation between V1 signals and visual awareness. Because brain areas further downstream in the visual system do not encode unconscious orientation information, it is plausible that the neural correlates of the contents of visual consciousness start later in the visual system. Interestingly, the same content-based neuroimaging techniques can be used to go even a step further in mapping conscious experience to brain processes. It is possible to investigate how relationships between elements of conscious experience are reflected in similar relationships between
82
The Neural Basis of Perceptual Awareness
brain activation patterns encoding them. Such an ‘isomorphism’ between phenomenal experience and brain activity can for example be found in the way objects are encoded in the human temporal lobe. The perceived similarity between objects is reflected in a similarity between the brain activation patterns in the object-recognition regions of the temporal lobe.
Awareness and Specific Brain Regions Area V1 or ‘striate cortex’ is an excellent showcase of the different types of arguments that are put forward when discussing the potential involvement of a brain area in conscious perception. The role of primary visual cortex in visual awareness has been heavily debated. Primary visual cortex is the first cortical stage of information processing and is thus an important entry point for visual information into visual cortex. Despite the fact that other entry points exist where subcortical regions directly project to more high-level visual areas, the majority of information enters the visual system through V1. The key role of V1 in awareness can be seen from the fact that a lesion in V1 will always cause a fully blind section or ‘scotoma’ in the visual field. The scotoma exactly matches the corresponding retinotopic location of the visual field. For example, if the upper left section of V1 is fully lesioned, this causes a blind region in the lower right quadrant of the visual field. Similar lesions in higher level visual areas typically do not cause a full dropout of visual sensitivity. They only affect the perception of specific features such as color, motion, or depth. Even lesions in V2 only affect visual acuity and contrast perception. Thus, it appears that V1 plays a special role among the visual areas in that it is necessary for visual awareness. This is supported by a number of demonstrations of close correlation between processing in primary visual cortex and visual awareness. Activity in V1 closely reflects perception of simple visual features such as contrast and brightness. Even simple forms of perceptual integration such as the perception of texture boundaries and contours can be explained from properties of V1 neurons. FMRI signals recorded in human V1 correlate with the
conscious percept during binocular rivalry. When experimental subjects train to see subtle differences between simple visual stimuli, a phenomenon called perceptual learning, the improvement in performance correlates with changes in tuning properties in primary visual cortex. However, when arguing for a role of a specific region in visual awareness from experimental correlations caution is required. The important question needs to be addressed, whether any correlations observed indeed reflect necessary conditions for awareness or whether they are purely incidental or ‘epiphenomenal.’ The retina can also be considered a necessary condition for visual awareness because a loss of both retinae causes a complete loss in visual perception. This does not mean that the retinae are strictly necessary for awareness. Instead it means that the retina is a necessary step in the normal causal chain of events leading to conscious perception. But conscious visual perception can also be caused by bypassing the retina and directly stimulating visual cortex using implanted electrodes in patients, during surgery or by using transcranial magnetic stimulation (TMS). Also, during visual hallucinations and visual imagery, V1 is not always involved. An alternative view would be that the disruption of visual perception following lesions to V1 simply reflects the loss of the major ‘input’ channel to the visual system. It does not directly imply that V1 is always required for every type of conscious visual experience. Furthermore, there are a number of reasons that have cast doubt on a close involvement of V1 in visual awareness. On the one hand there are theoretical reasons. For example, we can directly access, manipulate, and act upon information that is in our consciousness. To explain such access, it seems necessary to assume a direct projection from regions of the brain involved in high-level control of behavior to sensory regions encoding the contents of consciousness. However, there are no direct projections between prefrontal cortex and V1 that could support such access. A further argument can be obtained from the fact that there are striking dissociations between our conscious visual experience and encoding of information in V1. For example, we are not able to consciously tell whether a monocular stimulus is presented to the
The Neural Basis of Perceptual Awareness
left or right eye. But V1 encodes eye-of-origin and also many other features that fail to reach consciousness such as unconscious orientation-information and unperceived high-frequency flicker. In crowding, the visibility of a stimulus that is normally clearly visible is reduced by presenting other stimuli in the surrounding regions of the visual field. Stimuli that fail to reach awareness under crowding conditions are nonetheless processed at least up to primary visual cortex, because they leave traces of orientation-selective adaptation, suggesting that an encoding of information in V1 does not automatically lead to awareness. Even if V1 does not encode the contents of our consciousness, several findings suggest that it might nonetheless be required for awareness. In some patients large regions of visual cortex beyond V1 can continue to respond to visual stimuli despite lesions to V1 that preclude input arriving through primary visual cortex. This means that activity in such ‘extrastriate’ visual regions is not sufficient in itself to produce awareness without a contribution from V1. This is further supported by the fact that although residual visual sensitivity can remain in a scotoma, the subject does not subjectively feel to be seeing anything. A subject can be above chance in guessing which stimulus was presented, but they will have the impression to be seeing nothing in that region of the visual field. Such ‘blindsight’ is presumably mediated by pathways into extrastriate visual cortex that bypass V1, thus lending further support for the notion that activation of extrastriate cortex can be sufficient to support visually-guided behavior but is in itself not sufficient for awareness. The term ‘extrastriate cortex’ refers to a group of visually responsive brain regions beyond V1 or ‘striate cortex.’ An important feature of early extrastriate areas V2, V3, and V4 that they share with V1 is their retinotopic organization. This means that the topography of the visual field is largely preserved in the visual maps of V1–V4, despite undergoing a coordinate transform from Cartesian to polar coordinates. Because detailed spatial information is progressively lost at higher stages of the visual system beyond V4, this means that signals encoding the topographic spatial layout of visual perception can only be found in these early visual regions. Regions beyond V4 exhibit an increasing specialization for different visual features. Motion is
83
primarily processed in a dedicated area labeled MT located in the back of the temporal lobe. The main cortical color region is located in the fusiform gyrus at the bottom of the temporal lobe, although there is some debate as to whether this is the same region as retinotopic area V4. The main cortical region for object recognition is called the LOC and is located in the lateral occipital lobe and in the posterior fusiform gyrus. Visual contents can be selectively lost from consciousness. For example, color and motion can be independently disrupted following lesions to the corresponding brain regions. After lesions to specific regions of the ventral temporal lobe, patients can lose color perception, thus perceiving the world in shades of grey while other qualities of visual perception such as motion are spared. Similarly, lesions restricted to MT cause loss of motion perception without loss in color perception. The modular encoding of different aspects of visual perception is further supported by direct cortical stimulation to patients undergoing brain surgery. Depending on which visual region is stimulated, patients report seeing complex patterns, colors, or movement. An interesting property of cells at the higher levels of the visual system is that they respond independent of the detailed physical characteristics and context of a presented object. This phenomenon is termed ‘invariance.’ Say, an actress might be wearing a different dress and sporting a different hairdo, but still the cell would recognize the invariant person encoded in the image. But this invariance also comes at a price. If a cell responds to a complex visual object independent of the constituent features, it means the cell has lost all information about the fine-grained features such as brightnesses and colors that the object is made up of. This means that the encoding of contents of consciousness necessarily occurs at multiple levels. Brain regions coding complex, invariant aspects of our visual experiences can in principle not be coding the simple features. The fact that the processes underlying consciousness are fragmented, modular and multilevel is well documented, but this stands in direct contradiction to our impression that visual experience is unified rather than divided into a number of different features. The distributed encoding of the contents of consciousness in extrastriate visual areas requires additional
84
The Neural Basis of Perceptual Awareness
assumptions in order to explain the unity of consciousness. This problem is known as the ‘binding’ problem. The most prominent but also controversial explanation of binding assumes that features are bound by synchronization of cells in different brain regions. An important dissociation in the visual system is that between representations for action and representations for awareness. There are two major visual pathways in the extrastriate visual system. The dorsal pathway is involved in spatial transformations and actions, whereas the ventral pathway is involved in object recognition. Patients with lesions of the ventral pathway cannot describe a visual stimulus any more, but can still perform visually-guided actions to the input. Patients with lesions of the dorsal pathway show the opposite pattern of disorders. They cannot perform visually-guided behavior but can consciously describe the stimulus. This has been interpreted to mean that only the ventral visual pathway supports conscious perception, whereas the dorsal pathway supports unconscious visual guidance of behavior, similar to the blindsight that follows lesions to primary visual cortex. Supramodal regions beyond the visual system also play an important role in perceptual awareness. When a stimulus crosses the threshold to awareness, this also leads to changes in activation of regions of prefrontal cortex. For example, during studies where subjects are required to recognize masked versus unmasked words, conscious perception is correlated with increased activation in several prefrontal brain regions. Similarly, during perception in binocular rivalry, a frontoparietal network is involved in perceptual transitions, suggesting that these regions might be involved in awareness. One possibility is that these prefrontal brain signals reflect the global distribution of information as postulated by the so-called global workspace theory of consciousness. According to this theory, neural representations reach awareness when they are distributed to other brain regions. An alternative view is that content-specific brain activity within individual sensory brain regions is sufficient for awareness of these contents to occur, a theory referred to as microconsciousness. To date it is unclear whether the frontal and parietal processes involved in awareness are indeed content-specific and thus can be claimed to reflect a distribution of information. Instead they could
reflect unspecific processes, for example, they might be involved in causing awareness of a stimulus without themselves encoding the contents of consciousness. Alternatively they could be involved in the subject consciously noticing and reporting a change in the contents of their consciousness. Another important question for the neural mechanisms of awareness is whether they involve large-scale integrative processes that jointly involve a number of brain regions. One line of thought posits that awareness of representations is related to specific dynamic interactions between populations of neurons. For example, one controversial theory suggests that awareness and binding are both closely related to synchronization between the distributed population of cells encoding the different features of an object. A similar hypothesis states that awareness of representations is related to the stage of recurrent cortical processing where feedforward and feedback neural processes overlap in early sensory regions. This is supported by several invasive studies in monkey visual cortex, as well as by studies using TMS to knock out later stages of processing in V1. When MT, the main cortical motion-processing region, is stimulated using TMS, this can create the illusory perception of movement. When a second TMS pulse is administered to V1 just following the first pulse to MT, the perception of movement is abolished. This could mean that awareness of motion depends on intact feedback projections into V1. In contradiction to this finding, patients with the so-called Riddoch syndrome who have lesions to V1 can have selectively spared motion perception despite global blindness. Thus, recurrent processing involving V1 activity cannot be a strictly necessary condition for visual awareness of motion.
Attention and Awareness Awareness and attention are so closely related that many researchers consider them to reflect the same process. A key feature of both attention and awareness is their selectivity. Only a small amount of sensory information that is available at a given time reaches awareness. This limitation is most clearly apparent when there is a lot of competition for access to our consciousness. For example in a crowded room we might fail to immediately notice
The Neural Basis of Perceptual Awareness
a friend who is present although he is clearly visible. In such situations we have to consciously scan our environment by focusing our attention on one person at a time, and only when attention is focused on a known person we will be able to recognize them. Such cluttered scenes help understand the neural mechanisms that underlie the selectivity of visual awareness. According to the biased competition model of visual attention, stimuli are in competition for access to processing resources. When several stimuli are present in the visual field (or in the receptive field of cell), attention is required to bias or boost the processing of a selected visual feature, and the unselected stimuli are suppressed from further processing. Such effects have been demonstrated for single cells as well as for population responses measured with fMRI. The general finding is that competition increases at advanced stages of visual processing where receptive fields increase in size and competition occurs between an increasing number of elements in the visual field. This model of competitive interactions between processing units provides a model for understanding the selectivity of conscious visual experience. The selectivity of attention and awareness seems to imply that the two might be two aspects of the same process. In this view attention functions as a gatekeeper to consciousness. There are many demonstrations of the tight relationship between attention and awareness. Lesions to attentional control regions can cause a disorder of awareness known as visual hemineglect, where stimuli can fail to be noticed when presented in the contralesional visual field. Another striking example is a phenomenon called inattentional blindness, where subjects fail to notice highly salient events in their visual field when their attention is directed elsewhere, thus highlighting the importance of attention for visual awareness. A related phenomenon is changeblindness, where subjects fail to notice marked changes occurring in their visual environment. In the typical experiment, a person is monitoring a cluttered visual image that is repeatedly flashing on and off for any changes made between two successive presentations. Most people have the intuitive assumption that they perceive the entire visual scene surrounding them and thus would definitely notice a salient change. In contrast to this intuition, most people fail to perceive quite marked changes
85
unless explicitly paying attention to the particular region of the visual scene where the change occurs. These experiments can be useful for tracking down the neural correlates of consciousness. Major changes to visual scenes that go unnoticed can still continue to be registered and analyzed up to higher stages of the visual system. FMRI signals from high-level object-selective regions signal changes in a visual scene that are not registered. Similar findings have been obtained in single-cell recordings in human medial temporal areas related to object recognition and memory. Despite the close link between attention and awareness, there are also several findings that demonstrate that the two cannot be fully equated, although they are clearly closely related. One important question is whether stimuli outside the focus of our attention can really not reach our conscious awareness. There are several experiments that suggest that this might be the case. In one line of experiments, the attentional resources of subjects are engaged maximally at one region of the visual field, typically the center of gaze, by requiring them to perform a very difficult visual discrimination task. Despite the full engagement of attention, they can still continue to perceive certain stimuli presented at a different region of the visual field. Simple, salient stimuli such as colors or shapes can be identified without attention. Even more complex aspects of a visual scene can be perceived in parallel without requiring attention. For example, observers can readily and rapidly perceive whether an image contains an animal or not, even when the animal is hidden in a cluttered natural landscape. EEG signals recorded from humans during such tasks show signatures of rapid recognition as early as 120 ms after the onset of a stimulus. In contrast, complex combinations of geometric features cannot be perceived rapidly while attention is directed elsewhere. This suggests that the ambient ‘fringe’ surrounding the focus of our attention can still enjoy a fair degree of complex, high-level cortical processing and lead to awareness.
Summary A large number of experimental studies have contributed strongly to our understanding of the neural mechanisms of visual perceptual experience.
86
The Neural Basis of Perceptual Awareness
The contents of perceptual experiences are encoded in distributed subregions of modality-specific cortex. The access to consciousness appears to be regulated at multiple stages of processing, reaching from early subcortical processing to high-level regions involving prefrontal and parietal cortex. However, several theoretical debates regarding the specific processes involved still await further clarification. It is currently unclear whether representations that reach awareness are globally made available for further processing across the brain. And it is also unclear how the many subdimensions of perceptual space that are distributed across many brain regions give rise to a unified perceptual experience. See also: Attention: Change Blindness and Inattentional Blindness; Bistable Perception and Consciousness; Blindsight; Neglect and Balint’s Syndrome; Perception: The Binding Problem and the Coherence of Perception; Visual Experience and Immediate Memory.
Suggested Readings Baars BJ (2002) The conscious access hypothesis: Origins and recent evidence. Trends in Cognitive Sciences 6: 47–52. Crick F and Koch C (1998) Consciousness and neuroscience. Cerebral Cortex 8: 97–107.
Dehaene S and Naccache L (2001) Towards a cognitive neuroscience of consciousness: Basic evidence and a workspace framework. Cognition 79: 1–37. Driver J and Vuilleumier P (2001) Perceptual awareness and its loss in unilateral neglect and extinction. Cognition 79: 39–88. Engel AK and Singer W (2001) Temporal binding and the neural correlates of sensory awareness. Trends in Cognitive Sciences 5: 17–25. Gold JI and Shadlen MN (2001) Neural computations that underlie decisions about sensory stimuli. Trends in Cognitive Sciences 5: 10–16. Haynes JD and Rees G (2006) Decoding mental states from brain activity in humans. Nature Reviews Neuroscience 7: 523–534. Heeger DJ (1999) Linking visual perception with human brain activity. Current Opinion in Neurobiology 9: 474–479. Lamme VAF (2003) Why visual attention and awareness are different. Trends in Cognitive Sciences 7: 12–18. Lamme VAF, Supe`r H, Landman R, Roelfsema PR, and Spekreijse H (2000) The role of primary visual cortex (V1) in visual awareness. Vision Research 40: 1507–1521. Milner AD and Goodale MA (1995) The Visual Brain in Action. Oxford: Oxford University Press. Rees G, Kreiman G, and Koch C (2002) Neural correlates of consciousness in humans. Nature Reviews Neuroscience 3: 261–270. Thiele A and Stoner G (2002) Neuronal synchrony does not correlate with motion coherence in cortical area MT. Nature 421: 366–370. Tong F (2002) Primary visual cortex and visual awareness. Nature Reviews Neuroscience 4: 219–229. Zeki S (1993) A Vision of the Brain. Oxford: Blackwell.
Biographical Sketch
John-Dylan Haynes is Professor for cognitive and computational neuroimaging at the Bernstein Center for Computational Neuroscience, Berlin, Germany. His main interest lies in consciousness research where he has worked on the neural mechanisms of visual awareness and on unconscious neural determinants of human decisions. His research combines psychophysics, neuroimaging and advanced data analysis methods, such as connectivity measurements and multivariate pattern recognition. The latter can be used to decode mental states from patterns of human brain activity, thus constituting a rudimentary form of ‘‘mindreading.’’ He is currently a board member of the Association for the Scientific Study of Consciousness (ASSC) and will be organizing the 13th meeting of the ASSC in Berlin (with Michael Pauen).
Neurobiological Theories of Consciousness S Kouider, Laboratoire des Sciences Cognitives et Psycholinguistique, CNRS/EHESS/ENS-DEC, Paris, France ã 2009 Elsevier Inc. All rights reserved.
Glossary Neural correlates of consciousness – They are defined by Christoph Koch as ‘‘The minimal set of neuronal mechanisms or events jointly sufficient for a specific conscious percept or experience.’’ They allow to avoid the difficult problem of directly looking for neural bases. Panpsychism – Reflects the philosophical doctrine that everything (in Greek, ‘pan’) has a mind (‘psyche’) and is therefore conscious. Some theories presented in this article endorse a certain form of panpsychism in which anything that transmits information is in a way conscious. The hard problem – It is the problem of explaining how and why we have the subjective experience of consciousness. It is often contrasted with the easy problem, which consists of describing consciousness as the cognitive ability to discriminate, integrate information, focus attention, etc.
Cognitive Influences on Neurobiological Accounts Regarding the influence of cognitive theories, the majority of neurobiological accounts can be seen, in fact, as extensions of preexisting cognitive theories (e.g., for instance global workspace theories). Indeed, one of the main tasks exercised by neurobiologists in the last two decades has been to search for cerebral or neuronal equivalents to the functional elements constituting cognitive models (e.g., the dorsolateral prefrontal cortex for voluntary control, or long range axons for connecting brain regions associated with ‘unconscious’ and ‘conscious’ processing). Of course, many neurobiologists disagree with this approach. Consciousness, because it is a biological problem, should be reframed the other way around, by focusing primarily on its structural basis rather than relying on cognitive theories, often considered too speculative. Therefore, many neurobiologists consider that an ideal neurobiological science of consciousness should focus on neural structures and mechanisms in order to understand how the organic matter constituting the brain creates consciousness.
Introduction
The Hard Problem for a Neurobiology of Consciousness
Neuroscientists working on the issue of consciousness consider that it is a biological problem. They assume that we will understand how and why we are conscious by studying the cerebral and neuronal features of the brain. These theories have largely benefited from the recent advances in neuropsychology, neurophysiology, and brain imaging in particular. However, neurobiologists have also been influenced, on the one side, by cognitive theories aimed at characterizing the psychological determinants of consciousness, and on the other side, by philosophical issues related to the mind– body problem.
Yet, studying the neural mechanisms ‘leading to’ consciousness, trying to explain the ‘emergence’ of consciousness, or focusing on how the brain ‘creates’ consciousness, as often described in neurobiological literature, sounds as if it involved an immaterial soul that would magically arise from the brain. This is not a new issue for philosophers who have also been wondering about the equivalent mind–body problem since antiquity. More than a century ago, the contemporary ‘brainconsciousness’ problem was well captured by Thomas Huxley’s famous remark: ‘‘How it is that anything so remarkable as a state of consciousness 87
88
Neurobiological Theories of Consciousness
comes about as a result of irritating nervous tissue, is just as unaccountable as the appearance of the djinn when Aladdin rubbed his lamp in the story.’’ The same issue applies today: understanding consciousness as an ‘emergent’ property ‘arising’ from functional elements of the neurocognitive architecture, without falling on a dualistic position where consciousness lies somewhere outside of the brain, poses serious epistemological difficulties and leads to the so-called hard problem of consciousness. Indeed, many philosophers have concluded that there is not one single problem, but actually two problems that are faced by anyone trying to understand consciousness: they distinguish between the so-called easy problem and hard problem. In a nutshell, the easy problem consists in relying on objective measures of conscious processing in order to explain how one is able to discriminate sensory events, integrate information, report mental states, focus attention, etc. By contrast, the hard problem consists in explaining the firstorder, subjective nature of qualias and phenomenal states, the ‘what is it like to be conscious’ as well as how and why we experience consciousness at all. Addressing the hard rather than the easy problem of consciousness constitutes an important epistemological constraint put forward by philosophers. In particular, contemporary philosophers such as Joseph Levine and later David Chalmers have argued that trying to resolve the hard problem leads to an ‘explanatory gap’ that science is unable to cross, at least today. Indeed, they stress the fact that it appears impossible to demonstrate that a neural structure leads to a conscious experience, while denying the reverse possibility. In addition, given how different they are, reducing phenomenal states to neural states appears almost impossible. Looking for Neural Correlates, Not Neural Bases Should neurobiologists then give up on addressing this issue? Most neurobiologists acknowledge the existence of a hard problem. However, they also endorse the principle that further scientific investigations will ultimately allow us to resolve it. Others explicitly deny the existence of a hard problem in the Chalmerian sense. For some,
assigning too much importance to the explanatory gap might actually turn out to be counterproductive and impedes rather than facilitates scientific progress. Accordingly, neurobiologists have mostly focused on the easy problem, considering that this strategy will progressively get us closer to understanding the full issue. They extended the ‘contrastive analysis,’ originally put forward by Bernard Baars, from the cognitive to the neurobiological domain. While this method initially consisted in contrasting conscious and unconscious processes in order to characterize their cognitive features, the neurobiological approach aims at characterizing the neural features. A typical example consists in comparing the cerebral activity when subjects are presented with subliminal stimuli they cannot report (unconscious processing) with that of visible stimuli they can report (conscious processing). In other terms, the current first step in trying to understand the link between consciousness and the brain consists in finding out which neural components are specifically involved during conscious processing, but importantly not during unconscious processing. Francis Crick and Christoph Koch have coined the term ‘neural correlates of consciousness’ (NCC; see Glossary) in order to describe this epistemological approach. According to them, the best strategy for a neurobiological science of consciousness is to search for the NCC. Underneath this approach is the crucial principle that ‘correlates’ do not imply any relation of causality between the occurrence of conscious mental events and their associated physiological structure. Consequently, this strategy has the advantage of leaving aside, at least for the moment, the hard problem of finding the neural ‘bases’ of consciousness. In the following sections, I will provide an overview of the current most influential neurobiological theories of consciousness. These theories will be largely described in an independent manner, such that each of them can be understood individually, that is, without having to frame it in the context of alternative accounts. Only later, in the section labeled ‘Neurobiological standpoints on the hard problem’ will I evaluate their explanatory power by confronting them in relation to some important conceptual issues (e.g., dissociating access vs. phenomenal consciousness, dissociating attention vs.
Neurobiological Theories of Consciousness
consciousness, panpsychism). I will conclude by emphasizing how promising these theories are in getting us closer to resolve the issue of the hard problem. From Globalist to Localist Accounts of Consciousness Neurobiological theories of consciousness differ in many respects. One way to portray them in a coherent manner is to follow the large spectrum ranging from globalist to minimally localist accounts. By globalist or localist I refer to the size of the brain states that are assumed to be sufficient for consciousness (extended to large parts of the brain vs. focalized to specific and small brain areas, respectively). I will mainly restrict this review on the cerebral level and exclude alternative theories that are more globalist, in the sense that consciousness encompasses more than neural activity in the brain (e.g., theories proposing that consciousness reflects modification in the electromagnetic field surrounding the brain, as proposed by J. McFadden and by R. John), or, conversely, theories that are more localist, by focusing on single neurons or even lower structural levels (e.g., the quantum-level theory of microtubules by S. Hamerof and R. Penrose). Such theories remain excessively speculative and unspecified to be included in a serious review, at least for the moment. The Reentrant Dynamic Core Theory The Reentrant Dynamic Core theory, proposed by Gerald Edelman and Giulio Tononi, is arguably the most globalist account of consciousness. Indeed, in this framework, consciousness is not to be localized in dedicated brain areas or with particular types of neurons. Rather, it is the result of dynamic interactions among widely distributed groups of neurons in the entire thalamocortical network (the ‘dynamic core’). This theory offers an interesting tentative of unifying the hard and easy problems, providing a neurobiological explanation for qualias, and explaining both phylogenetic and ontogenetic aspects of the development of consciousness in humans and other species. Yet, although this theory is very appealing, especially given its explanatory power, it is also highly speculative, and based on several assumptions that
89
remain to be demonstrated. I will further discuss the speculative aspects of this theory later (see the section ‘Neurobiological standpoints on the hard problem’). For now, I shall provide an overview of the core assumptions underlying this theory. In order to appreciate the specificity of the Reentrant Dynamic Core theory, it is important to understand that, regardless of its explanation for consciousness, it offers an alternative view on brain structures, considering the wiring of the brain into neuronal assemblies as the result of variation and selection mechanisms that are analogous to those underlying evolutionary theories. This macrolevel account of brain development, formally developed by Edelman in 1978, and called the theory of Neuronal Group Selection (also called the theory of Neural Darwinism) constitutes a key element for understanding the development of consciousness. According to the Neuronal Group Selection theory, the brain is assumed to be a selectionist system in which variant groups of neurons are selected over others in three steps. The first one, termed developmental selection, happens during embryogenesis and early development, and it is largely influenced by epigenetic factors. It consists of several processes such as cell division, cell death, extension, which has the consequence of connecting neurons together into a large number of variant neuronal circuits (labeled primary repertoires), and elimination, which targets unconnected neurons in particular. The second step is called experiential selection and lasts from early infancy through all of the lifespan. It consists of reinforcing, through the influence of behavior and experience with the environment, the synaptic connections of some variants over others, leading to secondary repertoires of neurons. The final step, that of reentry, consists of the formation of massively parallel reciprocal connections among distant maps of neuronal repertoires, allowing them to exchange signals and be spatiotemporally coordinated. In the Reentrant Dynamic Core theory, consciousness results exactly from this mechanism of reentry among distant groups of neurons within the dynamic core of thalamocortical connections. The spatiotemporal coordination provided by reentry allows the binding of several elements into a single and coherent object or event, providing a solution to the binding problem. It also allows to
90
Neurobiological Theories of Consciousness
explain why conscious experiences appear to be unified. Particularly important in this theory is the reentry of information between groups of neurons dealing with perceptual categorization (i.e., in posterior areas) and the more frontally located systems responsible for value-category memory. Indeed, the latter will constrain the selectionist process by modulating or altering the synaptic connections within groups of neurons, as a consequence of their influence on behavior (pleasure, pain, etc.). Another key aspect to this whole framework is that groups of neurons are constantly in competition and their survival depends on creating or reinforcing synaptic links with other groups, such as to form large assemblies of reentrant neuronal maps in the dynamic core. The victorious assembly (or ‘coalition’ in Crick and Koch’s terms) will lead to consciousness, at least for a few hundred milliseconds, until a new coalition of neurons bypass it. Indeed, because the variant groups of neurons are assumed to be degenerate, which means that different variants of neuronal assemblies can actually carry out the same function and have the same output, each integrated state in the dynamic core is followed by yet another and differentiated neural state in the core. As such, coalitions of neurons leading to consciousness are temporally transient by nature and widespread along variant regions of the whole dynamic core. Because interactions within the dynamic core are constantly moving, it explains the diversity of consciousness, yet the constant integration through reentry explains the unity of consciousness. Note then that the victory of a coalition of neurons in leading to consciousness is not just a matter of its size; rather it depends on its complexity, that is, its ability to generate at the same time an integrated scene (i.e., appearing as a unitary event) and a differentiated scene (i.e., which can be highly discriminated from other scenes). Measures of complexity in relation to consciousness have been further developed by Tononi in his recent Information Integration theory of consciousness (see the article ‘cognitive theories of consciousness’ in this volume). According to Edelman and Tononi, these highly discriminatory properties of neural complex systems are the qualias that have been torturing philosophers, nothing more, nothing less! Finally, they distinguish between primary consciousness that
allows for a perceptual organization of the environment and whose characteristics have been presented above, and higher-order consciousness that is possessed by humans, and which is related to linguistic and symbolic mental activities. The latter requires further reentry with additional brain regions such as those involved in language production and comprehension. In sum, this theory assumes that the brain is a dynamic complex system in which consciousness emerges from the interactivity itself. Instead of involving top-down communication between dedicated areas, the Reentrant Dynamic Core theory involves regions in the cortex in active communication with one another and with associated nuclei in the thalamus. Consciousness arises from the differentiated integration of activity in these areas, as information is transmitted recurrently, with local groups of neurons performing their specialized and discriminatory function, while at the same time being unified with other neuronal groups of the dynamic core. Before concluding on the characteristics of this theory, it is important to point out that although the Reentrant Dynamic Core can be considered as a globalist theory, in the sense that it can involve a large set of regions in the brain, it is also a gradual theory of consciousness that can result from minimal neural networks. Indeed, reentry among even a small number of neurons, as far as it reflects differentiated integration, will induce consciousness to a certain degree. We will come back to this theoretical aspect below.
The Global Neuronal Workspace Theory The Global Neuronal Workspace theory proposed by Stanislas Dehaene with Lionel Naccache and Jean-Pierre Changeux is currently the most explicit and most functional account of the cerebral architecture underlying conscious access. Its functionality has rendered this theory very popular in neuroscience circles. Yet, as we shall see below in the section ‘Neurobiological standpoints on the hard problem,’ this theory has also been greatly criticized for probing functionality at the price of sacrificing some phenomenological aspects of consciousness. For now, let us focus on the main characteristics of this theory.
Neurobiological Theories of Consciousness
This theory is a perfect example of the neurobiological extension of a cognitive theory, that of a global workspace, originally proposed by Bernard Baars in 1988 (see the article ‘Cognitive theories of consciousness’ in this volume). Dehaene’s theory assumes that neurocognitive architecture is composed of two qualitatively distinct types of elements. The first type is represented by a large network of domain-specific processors, in both the cortical and subcortical regions that are each attuned to the processing of a particular type of information. For instance, the occipitotemporal cortex is constituted of many such domain-specific processors, or ‘cerebral modules,’ where color processing occurs in V4, movement processing in MT/V5, face processing in the fusiform face area (FFA), etc. Although these neural processors can differ widely in complexity and domain specificity, they share several common properties: they are triggered automatically (i.e., mandatorily, by opposition to voluntarily), they are encapsulated (their internal computations are not available to other processors), and importantly, they largely operate unconsciously. Conscious access involves only the second type of element, namely, the cortical ‘workspace’ neurons that are particularly dense in the prefrontal, cingulate, and parietal regions. These neurons are characterized by their ability to send and receive projections to many distant areas through longrange excitatory axons, breaking the modularity of the nervous system and allowing the domainspecific processors to exchange information in a global and flexible manner. The global workspace is thus a distributed neural system with long-distance connectivity that can potentially interconnect multiple cerebral modules through workspace neurons. It offers a common communication protocol, by allowing the broadcasting of information to multiple neural targets. One important aspect of this theory is that encapsulation and automaticity are less rigid than traditional modularist (i.e., Fodorian) accounts of the cognitive system (see the article ‘Cognitive theories of consciousness’ in this volume). Indeed, once a set of processors have started to communicate through workspace connections, this multidomain processing stream also becomes more and more ‘automatized’ through practice, resulting
91
in direct interconnections, without requiring the use of workpace neurons and without involving consciousness. Another important aspect that follows from this theory is that information computed in neural processors that do not possess direct long-range connections with workspace neurons, will always remain unconscious. This idea is rooted in the work of Crick and Koch, postulating that neural activity in V1, because it does not project toward prefrontal neurons, does not participate directly in visual consciousness. Note that for a mental object to become conscious, it is not sufficient that its activity gives input to the global workspace. Two other conditions have to be met. One is that the content of the mental object must be represented as an explicit firing pattern of neuronal activity, that is, a group of neurons that unambiguously indexes its relevant attributes. A final and important condition is that the top-down amplification mechanisms mobilizing the long-distance workspace connections render the content of consciousness accessible, sharpened, and maintained. A mental object, even if it respects the two first conditions (explicit firing and accessibility to workspace neurons) will still remain buffered in a ‘preconscious’ (and thus a nonconscious) store until it is attended to and its neural signal amplified. Therefore, top-down attention in this framework is a necessary condition for consciousness. Whether being conscious requires top-down attention constitutes, as we will see below, one of the most debated aspects among neurobiological theories of consciousness. The Duplex Vision Theory The Duplex Vision theory proposed by David Milner and Melvyn Goodale postulates that visual perception involves two interconnected, but distinctive pathways in the visual cortex, namely, the dorsal and the ventral stream. After being conveyed along retinal and subcortical (i.e., geniculate) structures, visual information reaches V1 and then involves two streams. The ventral stream projects toward the inferior part of the temporal cortex and, according to this theory, serves to construct a conscious perceptual representation of objects, whereas, the dorsal stream projects toward the posterior parietal cortex and mediates
92
Neurobiological Theories of Consciousness
the control of actions directed at those objects. Apart from these structural considerations, the two streams also differ at the computational and functional levels. On the one side, the ventral stream conveys information about the enduring (i.e., longlasting) characteristics that will be used to identify the objects correctly, and subsequently to link them to a meaning and classify them in relation to other elements of the visual scene. Computing these enduring characteristics involves relatively long and costly computations. On the other side, the dorsal stream can be regarded as a fast and online visuomotor system dealing with the moment-tomoment information available to the system, which will be used to perform actions in real time. It should be noted that this dissociation between ventral and dorsal pathways has sometimes been misunderstood as equivalent to conveying the ‘what’ and ‘where’ information in the visual cortex, respectively. However, structural and spatial attributes can conjointly be used by both systems. For instance, visual information such as the size, geometrical structure, and location of a target object might be computed both by the dorsal steam, in order to grasp and reach the object, and by the ventral stream in order to segregate the object from a complex visual scene. While the ventral steam involves object perception by comparison with visual perceptual attributes stored in memory, the dorsal steam, because it is fast and updated in real time, involves no storage of the visuomotor attributes extracted from the object, nor the motor program resulting from actions upon that object. In other terms, while the ventral stream can be seen as conveying vision for perception, the dorsal stream is concerned with vision for action. Both systems work together in the production of adaptative behavior, but they can also function independently as revealed by clinical studies. Indeed, the development of this theory mainly results from neuropsychological investigations, demonstrating a double dissociation between vision for perception and vision for action. On the one side, patients with a lesion in the posterior parietal cortex, the area terminating the visual dorsal stream, suffer from ‘optic ataxia,’ a deficit in the control of reaching and grasping objects. Despite this deficit, these patients are perfectly able to verbally describe the unreachable objects, either as a whole or in terms
of their attributes. In other terms, they can use vision for consciously perceiving objects in their environment, but not for controlling real-time actions directed at those objects. On the other side, patients with damage to a ventral region known as the lateral occipital complex are unable to recognize everyday objects or even simple geometric forms, a deficit labeled ‘visual form agnosia.’ Yet, such patients are strikingly efficient at grasping objects correctly (for instance by opening their hand as a function of the object size, or by rotating their hand according to the object orientation). Although such patients cannot consciously perceive the object and even its visual attributes (size, shape, or orientation), they can use the same object attributes to control their object-directed actions. One important consequence of this theory, according to Milner and Goodale, is that because the processes performed by the dorsal stream are very fast and largely automatic, they are largely unconscious, while, by contrast, those conveyed by the ventral stream are assumed to constitute the core of conscious perception. The visual phenomenology generated in ventral regions will in turn be transferred to working memory components in order to use information off-line, when the objects are not stimulating the visual system anymore. According to this theory, although we are typically aware of the actions we perform, no phenomenology is associated with the visual information used by the dorsal stream to control those actions. Hence, the neural computations performed in the dorsal stream remain quite inaccessible to consciousness. It is of note that when this theory was proposed more than a decade ago, evidence from binocular rivalry (in which conscious perception alternates between two images, say a face and a house presented to each eye) revealed a very strong correlation with conscious perception in the ventral stream (e.g., in the FFA for faces). Therefore, this evidence misled many researchers at that time to deny the possibility of unconscious perception in the ventral stream. However, Geraint Rees and colleagues found that patients with unilateral neglect, a deficit in which they fail to pay attention and then report stimuli on half of their visual field, still exhibit FFA activity for faces in the neglected field. Since neglected stimuli are perfectly reported when cued carefully, neglect is considered as an inability to
Neurobiological Theories of Consciousness
report efficiently because of a lack of attention. Given the possibility that consciousness and attention might be distinct (see below), it still remains unclear whether. Rees’s finding reflected ventral neural activity without consciousness or just without attention. Yet, more recent evidence with visual masking, obtained in my laboratory has revealed unconscious neural activity in ventral regions, including the FFA, during subliminal face perception, thus clearly demonstrating that the ventral stream is not exclusively related to conscious perceptual processes. This type of evidence is problematic for the Duplex Vision theory, since this theory predicts that conscious perception should be proportional to neural activity in the ventral stream. Although the possibility of unconscious ventral processing was not taken into account in the original theory, it can be accommodated by assuming a threshold mechanism, as proposed by Zeki for the Minimal Consciousness theory reviewed below. However, including this threshold leads the theory to lose it former appeal, since consciousness is ‘only partially’ correlated with activity in the ventral stream. Conversely, various recent evidences have shown that the dorsal stream can, under some circumstances, be associated with the consciousness of actions. Thus, although this theory might be effective for distinguishing the neural mechanisms that are responsible for vision for perception and vision for action, this dissociation might turn out to be orthogonal to the dissociation between conscious and unconscious processing. The Local Recurrence Theory The Local Recurrence theory put forward by Viktor Lamme is mostly concerned with vision for perception rather than action. It distinguishes between three hierarchical types of neural processes related to consciousness. The first stage involves a ‘feedforward sweep’ during which the information is fed forward from striate visual regions (i.e., V1) toward extrastriate areas as well as parietal and temporal cortices, without being accompanied by any conscious experience of the visual input. Only during the second stage, involving the ‘localized recurrent processing,’ is the information fed back to the early visual cortex. It is these recurrent interactions between early
93
and higher visual areas which, according to this theory, lead to visual experience (i.e., phenomenal consciousness, see below). The third and final stage consists of ‘widespread recurrent processing,’ which involves global interactions (similar to the global workspace model) and extends toward the executive (i.e., the frontoparietal network) and language areas. This final step also involves the attentional, executive, and linguistic processes that are necessary for conscious access and reportability of the stimulus. An interesting aspect of the Local Recurrence theory is that it offers an explanation for the difference between conscious and unconscious perception in mechanistic rather than in architectural terms. Here, the distinction between subliminal and conscious perception involves the same regions. Subliminal perception reflects the fact that the visual signal is fed forward to higher visual areas without being fed back to early areas (for instance a second stimulus replaces the first one in V1 and then prevents the setting up of recurrence between higher regions and V1). Another interesting aspect of this theory, although provocative and speculative, is that consciousness should not be defined by behavioral indexes such as the subject’s introspective reports. Instead, according to Lamme, one should rely on neural indexes of consciousness, one of which is neural recurrence. Indeed, Lamme is concerned with the question of defining phenomenological consciousness when a report is impossible. According to Lamme’s theory, involvement of the second step (local recurrence) without involving the third step (global recurrence) represents exactly this situation. We will come back to this aspect when discussing the neurobiological standpoints on the hard problem. In sum, the theory of Local Recurrence explicitly stipulates that recurrent activity is sufficient for consciousness. Yet, one main difficulty with this theory is that is fails to take into account the recurrent connections that exist between regions that are not associated with consciousness (for instance between V1 and the thalamus). It remains possible that consciousness involves local recurrence ‘between some specific cerebral components.’ However, local recurrence cannot then be considered as a sufficient condition for consciousness anymore, since it requires the involvement of
94
Neurobiological Theories of Consciousness
additional mechanisms for explaining why it only applies to a restricted set of brain regions. The Microconsciousness Theory The microconsciousness theory put forward by Semir Zeki and colleagues is arguably the most localist account of the cerebral architecture underlying consciousness. It is assumed in this theory that instead of a single consciousness, there are multiple consciousnesses that are distributed in time and space. This theory has also been mainly developed in the context of vision research and reflects the large functional specialization of the visual cortex. For instance, evidence from various sources (clinical, brain imaging, etc.) have shown that while the perception of colors is associated with neural activity in area V4, motion perception reflects neural activity in MT/V5. In particular, neuropsychological investigations have demonstrated that the respective cerebral lesions in these two functional sites lead to dissociated forms of conscious perception. Lesions in V4 lead to ‘achromatopsia,’ the inability to see the world in colors, while motion remains intact, and conversely lesions in MT/V5 result in visual ‘akinetopsia,’ the inability to perceive visual motion, while color perception remains unaffected. Furthermore, because of the existence of direct connections between subcortical structures and MT/V5 (i.e., without having to be mediated by V1), patients with a lesion in V1 are unable to perceive objects, while they can still experience motion when these objects are moving. Zeki takes these findings as evidence that consciousness is not a unitary and singular phenomenon, but rather that it involves multiple consciousnesses that are distributed across processing sites (also called essential nodes in this theory), which are independent from each other. Another form of evidence in favor of this theory is that the conscious perception of different attributes is not synchronous and can respect a temporal hierarchy. For instance, psychophysical measures have shown that color is perceived a few tens of milliseconds before motion, reflecting, according to Zeki, the fact that neural activity during perception reaches V4 before reaching MT/V5. This observation is congruent with the microconsciousness theory,
where it is postulated that microconsciousnesses are not only distributed in space, but also in time. A critical characteristic of the microconsciousness theory is that it considers these processing sites to be equivalent to perceptual sites, which means that conscious perception of one specific attribute (e.g., color) is proportional to the strength of activity in its respective processing site (i.e., V4). According to this theory, a microconsciousness associated with a processing site does not necessitate top-down influences from higher (i.e., frontal) areas although these regions might play a role in visual consciousness. Note that although the correlation between neural activity in a processing site and conscious perception is predicted to be highly positive, it cannot be perfect in this theory. Indeed, in order to take into account some clinical and experimental evidences revealing unconscious processing, notably those obtained by Zeki himself, neural activity in a processing site must reach a certain height for a conscious correlate to be generated. Therefore, this theory does not deny the existence of unconscious processing. In fact, a consequence of this postulate is that it obviates the need to separate brain regions that are linked to consciousness from those that are linked to unconscious processing, as we saw above if the Local Recurrence theory. The transition between unconscious and conscious perception reflects the crossing of a threshold ‘within’ the processing site, though the neural and behavioral characteristics of this transition remain to be specified. Note that this theory does not deny that the attributes of each processing site are bound together at one point. However, it assumes that binding is a post-conscious process occurring ‘after’ consciousness of the specific attributes to be bound together has taken place. This second step is termed ‘macroconsciousness.’ Although microconsciousness involves only one perceptual attribute (e.g., color), macroconsciousness reflects the phenomenal experience associated with the bound object (i.e., including its form, color, motion, etc.). It occurs higher and later in the hierarchy, and depends upon the presence of the previous one. In addition, Zeki proposes that there is also a third and last level coined as ‘unified consciousness,’ reflecting a more global form of consciousness,
Neurobiological Theories of Consciousness
which involves linguistic and communication skills. This third level remains largely unspecified, but it is roughly equivalent to the brain regions leading to consciousness in the global workspace model. One main difficulty with this theory is that any processing region in the brain should, at first glance, constitute an NCC in the multipleconsciousness theory. As such, it remains unclear why conscious perception is not associated with activity in most brain regions, including the cerebellum and subcortical regions, especially those conveying visual signals (e.g., the Lateral Geniculate Nucleus). Zeki takes this problem into account, at least by admitting it, and tries to solve it by explaining that neural activity in a processing site probably also necessitates the involvement of additional systems, in particular the reticular activating system maintaining arousal. Yet, the theory loses its force, since neural activity, in certain perceptual sites, is not a sufficient condition for microconsciousness anymore (not mentioning the fact that the reticular system must be involved during unconscious processing, since it also influences the subcortical regions). In addition, another difficulty for the Multiple Consciousness theory is that visual regions can lead to the binding of several attributes in the absence of consciousness, as shown both by Dehaene and colleagues using visual masking, and by Wojeiulik and Kanwisher in a patient with a bilateral parietal damage suggesting that the binding mechanisms that are supposed to lead to macroconsciousness can in fact operate in the absence of consciousness. Therefore, empirical evidence contradicts the claim that binding has to be a post-conscious process. Finally, and maybe more crucially, the inclusion of a threshold mechanism between unconscious and conscious processing, and hence the possibility of unconscious processing, is far from being obvious in this framework. This constitutes, as we saw above for the Duplex Vision theory, an empirical constraint (i.e., there has to be a threshold even if its nature remains unspecified) that implies a radical change in the main message conveyed by these two theories (i.e., consciousness is ‘only partially’ correlated with activity in the ventral/perceptual sites). In particular, it then becomes difficult to understand the frequent claim by Zeki that processing sites in the visual brain are also perceptual sites.
95
Neurobiological Standpoints on the Hard Problem Now that we have seen the main characteristics of these competitive theories individually, the next sections will describe how they face (or deny) conceptual issues related to the hard problem of explaining conscious experience. Relation to Access and Phenomenal Consciousness The neuroscientific and philosophical issues of consciousness have never been so closely related. A consequence of this close interplay is that conceptual issues raised by philosophers are progressively influencing neurobiological theories. Arguably, the most influential issue in recent years has been the potential distinction between phenomenal and access consciousness proposed by Ned Block. In short (and in the context of visual perception), this dissociative approach assumes that the NCC for phenomenal consciousness reflects the qualitative subjective experience (i.e., the qualia) associated with the percept. According to Block, the phenomenological richness that one can experience when seeing a complex visual scene goes far beyond what the observer can report. Hence, conscious access is assumed to reflect another NCC that is more linked to reportability, stimulus discrimination, introspection, etc. This distinction has been appealing to many neuroscientists in recent years, and several research programs are currently investigating the possibility to dissociate different NCCs for these two forms of consciousness. An important contrastive feature regarding the five theories reviewed above is whether they accept or reject this dissociative approach. It turns out that all of them, except the Global Neuronal Workspace theory, have recently acknowledged or even incorporated this distinction. In the Reentrant Dynamic Core theory, the distinction between phenomenal and access consciousness is analogous to the one between primary consciousness and higher-order consciousness, respectively. In a recent extension of the Duplex Vision theory, neural activity in the ventral stream has been directly associated with phenomenal consciousness by Goodale. Similarly, Zeki has recently
96
Neurobiological Theories of Consciousness
linked micro- and macroconsciousness with the phenomenal consciousness of specific attributes (colors, contrasts, etc.) and bound objects, respectively, while unified consciousness is analogous to access consciousness. In the Local Recurrence theory, phenomenal consciousness has been a very central concept at the origin of the construction of this framework. It is assumed to be a by-product of local recurrent loops, while access consciousness involves additional widespread loops between posterior and anterior regions of the brain. Importantly, all these theories assume that the NCC for access consciousness involves more or less the network of brain regions constituting the workspace in Dehaene’s theory. Therefore, the Global Neuronal Workspace theory is often considered as a restrictive theory of access consciousness, and has been criticized for confounding the subjective experience of consciousness with the ‘subsequent’ executive processes that are used to access its content. Dehaene and his colleagues not only reject this dissociation in terms of two NCCs, but they also put doubts on its psychological reality. In particular, they deny the possibility of phenomenal consciousness without access and consider this dissociation to be flawed for at least two important reasons. First, following Larry Weiskrantz, they assume that reportability is the only reliable index of consciousness and thus stimuli that cannot be reported/accessed are by definition not conscious in any way. Second, following Kevin O’Regan and Alva Noe¨, they consider that apparent cases of a rich phenomenal experience without access actually reflect the so-called illusion of seeing. For instance, in ‘change blindness’ situations, observers are usually overconfident about their capacity of seeing an entire visual scene, while actually, they fail to notice when important modifications occur at unattended locations. As such, the illusion of phenomenal richness reflects the fact that observers think that they can see more than they actually do. Because of this overconfidence, it remains unclear how one can empirically probe observers to describe the content of unattended or inaccessible perceptual events. As one can see here, this conceptual issue of phenomenal consciousness without access is closely linked to the empirical issue of consciousness without attention. This latter issue is addressed in the next section.
Relation to the Distinction between Attention and Consciousness Although attention and consciousness have long been considered to be similar, if not identical phenomena, it is largely acknowledged today that they can be dissociated. Yet, a double dissociation remains to be demonstrated. Indeed, since the seminal study by Naccache, Blandin, and Dehaene showed that top-down attention modulates subliminal priming, there is now a general consensus regarding the fact that attention is independent of consciousness and that subjects can attend to objects even if they do not consciously perceive them. However, the reverse dissociation where consciousness is itself independent of attention remains highly debated. More precisely, the possibility that subjects can be conscious of objects without any attention is at the center of a very intense controversy. Neurobiologists acknowledging Block’s dissociative approach actually assume that top-down attention is driven by workspace regions and correspond to a central mechanism, which, along with language and working memory, define access consciousness. For instance, in the case of visual perception, the signal in the posterior visual regions (i.e., in the occipitotemporal cortex) is assumed to be attentionally amplified in a top-down fashion by the same widespread parietofrontal areas as the workspace regions in Dehaene’ theory. Thus, similarly to the possibility of phenomenal consciousness without access, all the neurobiological theories seen above, except that of the Global Neuronal Workspace, acknowledge the possibility of consciousness without attention. Some a priori support for the dissociative approach can be found in studies showing that visible (i.e., supraliminal) though unattended stimuli do not involve workspace regions, but rather an increase of activity in posterior regions, as I have recently evidenced myself. Yet, as we will see below, it remains extremely difficult to demonstrate that there is any form of consciousness for visible but unattended stimuli. Here also, Dehaene and colleagues disagree with this dissociation and consider, on the contrary, that attention embodies conscious processing. According to them, only attended stimuli can be reported and are thus consciously
Neurobiological Theories of Consciousness
processed: demonstrating that an observer is conscious of an unattended stimulus, without relying on any sort of report by the subject, seems extremely difficult or even impossible. In other words, it appears that in order to assess consciousness of the stimulus, one necessarily needs to direct the observer’s attention on the stimulus, leading to the conclusion that consciousness without attention is an illogical possibility! It is of note that a few attempts, notably by Koch and colleagues, have been made to demonstrate consciousness without attention. In particular, they have relied on situations termed ‘near-absence of attention’ and where a stimulus is presented in the periphery while the subject is performing the task on a central target. Under these conditions, Koch and colleagues have found that subjects can still consciously perceive the peripheral stimulus, at least indistinctively. Yet, unfortunately, it remains impossible to demonstrate that there has not been any residual attentional amplification in this situation (additionally, the mere fact that this situation is called ‘near-absence of ’ and not ‘full-absence of attention’ appears to be highly symptomatic and suggestive of the difficulty to demonstrate consciousness without attention). Importantly, the majority of neurobiologists including Dehaene and colleagues, acknowledge that the processing of supraliminal but unattended stimuli has a special status, in-between the first step of subliminal processing, and the last step of conscious access. However, they totally disagree on a crucial question which is, ‘whether there is any form of consciousness associated with this intermediate level of processing or rather whether it is just another form of nonconscious perception.’ Proponents of the Global Neuronal Workspace theory have been very explicit on this issue and consider that this intermediate level involves just another, more elaborated form of nonconscious processing. They termed it the ‘preconscious’ stage. According to them, a stimulus reaching this preconscious state becomes potentially accessible (it could quickly gain access to conscious report if it was attended), but it is not consciously accessed at the moment, mainly because attention is maintained away (e.g., by processing a concurrent stimulus as in situations of attentional blink, inattentional blindness, or change blindness).
97
It is interesting to note that since the recent inclusion of the preconscious stage in the Global Neuronal Workspace theory in 2006, it is becoming harder to distinguish it from the Local Recurrence theory on a purely structural basis. Indeed, a tripartite taxonomy is provided in both accounts. The initial stage reflects local feedforward activity at the neural level and corresponds to unconscious (i.e., subliminal) processing, while the third stage involves a global form of recurrence in the brain and reflects conscious access. Yet, the intermediate stage is also structurally analogous in these two theories, as it involves a local form of recurrence in both cases. Rather, it is on the psychological dimension that these two theories diverge, Dehaene considering this intermediate level to reflect a nonconscious stage, while Lamme takes it as reflecting phenomenal consciousness before access. This shows us how much neurobiological theories are still dependant on psychological (i.e., subjective) indexes of consciousness. In sum, apart from proponents of the Global Neuronal Workspace theory, there is an increasing tendency in most neurobiological accounts to consider that there is an intermediate level associated with phenomenal consciousness without access. Consequently, new indexes that do not rely on subjective reports and that mainly reflect neural mechanisms are being proposed. For instance, Lamme proposes that consciousness should be indexed by the observation of local recurrent loops. In Edelman and Tononi’s theory, consciousness is indexed in terms of neural complexity reflected by differentiated integration. Before addressing these approaches in the next section, it is necessary to remind the reader that the dissociation between consciousness with and without attention, similarly to the distinction between access and phenomenal consciousness, remains highly difficult to tackle experimentally. This difficulty is mainly due to the fact that subjective reports, which necessarily involve access/attention components, remain so far the best index of whether someone is conscious. Unless a more reliable index of consciousness is found, these dissociations might possibly turn out to be immune to scientific investigation. Here also, maybe we should wait and see; or maybe new and less constrained epistemological directions should be explored.
98
Neurobiological Theories of Consciousness
Relation to Neurocognitive Panpsychism As we have seen in the previous section, there is an increasing tendency to consider that subjective reports cannot be trusted when indexing consciousness. Among the theories we have seen above, proponents of two theories in particular, that of Local Recurrence and that of the Reentrant Dynamic Core (and in particular the resulting Information Integration theory put forward by Tononi), have proposed alternative indexes that are supposedly more reliable than subjective reports. Both theories consider consciousness as an emergent property of any system sharing specific core mechanisms of either recurrence or differentiated integration, respectively. Therefore the most reliable index of consciousness should, according to these accounts, reflect the quantification of these core mechanisms. In particular, these neural indexes are assumed to offer a more reliable estimation of phenomenal experience, well above the poor estimation provided by subjective reports. This approach can be appealing since it could potentially lead to a quantification of consciousness in nonhuman species, in preverbal infants, and in artificial intelligence systems. Yet, it leads to a form of panpsychism (see Glossary) that one can term as ‘neurocognitive panpsychism.’ Both the Local Recurrence theory and the Reentrant Dynamic Core theory are leading to a similar, though more restrictive panpsychist views: any system in the world can be conscious as long as it shares information following the core mechanisms that are specific to these theories. They constitute neurocognitive versions of panpsychism, because these theories are primarily aimed at offering neurobiological accounts of consciousness, while they can also be extended to any cognitive system in the world that shares the same mechanical properties for information-processing. As such, both theories follow the principle that consciousness arises when information is transmitted among neurons interacting along the core mechanisms of these theories. Yet, there are several conceptual difficulties with these theoretical approaches (the term conceptual difficulties and not impossibilities is preferred here because none of these issues is impossible to resolve per se; however, they might turn out to be extremely difficult to demonstrate as well). A first conceptual
difficulty is that this doctrine leads to the conclusion that any system displaying these specific core mechanisms will experience consciousness, whatever implementation supports it and, more importantly, whatever its size. This fact relates to Chalmers’ argument that even a system as small as a thermostat, because it recurrently shares information between a temperature sensor and a controller for switching the heater on or off, will experience consciousness to some degree. Indeed, the theory of Local Recurrence leads to the fact that any two neurons in interaction would be sufficient for consciousness. The Reentrant Dynamic Core theory would require a few more neurons, but not that many to generate consciousness (precisely nine neurons would be sufficient according to Herzog, Esfeld, and Gerstner). Although such possibilities can hardly be totally rejected, it remains as unclear how they could ever be demonstrated. A second main issue lies in the paradox of trying to prove that neural indexes are more respectable because they supposedly probe phenomenal and not access consciousness. Indeed, although it is clear that neural indexes offer interesting possibilities when report is impossible (for instance in cases of locked-in syndromes or for prelinguistic babies), they still cannot be taken as reflecting more than conscious access. This principle follows from the fact that neural indexes of any sort have to be validated by confronting them at one point with some kind of report, hence with access and not phenomenal consciousness. For instance, demonstrating that recurrent processing is sufficient for consciousness initially requires consciousness to be indexed by probing whether the system is indeed conscious, and the best way to do so is still to rely on subjective reports in humans. Then, a neural index can only be validated when a correlation has been established with subjective reports. Assuming that one can demonstrate that recurrence, even among two neurons, correlates with some degree of consciousness, this can be achieved only through report and that degree of consciousness will only reflect a partial form of access, not phenomenal consciousness. Once again, the interesting aspects of this approach cannot be neglected, since it can allow to subsequently probe whether the core mechanisms represented by this
Neurobiological Theories of Consciousness
index can be found in systems or species that cannot report, suggesting that they ‘might probably’ be conscious. However, neural indexes cannot be taken as reflecting a superior or richer phenomenal form of consciousness since they can only be validated through reports. In other terms, these theories paradoxically emphasize that their core mechanisms do not reflect access consciousness, while actually only this form of consciousness can be used to demonstrate their validity. A third conceptual issue is that even if it turns out that one of these neural indexes turns out to be perfectly correlated with consciousness, and thus becomes a perfectly reliable measure of consciousness, then one might still ask whether we have made any progress. This issue is less problematic and might turn out to be resolved in the future. However, it is worth mentioning, since it remains unclear whether we are searching in the right direction. Indeed, although neural indexes might turn out to be very useful in many circumstances, they do not escape the limitation of NCCs. They would still primarily inform us on whether someone (or something) is or is not conscious. Yet, they would not directly help us understand why they are conscious. If it turns out that recurrence or differentiated integration are sufficient for consciousness, we would still have to accept that consciousness is ‘emerging’ from the brain each time these core mechanisms come into play. This issue is more moderate because once we have found a perfect neural index of consciousness, it might get us closer to the explanatory gap. Yet, it remains unclear how it will allow us to cross that gap.
Conclusion By the end of this article, one might wonder where we stand with regard to the original issue of how the brain leads to consciousness. Over the last two decades, dozens of theories of the NCC have been proposed (leading David Chalmers to ironically speak about an ‘NCC zoo’) and a kind of trialand-error strategy has been applied to evaluate them, and in most cases, to reject them one by one. Here we have focused more specifically on five recent neurobiological frameworks that are
99
the most popular accounts in this field of research. We have seen that, actually, consciousness can hardly be restricted to neural activity in the ventral stream, nor is it simply reflecting recurrent activity or minimal activity in processing sites. It remains unclear whether consciousness reflects or does not reflect differentiated integration within a dynamic core, as emphasized by Edelman and Tononi’s speculative, but powerful theory, or whether it simply reflects activity in workspace regions, as outlined by Dehaene’s well-specified, but apparently restrictive account of conscious experience. One main difficulty in disentangling this issue and contrasting these theories, is that it remains unknown whether neurobiological frameworks should be restricted to conscious access and subjective reports, or whether they should extend to other (i.e., inaccessible, unattended, phenomenal) forms of consciousness. Therefore, future research will unavoidably be bound to decide whether neurobiological accounts should take into account or reject the hard problem of understanding a subjective form of conscious experience that cannot be simply defined as conscious access. Thus, facing the hard problem will probably not be any more a main issue on its own. Instead, it will be the problem of deciding whether there is a hard problem at all, leading to a ‘super-hard problem’ of consciousness and obviously complicating the whole issue. Whether we should stay focused on variants of the NCC until the solution pops out and reduces the hard problem to an easy one, or whether we actually need new laws (of physics, life, or anything else) to cross the explanatory gap, only time will tell. Yet, although none of the neurobiological theories provided so far have been able to provide a ‘conclusive’ explanation regarding this matter, the search for a neurobiological explanation of consciousness still constitutes one of the most exciting challenges of contemporary science. See also: The Neurochemistry of Consciousness.
Suggested Readings Baars BJ (1988) A Cognitive Theory of Consciousness. New York: Cambridge University Press. Block N (2005) Two neural correlates of consciousness. Trends in Cognitive Science 9: 46–52.
100
Neurobiological Theories of Consciousness
Chalmers D (1996) The Conscious Mind. New York: Oxford University Press. Crick F and Koch C (1990) Towards a neurobiological theory of consciousness. Seminars in Neuroscience 2: 263–275. Dehaene S, Changeux JP, Naccache L, Sackur J, and Sergent C (2006) Conscious, preconscious, and subliminal processing: A testable taxonomy. Trends in Cognitive Sciences 10: 204–211. Dehaene S and Naccache L (2001) Towards a cognitive neuroscience of consciousness: Basic evidence and a workspace framework. Cognition 79: 1–37. Edelman GM and Tononi G (2000) A Universe of Consciousness: How Matter Becomes Imagination. New York, NY: Basic Books. Gallese V (2007) The ‘‘conscious’’ dorsal stream: Embodied simulation and its role in space and action conscious awareness. Psyche 13(1): 1–20. Goodale M (2007) Duplex vision: Separate cortical pathways for conscious perception and the control of action. In: Velmans M and Schneider S (eds.)
The Blackwell Companion to Consciousness, pp. 616–627. Oxford: Blackwell. Koch C (2004) The Quest for Consciousness: A Neurobiological Approach. Denver, CO: Roberts. Koch C and Tsuchiya N (2007) Attention and consciousness: Two distinct brain processes. Trends in Cognitive Sciences 11: 16–22. Kouider S and Dehaene S (2007) Levels of processing during non-conscious perception: A critical review. Philosophical Transactions of the Royal Society of London B 362: 857–875. Lamme VA (2006) Towards a true neural stance on consciousness. Trends in Cognitive Sciences 10: 494–501. Tononi G (2004) An information integration theory of consciousness. BMC Neuroscience 5: 42. Zeki S (2007) A theory of micro-consciousness. In: Velmans M and Schneider S (eds.) The Blackwell Companion to Consciousness, pp. 580–588. Oxford: Blackwell.
Biographical Sketch
Sid Kouider is a cognitive neuroscientist working at the Ecole Normale Supe´rieure (Paris, France) on the neurobiological and psychological foundations of consciousness. His work focuses on contrasting conscious and unconscious processes, both at the psychological and neural level, using various behavioral and brain imaging methods. Recently, he extended this line of research to study the neural correlates of consciousness in prelinguistic babies.
The Neurochemistry of Consciousness J Smythies, University of California, San Diego, CA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary AMPA – Alpha-amino-3-hydroxy-5-methyl-4isoxazolepropionic acid CaMK11 – Alpha Ca2+-calmodulindependent kinase 11 EPSP – Excitatory post-synaptic potential ERK – Extracellular signal-regulated kinase GABA – Gamma-amino butyric acid GHB – Gamma-hydroxy butyric acid GIRK – G-protein mediated inwardly rectifying K+ JaK – Janus kinase LDTN – Lateral dorsal tegmental nucleus LSD – Lysergic acid diethylamide MAP – Microtubule-associated protein NMDA – N-methyl D-aspartate Oct-6 – Oct-6 is a transcription factor that belongs to the POU transcription factor family (class-3 subfamily) PPN – Pedunculopontine nucleus PSD – Postsynaptic density S-100 – A brain protein SAP – Synapse-associated kinase
Introduction The term ‘consciousness’ is commonly used in three different ways. One is the clinical sense in which we distinguish a person in coma from a normally alert being. The second sense (as used e.g., by Crick, 1998) is ‘awareness,’ or ‘alertness’ that is closely allied to ‘attention.’ The third is the phenomenal sense, in which we refer to the totality of a person’s sensations, images, emotions, and thoughts, available to introspection, and studied by introspectionist psychologists, such as Ramachandran and Gregory. In the first we study neural mechanisms whose malfunction lead to coma. The second covers a broad range of studies. In the third, we study the
complex brain mechanisms that construct the structures (including the visual field and the somatic sensory field – also known as the ‘body image’) that compose a person’s phenomenal consciousness.
Clinical Consciousness Coma can be induced by a wide variety of agents and diseases. Among the most interesting, from the neurochemical point of view, are general anesthetics, and the remarkable effect induced by the combination of desferrioxamine and prochlorperazine. General Anesthetics Over a century ago Meyer and Overton proposed that anesthetic agents act by interacting with lipophilic components of cell membranes. Today it is still recognized that lipophilicity is important, but attention has now focused on specific receptors. Most anesthetics act at multiple molecular sites. The most prominent action is enhancement of GABA(A) receptor-mediated neuronal inhibition, in both cortical and subcortical regions, with selectivity for receptors that contain beta2 and especially beta3 subunits. This can be modulated by simultaneous inactivation of structures, such as the median septum, and ventral tegmental area, that normally activate the cortex, or that mediate the hippocampal output, such as the nucleus accumbens and ventral pallidum. Every general anesthetic acts on at least one type, and in some cases several types, of ligandgated ion channels. (The ‘family’ of such channels composes nicotinic, GABA(A), glycine, serotonin, and glutamate receptors). There is evidence that some anesthetic agents act partly at inhibitory glycine channels, and that some depress excitatory acetylcholine nicotinic receptors. They may target specific sites within the channel lumen (ACh), or on the outer side of the alpha helix lining that 101
102
The Neurochemistry of Consciousness
channel (GABA(A) and glycine channels). Some also inhibit MGluR5 receptors. Recently individual amino acid residues have been identified that are necessary for anesthetic function. These are located on the extracellular portions of the alpha subunits of the GABA(A) receptor. Gamma-hydroxybutyric acid (GHB) is a naturally occurring metabolite of GABA that may produce coma by action at its own receptors (GHBR) and at GABA(B) receptors. The situation is complicated by the fact that GHB can also be metabolized to GABA. The coma induced by traumatic brain injury is usually attributed to glutamate neurotoxicity, there is also evidence, however, that potentiation of GABA(A) receptor function, via activation of CaMK11, may also be involved. GABA(A) receptors may also be involved in the comas induced by inert gases such as nitrogen. In contrast, two anesthetic agents – ketamine and nitrous oxide – act primarily by blocking excitatory NMDA receptors (to induce coma), and by enhancing analgesic opioid mu receptors.
The Desferrioxamine/Prochlorperazine Coma In 1985 Blake and his coworkers reported the serendipitous discovery in human subjects that a combination of normal doses of the iron-chelating agent desferrioxamine with the dopamine-receptor blocker prochlorperazine induced a deep and longlasting coma (1–3 days with complete recovery). Neither drug by itself induced any such effect. Catecholamines, such as dopamine, are known to take part in conjunction with iron ions in vitro in the complex O’Brien cycle. In this, iron ions cycle between the ferric and the ferrous states, and the catecholamine molecules cycle with their quinones. This system acts as an effective dismuter of superoxide, converting five molecules of the highly toxic superoxide per cycle into two molecules of oxygen and three of the less toxic hydrogen peroxide. Thus, the combination of dopamine and iron in the O’Brien cycle may provide an essential mechanism in the neuron to protect against superoxide toxicity. Mitochondria convert 5% of their oxygen into superoxide. If iron is removed from the neuron by desferrioxamine, and dopamine is prevented from
entering the neuron by prochlorperazine, this may result in the failure of the O’Brien cycle allowing neurotoxic levels of superoxide to build up, leading to coma. This hypothesis needs some further explanation of the role of catecholamines like dopamine in the brain. It is usually thought that such neuromodulators exert their effects on the postsynaptic neuron only by activation of their specific receptors and subsequent postsynaptic cascades. The neurotransmitter molecule itself, in this case dopamine, is thought to stay outside the cell, to be dealt with by reuptake into the terminal, or by further metabolism. However, it is now known that, when a receptor (in this case the dopamine D1R) is activated, it is usually immediately endocytosed carrying the molecule of, in this case, dopamine into the endosome apparatus. Here the dopamine molecule and the protein receptor separate, and the latter is recycled back to the surface, or, if damaged, to the proteosome to be broken down into its constituent amino acids for reuse. The iron transporter molecule transferrin, on binding a molecule of iron, is also endocytosed into the neuron together with its cargo, and is trafficked to the same endosome to which the dopamine receptor is trafficked. Thus the ingredients of the O’Brien cycle – free iron, dopamine, and superoxide (from a neighboring mitochondrion) – can come into the required close contact, in the endosome, for the O’Brien cycle to operate. This cycle may play an important role in the normal cell. Thus catecholamines, including dopamine, may act on neurons, in part, by their direct antioxidant chemical properties, rather than by only triggering postsynaptic cascades. Catecholamines are potent direct freeradical scavengers and antioxidants. Dopamine D2 receptors also promote antioxidant activity by inducing the synthesis of the antioxidant enzyme superoxide dismutase, and the neuroprotective enzymes glutathione synthase, gamma-glutamylcysteine synthesase, glutathione peroxidase, glutathione reductase, and glutathione-S-transferase.
The Cholinergic System The cholinergic system in the brain plays a vital role in the maintenance of the clinical conscious
The Neurochemistry of Consciousness
state. Brain injury studies show that much of the brain can be disabled, or removed, without affecting consciousness. However, damage to the midbrain often results in coma. The key nuclei involved in this reaction are the cholinergic pedunculopontine nucleus (PPN) and lateral dorsal tegmental nucleus (LDTN) in the pons and medulla. These project widely to all thalamic nuclei, including the reticular nucleus, and to the limbic system, as well as to the higher cholinergic nucleus basalis of Meynert in the basal forebrain. They do not project widely to the neocortex, except for the medial prefrontal cortex. The nucleus of Meynert supplies most of the cortex with cholinergic fibers. Lesions of the nucleus basalis do not commonly lead to coma. Another brain area, lesions of which commonly lead to coma, contains the complex intralaminar nuclei of the thalamus. This sends a glutamatergic projection to the cortex. During waking this generates single action potentials at regular intervals that activate the cortex. In contrast during sleep this changes to burst firing that generates the synchronized EEG pattern typical of sleep. A brain area essential for rapid eye movement (REM) sleep is the nucleus reticularis pontis oralis. Crick, Llinas, and others have suggested that the correlates of consciousness in the brain are provided by reentrant thalamocortical projections, especially linked to the reticular nucleus. This system is powerfully modulated by the brain stem cholinergic nuclei PPN and LDTN. Crick suggested that the control of the ‘searchlight of attention’ was mediated by the reticular nucleus of the thalamus. However, this searchlight includes olfaction as well as the other four senses. We can switch our attention to an experienced odor as quickly and easily as we can to a sight or a sound. Now, there are no anatomical connections between the olfactory cortex and the thalamic reticular nucleus. Therefore the basic switch that controls conscious attention may be provided by the midbrain cholinergic nuclei acting in conjunction with the reticular nucleus. Thus the global control of consciousness may be affected by this cholinergic system. The further effects of this system on ‘awareness’ and phenomenal consciousness will be discussed later.
103
Awareness/Alertness The degree of alertness of a person, and the direction of selective attention is controlled by a most complex brain mechanism in which a number of anatomical structures and neuromodulators take part. Amongst these the reticular nucleus of the thalamus plays a key role. It sits sequentially astride all the sensory inflow tracts in the brain, except the olfactory input, and receives glutamatergic input from these. Its output consists wholly of GABAergic neurons whose axons are widely distributed sequentially to other thalamic nuclei. This nucleus also receives extensive projections from the cholinergic PP/LDTN, raphe, and locus coeruleus in the brain stem. Thus it is in a position to modulate much of the sensory input to the brain and direct selective attention. The neuromodulators associated with ‘alertness’ and ‘selective attention’ are acetylcholine, norepinephrine and dopamine, adenosine, histamine, and hypocretins. The currently available evidence suggests a role for noradrenaline in alerting/arousal, whereas acetylcholine may control arousal and selective attention. Dopamine is principally concerned with reinforcement. The role of serotonin is very complex and is dealt with here and below. Acetylcholine. Collaterals of the sensory tracts in the midbrain enter the reticular formation where they activate the cholinergic PP/LDTN. These nuclei in turn project to the sensory relay nuclei and to the reticular nucleus of the thalamus. In the relay nuclei this input activates neuronal transmission via nicotinic receptors. In the reticular nucleus this input from the PP/LDTN inhibits (via muscarinic type 2 receptors) the inhibitory GABAergic cells of the reticular nucleus, which in turn further promotes relay of sensory information to the cortex. During sleep, activity in the sensory tracts diminishes, and this cerebral activation is reversed. Muscarinic blockade by scopolamine leads to an altered state of consciousness (‘twilight sleep’). Nicotine administration leads to improved attentional performance in a number of tests (e.g., the Stroop test). It has been suggested that an important role of acetylcholine related to consciousness is to modulate the intradendritic cytoskeletal structure of cortical pyramidal cells by a mechanism that includes MAP-2.
104
The Neurochemistry of Consciousness
Norepinephrine (NE) and dopamine (DA). Both of these neuromodulators promote the alert state: dopamine promotes ‘pleasure’ as well, having a powerful role in reinforcement mechanisms. Reduction in NE activity facilitates disengagement of spatial attention, and a frontal NE system promotes focused attention by attenuating the effects of distractors. DA plays a key role in shifting attention. It has been suggested that a DA pathway mediates a ‘lower’ tonic arousal system, whereas an ‘upper’ phasic NE arousal system is involved in response selection. During slow wave sleep (SWS) the NE neurons in the locus coeruleus are inhibited by an input from GABAergic neurons in the lateral preoptic area. During REM sleep inhibition by a second population of GABAergic neurons in the periaqueductal gray and dorsal paragigantocellular nucleus leads to a complete cessation of NE cell activity. Serotonin. This neuromodulator promotes waking and inhibits sleep activity. Raphe neurons (as well as neurons that release norepinephrine and histamine) fire steadily during waking, more slowly during SWS and hardly at all in REM sleep. This latter effect is brought about by 5HT activation of long-range GABAergic projection exhibiting 5HT 2A/2C receptors that inhibit the cholinergic cells of the PP/LDTN. Histamine. Histamine increases alertness and inhibits both SWS and REM sleep. Histaminergic cell bodies are located in the tuberomammillary nucleus in the posterior hypothalamus, and project widely to cortex, basal forebrain, thalamus, and tegmentum. Histamine, acting on H1 receptors on relay cells in thalamic relay nuclei, changes their firing from burst to single tonic firing, thus promoting transmission of information. H2 receptors facilitate cortical and hippocampal activity. Hypocretins/orexins. Two polypeptides, hypocretin 1 and 2 (also known as orexin A and B), are produced by cells in the lateral and posterior hypothalamus. They stimulate wakefulness by binding to their G-protein linked receptor. Defects in their action lead to narcolepsy. Adenosine. Adenosine is thought to be a sleepinducing factor. During sleep deprivation extracellular levels of adenosine rise steadily in the basal forebrain and inhibit the basic forebrain cholinergic neurons that promote wakefulness. Caffeine,
which promotes wakefulness, acts as an adenosine receptor inhibitor.
Phenomenal Consciousness This section discusses the neurochemical factors related to the processes of constructing the elements of phenomenal consciousness – sensations, images, emotions, mood, and thoughts. The Serotonin System Psychedelic drugs, such as mescaline and lysergic acid diethylamide (LSD), produce remarkable changes in conscious perception and cognition well described by Aldous Huxley in The Doors of Perception. Their principal mode of action is as partial agonists at serotonin 2A receptors (with some involvement of 2C receptors). In rat brain, LSD greatly increases 5-HT(2A)R-mediated Foslike immunoreactivity in some limbic areas (medial prefrontal cortex, anterior cingulate, central nucleus of the amygdala, but not nucleus accumbens). Activation of 5-HT(2A)Rs by LSD also enhances nonsynchronous, late components of glutamatergic excitatory postsynaptic potentials (EPSPs) at apical dendrites in layer V cortical cells. The cell bodies of all serotonin neurons in the brain are contained in the raphe nuclei of the midbrain and pons (as well as two small adjacent areas – the ventrolateral and dorsal reticular formation). There are three raphe nuclei. The dorsal raphe nucleus projects by thin axons mainly to the cortex, the median nucleus projects by thick fibers bearing large varicosities to limited limbic and frontal cortical areas, and to the hippocampus, and the caudal nucleus which projects to the spinal cord. The raphe nuclei supply every portion of the grey matter (with the exception of the superior olivary complex) with a dense network of fine axons that bear many boutonsen-passage. Many of these are extrasynaptic, so that serotonin uses mainly volume transmission. In the case of many transmitters, transmission is mostly synaptic, so that a precise network of connections is maintained. In volume transmission the transmitter molecule is, as it were, sprayed throughout the target cortex, carrying its informational signal as widely as possible at minimal informational cost.
The Neurochemistry of Consciousness
Many raphe neurons also contain glutamate, which is released concomitantly with serotonin upon activation at the synapse. Serotonin is an important cellular chemical, even very low down in the phylogenetic chain. In plants it plays an important role in photosynthesis. In the brain it has many cellular functions including modulation of neurogenesis, apoptosis, dendritic and spine refinement, cell migration, and synaptic plasticity. It likewise modulates many neuronal functions including autonomic, circadian rhythms, appetite, aggression, sensorimotor activity, sexual behavior, learning, memory, personality, as well as perception and mood. There are seven main types of serotonin receptors. The ones of current interest in this review are 5-HT(2A)Rs that are located both presynaptically and postsynaptically in all cortical layers except layer IV. They are located mainly on pyramidal cells, particularly on their apical dendrites, on inhibitory neurons, and on axon terminals. In human cortex they are ranked quantitatively cortex > hippocampus > > thalamus, with none in the cerebellum and brainstem. The effect of 5HT receptors on neurons is very complex, depending on target cell and species. 5-HT(2A)Rs usually activate the target neuron. For example in rat, neocortical layer V pyramidal cells 5-HT(2A)Rs induce EPSPs by both pre- and postsynaptic mechanisms. In mouse, prefrontal pyramidal cells 5-HT(2A)Rs reduce rapidly inactivating Naþ currents. This does not affect somatic action potentials in the target neuron, but has the important effect of reducing the amplitude of Hebbian back-propagated dendritic potentials. 5-HT(2A)Rs depolarize neurons in human neocortical slices, and guinea-pig motoneurons, by inhibiting a resting Kþ currant conductance. However, in the rat, 5-HT(2A)Rs hyperpolarize C6 glioma cells by activating an outward Kþ current, via increased phosphoinositol hydrolysis, and increased intracellular calcium levels. Serotonin also produces EPSPs in target neurons by promoting the presynaptic release of glutamate from glutamatergic terminals via activation of phosphokinase C. Serotonin, acting on its 2A receptors, also excites neurons by inhibiting inhibitory GABA interneurons, via phosphorylation of the GABA(A) receptor,
105
consequent on activation of phosphokinase C anchored to the scaffolding protein RACK. 5-HT (2A)Rs also release dopamine at its terminals, and inhibits nitric oxide synthase activity induced by cytokines in C6 glioma cells. In contrast, 5-HT(1A)Rs usually inhibit the target neuron by opening a channel mediating a G-protein mediated inwardly rectifying Kþ (GIRK) inhibitory current, and by opening a channel mediating a slowly activated outwardly rectifying Kþ current. It also inhibits AMPA-evoked currents by down regulation of Ca2þ/calmodulin-dependant PK11 and by inhibiting cGMP responses. Serotonin (2A) receptors utilize a number of postsynaptic cascades including the phosophoinositol pathway, activation of phospolipases, activation of ERK, JaK, and the transporter processes Naþ-proton exchanger and Naþ/Kþ-ATPase. 40 Hz gamma oscillatory behavior in populations of neurons has been linked to the genesis of conscious phenomena. It is therefore of interest that 5-HT(2A) receptors can inhibit oscillations in rat inferior olive nuclei by depressing both T-type Ca2þ currents and a resting Kþ current. The conclusion that can be drawn from this data supports Aghajanian’s hypothesis that psychedelic drugs act by promoting glutamatergic neurotransmission in selected areas of the brain. However, we need to be more specific than that. In the first place, why should increased glutamatergic activity in the visual brain induced by serotonin acting at its 2A receptors result in the marvelous hallucinations witnessed by the subject, who has taken the drug, when agents that act directly at glutamatergic receptors do no such thing? Furthermore, why should the hallucinations be so beautiful? What does it tell us about the neurophysiological basis of esthetics? The great British neurologist Macdonald Critchley summed it up as follows: The usual emotional content of the hallucinosis is best described as one of amazement, awe, interest, and delight. The character of the visions is such as to impress the most prosaic and unimaginative scientific observer in a manner of which no natural beauty or grandeur is capable, almost all writers have insisted that the most skilful pen or brush could not do justice to the marvel of the hallucinations.
106
The Neurochemistry of Consciousness
It seems most remarkable that a person, who shows in normal life little or no artistic ability, nevertheless has a brain that can produce all this amazing abundance of high art simply because its 5-HT(2A) receptors are stimulated. Psychedelic drugs simply modulate the firing of cerebral neurons by action on mundane receptors, ion channels, and postsynaptic cascades, just as all the other nonpsychedelic neurotransmitters and neuromodulators do. Aldous Huxley was so struck by this paradox that he invoked Henri Bergson’s suggestion that one function of the brain might be to limit our contact with reality, acting as some sort of filter keeping certain aspects of reality out of consciousness. Huxley suggested that our inability to see these phenomena under normal conditions is because this part of the mind is under continual inhibition by certain brain mechanisms, and, when the latter in turn are inhibited by drugs such as mescaline and LSD, then these ‘antipodes of the mind’ become visible. Jung (personal communication) has suggested that the origin of the visions would lie in the collective unconscious, or mundus archetypus. An alternative hypothesis is that psychedelic drugs modulate the nonlinear dynamics of holographic brain mechanisms in some as yet undetermined manner. Visual beauty depends on those subtle properties of color, shape, balance, proportion, and harmony explored by great artists. The very simple line drawings of Matisse, for example, evoke the utmost pleasure in the viewer. A very simple deviation destroys this effect. These particular forms drawn by Matisse evoke in the visual cortex certain patterns of neuronal excitation, that must lead in turn to a discharge in those efferent cortical neurons that project to the pleasure centers of the brain, such as the septal nuclei and the ventral tegmentum. In contrast, geometrically very similar patterns drawn by Tom, Dick, or Harry fail to stimulate these efferent neurons. How this is done, and what evolutionary purpose this reaction serves, remain problems in search of explanations. Similar considerations apply to how neurocomputational mechanisms in the auditory cortex react to Mozart’s music by stimulating comparable efferent neurons to limbic pleasure centers. The acetylcholine system is also involved in type 3 consciousness as muscarinic agonists can
induce delirium, which is a global disorder of consciousness. Opiate receptors are also involved in type 3 consciousness as their stimulation induces profound changes in perception and mood. Salvinorin A stimulates only kappa opiate receptors and is one of the most potent psychedelic drugs known. Acetylcholine, Virtual Reality, and Television Technology Recently a number of experimental observations have established the important facts, that the visual brain uses virtual reality in many of its functions, and that the neuromodulator acetylcholine plays a major role in this process. For years it was thought that the visual brain works as follows. Photons bounce off physical objects in the stimulus field to land on the retina, there to form a topographic map of the external world. The retina functions as a digital camera that translates this spatial pattern of stimulation of its rods and cones, into a spatiotemporal pattern of excitation of the optic nerve and optic radiation. These patterns are fed to the primary visual cortex where they are processed by a series of parallel neurocomputational mechanisms. The output of these mechanisms is then fed to three separate and parallel higher visual mechanisms that perform neurocomputations relating to shape, color, and movement. This process involves both topographic and vectorial coding. Finally, by a process not at present understood, the final result of these computations are functionally reunited to produce the unified percept that appears in the visual field in consciousness, in which the color is inside the shape and both move together. The fact that there are three separable processes involved is supported by observations of how vision returns after lesions of the occipital lobe. A patient, whose vision is returning after such an injury, at first sees only pure movement, usually rotary, without any objects or color. Then light appears as a pure Ganzfeld – a featureless field of white light. Then colors return, first as space or film colors free floating in visual space. This is followed by the appearance of objects, usually in fragments. Finally these fragments join up to form complete objects into which the film colors enter. But, in all this explanation, it was held that the final picture produced in perception in consciousness represented,
The Neurochemistry of Consciousness
as faithfully as it could, what was actually ‘out there’ at any one moment. It is now known that this assumption is wrong. We actually see, not what is ‘out there’ but what the brain computes is most probably ‘out there’. The first evidence for this was provided by Ramachandran, who showed that the brain fills in scotomata. A small lesion in the visual cortex will result in a small equivalent blind spot in vision. A natural form of this is the blind spot that we all have over the site where the optic nerve leaves the retina. But we do not experience any such blind spot as such. If we locate the blind spot over a plaid pattern we will see a complete plaid pattern, because the brain ‘fills in’ the gap from its memory banks. Since then many experiments have been carried out that demonstrate this phenomenon. In one particularly illuminating experiment subjects were shown two different pictures one to each eye, so that retinal rivalry occurred. The first presentation was of two photographs, one to each eye, the first of a group of monkeys and the other was filled with leaves in a tropical jungle. They saw what was ‘out there,’ that is, monkeys alternating with leaves in retinal rivalry. The experimenters cut up both photographs into irregular bits and then rearranged these to create two pastiches A and B, so that wherever A showed monkeys, B showed leaves, and vice versa. Now, when the subjects looked at A and B in the same way, they did not see the actual two pastiches A and B in retinal rivalry (which was the actual stimulus), but they saw a complete picture of monkeys alternating with a complete picture of leaves. In other words the brain had rearranged the stimuli, and had suppressed the most unlikely pastiches in favor of what it was more familiar with, and hence was more probable – a monkey picture and a leaf picture – which it created with assistance from its memory banks. A similar process occurs during a saccade – the REM from one visual fixation point to another. Experiments have shown that, during a saccade, we are technically blind. The input from the retina to the cortex is suppressed and the visual field during the transition is filled in from the brain’s memory banks. The purpose of this is to avoid the intense vertigo that would otherwise result. The reader can demonstrate this him or herself. Look into a mirror and direct your attention to your left eye. Then do a saccadic movement to look at your right eye. You
107
will notice that you cannot observe any movement of your eyes. Yet, if you stand next to someone else, and look at their eyes in the mirror whilst they are repeating this maneuver, you will notice that you can easily see the movement of their eyes. This use of virtual reality by the visual system can be explained by reference to television technology, and, in particular, to information compression. Television engineers have found that transmitting all the information picked up by the TV cameras to form the final picture on the TV screen is needlessly expensive and wasteful. This is because, in most TV scenes, only a few targets change much and much of the background does not. The latter can be ‘filled in’ much more cheaply by probability–estimates-based virtual reality, provided by the memory banks of the system. So the art of the TV engineer is to create the optimum mixture of actual events in the studio (expensive but accurate) with fill-ins from memory banks (cheap but progressively more inaccurate). The brain, astonishingly, does the same. The cholinergic input to the cortex comes preponderantly from the nucleus basalis of Meynert. This has two target areas in the cortex. The first (A) is to layer IV, where it up-regulates afferent (visual) input by means of excitatory nicotinic receptors on layer IV pyramidal cells. The second (B) is to superficial layers, where it down-regulates corticocortical transmission via muscarinic receptors. In the resting state, system B, that carries information from memory banks, is dominant, so that the visual system is tilted towards virtual reality (computationally parsimonious) presentation. Then, any unexpected, or potentially important, new stimulus causes activation of cholinergic neurons in the nucleus basalis. The acetylcholine thus released in the cortex activates the retinal input A (carrying computationally expensive ‘reality’) and inhibits the corticocortical input B (carrying virtual reality), so that the new potentially important stimulus can receive full attention. This convincing evidence that the brain uses television technology has an important impact on philosophical theories of consciousness and perception. ‘Common sense,’ as well as many philosophers who do not pay much attention to neuroscience, still believe that the theory of perception known as naı¨ve or direct realism is true. In the case of vision this
108
The Neurochemistry of Consciousness
theory holds that phenomenal objects – that which we experience – are literally physical objects, or at least the surfaces of physical objects. Sensations, they say, are just the way that such external objects appear to us. The central nervous system is supposed to mediate this process in some unexplained way. In other words vision works something like a telescope, and not at all like television. The rival scientific theory of physiological realism holds, in contrast, that phenomenal objects (our sensations) are literally constructs of the nervous system. In between the retina and the visual field in consciousness lies bank upon bank of neurocomputational mechanisms, that use, we now know, aspects of television technology. In other words, vision works like television, and not at all like a telescope. The Adrenergic System in the Medulla and the Neurobiology of Bipolar Disorder The relevance of studies of ‘abnormal’ mental states (i.e., those found in certain diseases) to attempts to discover the normal function of the brain is based on the fact that this technique belongs to a tried and tested tradition in clinical neurology and neuropsychiatry. A great deal of our understanding of how the normal brain works has come from such clinical studies. The names of Broca, Wernicke, Head and Holmes, Penfield, Ramachandran, and many others come to mind. In this article I will focus particularly on bipolar disorder and schizophrenia. One important feature of consciousness is the way that, although composed of many different functional components (perception, ideation, emotions, cognition, action), it nevertheless seems often to act as a unity. This is particularly true with regard to the phenomenon of mood. Emotions are transient episodes that usually are short-lived. Moods, in contrast, last longer and color all aspects of mental life. The basic neurobiological function of mood seems to be linked to the fundamental behavioral instincts of approach behavior and retreat behavior. When we are depressed the world looks a gloomy place. We dwell on its bleak sadness. Trivial worries grow to feelings of despair. Our minds become filled with gloomy thoughts. Any action becomes a burden. We tend to retreat from the world. Everything is repulsive. In contrast, when we are in a good mood, the world
appears bright and cheerful. If young, our thoughts lightly turn to thoughts of love. We view the growing mortgage with indifference. Everything appears to be for the best in this, the best of all possible worlds. Everything appears attractive. Furthermore, abnormalities in this system seems clearly responsible for the clinical condition of manic-depressive psychosis (bipolar disorder), in which all the phenomena described above for mood are taken to extremes. In mania all sensations are intensified. Colors are more intense, and everything appears more bright and beautiful with an intense inner meaning (salience). The sense of beauty is vastly increased. Natural scenes become transcendentally beautiful beyond belief. The patient is also flooded with religious feelings and ideas. Everything becomes attractive. These symptoms strongly recall the effect of psychedelic drugs such as LSD. Other symptoms, such as extreme ecstasy and mental clarity, resemble the effects of the dopamine-releaser cocaine. Thus mania resembles the effect of simultaneous over activity in the brain dopamine system and at serotonin 2A receptors. In major depression all these symptoms are reversed to their opposites. In place of ecstasy there is unutterable misery, and unbearable physical malaise. Patients are consumed by a sense of complete isolation from God, from all fellow humans and from the entire world. Everything is guilt, fog, and darkness. The dominant emotion is fear – overpowering, paralyzing fear, and despair. Everything becomes repulsive. So what could be the neurobiological basis of this close synchronization of mood with sensation, emotion, ideation, cognition, and behavior? Mood and emotion are regulated by extremely complex brain (and visceral) mechanisms in which all the neuromodulators take part. This synchronization typical of mood swings could be engineered by multiple interactions between the many nerve networks involved, or it is possible that one component has a greater regulatory role than the others. In other words, the orchestra may have a conductor. One candidate for this position is the adrenergic system in the brain. Very little is known about the functions of this system, which has achieved a Cinderella status amongst the brain’s neuromodulators. It is certainly a very primitive system, since its C1–C3 cell bodies are located in a more primitive location – the medulla – than the
The Neurochemistry of Consciousness
others, which have their cell bodies higher-up in the pons, midbrain, and basal forebrain. Some of its known activities are related to low-level functions, such as cardiac and respiratory control. This has led to its relative neglect. However, it has extensive rostral projections that reach as far up as the thalamus, and, more particularly from our present point of view, to the nuclei of all the other neurotransmitters – extensively to the substantia nigra (DA), the locus coeruleus (NE), and the raphe nucleus (5-HT), plus a small projection to the LDTN (cholinergic); as well as extensive projections to key limbic structures, such as the hypothalamus, the central tegmentum, the periaqueductal grey, and the midline paraventricular nucleus of the thalamus (a key center involved in correlation of emotion with cognition). It is difficult to see why a system, that is supposed to control only low-level functions, such as heart rate, respiration, and blood pressure, would need this extensive projection to most of the key higher areas of the limbic system that are so closely involved in all ‘higher’ mental functions including consciousness. Progress has been hampered by the fact that the usual lesion and stimulation experiments, used to explore function on other brain areas, are difficult in the medulla owing to the proximity of vital centers. Therefore this system may have some other as yet undiscovered ‘higher’ function. Thus, the medullary adrenergic system is well placed and well connected to act as the putative conductor of this orchestra, that controls the basic mechanisms of mood, and that waves its baton too excitedly in mania, and too lethargically in depression. Just as the cholinergic PP/LDTN in the pons and midbrain orchestrate consciousness in the brain, so the even more fundamental C1–C3 adrenergic nuclei in the medulla could orchestrate the basic behaviors of approach and retreat, with all their emotional and cognitive accompaniments, that constitute mood. Further research on this neglected system seems indicated.
The Microanatomy of Cognition and Schizophrenia There is a quantity of data that supports the conclusion that reactive oxygen species (such as hydrogen peroxide and superoxide) and antioxidants (such as ascorbate and glutathione) play key roles in many
109
brain mechanisms involved in neurocomputation, synaptic plasticity, and the brain activities that construct consciousness. The complexity of this system is such that it cannot be described under any simple rubric. But its relation to consciousness can be demonstrated by showing its disorders found in schizophrenia (as well as the desferrioximine/prochlorperazine coma discussed above). Schizophrenia is characterized by hallucinations, delusions, and a general fragmentation of the conscious mind. Until recently it was thought to be caused by the production of some endogenous neurotoxin, or by some imbalance in the brain’s neurotransmitter system, such dopamine or glutamate. There is now, however, strong evidence that schizophrenia is due to a disorder of synaptic plasticity in the brain, in particular damage to the neuropil. The neuropil consists of the fine network of axons, dendrites, and dendritic spines that fill the space in the brain between neuronal and astrocytic cell bodies. In many cases of schizophrenia, particularly in type 2 cases, the volume of the neuropil is reduced by up to 50%. This means that the communications between neurons, on which neurocomputation depends, has been severely jeopardized. Computer modeling has shown that, in a complex computational network, if the quantity of interunit connections is reduced by this amount, the whole computational network malfunctions. In place of its usual appropriate activity, it becomes dominated by fixed, circulating masses of inappropriate activity, comparable to the hallucinations and delusions found in schizophrenia. This faulty neuropil is thought to be produced partly by genetic or epigenetic factors, so that the synaptic construction and modulation system is not built properly in the first place, rendering it to malfunction at some point in its history, and partly by environmental insults such as viral infections in utero. One relevant factor is thought to be neurotoxic oxidative stress. Schizophrenic patients show many signs of synaptic pathology and oxidative stress. These include the following pathologies:
Abnormalities in postsynaptic proteins found in the postsynaptic densities of NMDA and AMPA receptors, such as PSD95 and SAP102. The expression of growth-associated protein-43 (GAP-43). This is a presynaptic lipoprotein involved in the control of neuronal development and synaptic plasticity.
110
The Neurochemistry of Consciousness
Reduced expression of synaptophysin and the vesicular glutamate transporter. A robust decrease in reelin expression. This is a key protein in neural development secreted by GABAergic neurons that modulates inter alia spine maturation. Oxidative damage to membrane lipids has been reported as well as several other abnormalities in indices of oxidative stress. Other preliminary findings suggest that there may also be abnormalities in the MAP-kinase cascade that modulates synaptic plasticity, in scaffolding proteins in the postsynaptic density, and in the function of brain-derived neurotrophic factors, oct-6 and S100R. LSD induces expression of many genes (such as c-fos, ARC, aria3, and others) related to proteins involved in modulating synaptic plasticity.
These studies show that the brain in schizophrenia presents evidence of various abnormalities in the complex neurochemical processes involved in synapse formation, augmentation, maintenance, and deletion.
Suggested Readings Karczmar AG (2007) Exploring the Vertebrate Central Cholinergic Nervous System. New York: Springer. Monti JM, Pandi-Perumal SR, Jacobs BL, and Nutt DJ (eds.) (2008) Serotonin and Sleep. Basel: Birkha¨user. Perry E, Ashton H, and Young AH (eds.) (2002) Neurochemistry of Consciousness: Neurotransmitters in the Mind. Amsterdam: John Benjamins. Rowell PP, Yolk KA, Li J, and Bickford ME (2003) Investigations of the cholinergic modulation of GABA release in rat thalamus slices. Neuroscience 116: 447–453. Smythies J (1996) The functional neuroanatomy of awareness. Consciousness & Cognition 6: 455–481. Smythies J (1997) The biochemical basis of synaptic plasticity and neural computation: A new theory. Proceedings of the Royal Society of London. Series B 264: 575–579. Smythies J (1999) The biochemical basis of coma. Psycholoquy, www.cogsci.soton.ac.uk/psyc. Smythies J (2002) The Dynamic Neuron. Cambridge, MA: MIT Press. Smythies J (ed.) (2004) Disorders of Synaptic Plasticity and Schizophrenia. Vol. 59: International Review of Neurobiology. San Diego: Elsevier. Smythies J (2005) The Neuromodulators. Vol. 64: International Review of Neurobiology. San Diego: Elsevier. Steriade M and McCarley RW (2006) Brain Control of Wakefulness. New York: Springer. Yu AJ and Dayan P (2002) Acetylcholine in cortical inference. Neural Networks 15: 719–730.
See also: Altered and Exceptional States of Consciousness; General Anesthesia; Neurobiological Theories of Consciousness; Psychoactive Drugs and Alterations to Consciousness.
Biographical Sketch
John Smythies MD is a graduate of Cambridge University and is currently a research scientist at UCSD and a senior research fellow at the Institute of Neurology, University College, London. Previously he has held the positions of C.B. Ireland professor of psychiatric research at the University of Alabama Medical Center, and reader in psychiatry at Edinburgh University. He has served as President of the International Society for Psychoneuroendocrinology (1970–1974), consultant to the World Health Organization (1963–1968), and editor of the International Review of Neurobiology (1956–1991). He is a fellow of the Royal College of Physicians (London) and was elected a member of the Athenaeum in 1968. He is the author of over 200 scientific papers and 16 books.
Neuroscience of Volition and Action M Jeannerod, Institut des Sciences Cognitives, Lyon, France ã 2009 Elsevier Inc. All rights reserved.
Introduction The main feature of voluntarily controlled actions is that they are generated from within, and that their generation requires the preexistence of an internal state (a representation) whereby they can be encoded, stored, and ultimately performed independently from external influences. Thus, they cannot be accounted for by automatic sensorimotor mechanisms, where a causal relationship between an immediate (sensory) cause and a visible behavior can be identified. Instead, voluntary actions are to be considered, not as mere responses to a stimulus, but as the self-generated expression of conscious cognitive states. A central question in this article will be to determine to what extent voluntary actions can be consciously controlled. This issue encompasses a wide range of technical as well as theoretical problems in neuroscience, psychology, and even philosophy.
Historical Landmarks The existence of actions generated in the absence of sensory input was a central issue in nineteenth century neurology. Those who were influenced by thinkers like Charlton Bastian or William James defended the idea that peripheral influences, immediate or delayed, was a necessary condition for an action to appear. Sherrington, in showing the absence or poverty of spontaneous motor behavior in monkeys with deafferented limbs, provided experimental support to this theory. Alternatively, Hugo Liepmann, starting from the background of clinical neurology, abandoned the opposition between peripheral or central origin of the action. Instead, he proposed a completely new scheme in concentrating on how an action can be assembled from its elementary constituents. According to Liepmann, an action must proceed from an internal ‘plan’: ‘‘The main representation
of the goal can only be reached if a plan (Entwurf ) is built internally, concerning the direction, contiguity, succession, and rhythm of the elementary acts.’’ To account for the implementation of the plan, he proposed that the elementary bits of the action were assembled according to the main representation: The result of this process was what he called a ‘movement formula’ (Bewegungsformel), that is, an anticipatory hierarchical structure where all the aspects of an action were represented. Liepmann’s legacy is still quite influential in neuropsychology and cognitive neuroscience of action. Later authors replaced the term of movement formula by those of ‘engram,’ ‘schema,’ or ‘internal model,’ but kept the notion of a representational level with a hierarchical organization. Another, closely related, debate was about whether actions could be consciously monitored and controlled. This debate became known as the ‘Two Williams debate.’ On one side, William James, although he clearly considered will as an internal driving force, defended the opinion that the consciousness of our movements is based on a posteriori information from sensory organs. On the other side, Wilhelm Wundt held that conscious knowledge about our actions was based on a priori efferent information of a central origin. Wundt thought that we can perceive specific feelings, which he called the feelings of innervation (Innservationsgefu¨hl), when we produce a movement. He was a proponent of the notion that the mind has a mental content that can be probed by appropriate methodology: To him, voluntary actions were an expression of this mental content.
What Do We Know about Our Actions? Whoever of James or Wundt was right, the fact is that we remain unaware of many of our own actions. It is common experience that when leaving 111
112
Neuroscience of Volition and Action
home we ask ourselves questions like ‘‘Did I lock the door?’’ or ‘‘Did I turn off the light?’’ immediately after having done it. We may also drive the car back home and suddenly realize that we are at destination without having the least idea of how we did it. But, to the contrary, there are situations where we remain fully aware throughout the action. When I do something for the first time, for example, start a new laptop computer, I try to carefully follow the instruction manual and to consciously control each step of the action. Here, we first examine the question of what consciousness of action is about. The first answer to this query that comes to mind is that consciousness of action refers to what the action is about. Actions have goals: To be aware of the goal one reaches for is one way of being conscious of the action undertaken to reach that goal. Being aware of the goal, however, does not imply to be aware of how it is being reached. Indeed, our above definition of voluntary control should imply that an agent can consciously access the internal representation of the action he is performing. To examine this point, we need experimental situations where the subjects have to ask themselves questions about what they really did to achieve a motor task. An example of such situations are those that produce a conflict between the different signals that are generated at the time of execution of an action (e.g., visual signals, proprioceptive signals, signals arising from central motor commands), and that are normally congruent with each other. Subjects’ reports about their feelings in situations where these signals become mutually incongruent thus provide a direct insight into their ability to consciously monitor these signals. This type of experiments involving a sensorimotor conflict has a long tradition in psychology. They can typically be created by optical devices: Looking at one’s moving hand through an inverting prism, for example, makes the movements to appear inverted, such that, in order to reach for a visual target seen on my right, I have to move my hand to the left. The main objective of these classical experiments was to study the process of adaptation, through which visuomotor coordination progressively rearranges for matching the hand movements to the apparent location of the visual targets. But only a few systematic attempts were made at studying
the motor awareness of the subjects, that is, their insight about how they performed their own movements when unknowingly faced with this type of conflict. The research on motor awareness was initiated by Torsten Nielsen more than 40 years ago. In a more recent version of Nielsen’s experiment performed by P. Fourneret and M. Jeannerod, the subjective reports of subjects about what they had just done in trying to reach for a visual target were quantified. Participants were instructed to draw straight lines between a starting position and a target, using a stylus on a digital tablet. The output of the stylus and the target were displayed on a computer screen. The participants saw the computer screen in a mirror placed so as to hide their hand. On some trials, the line seen in the mirror was electronically made to deviate from the line actually drawn by the subject. Thus, to reach the target, the subject had to deviate his or her movement in the direction opposite to that of the line seen in the mirror. At the end of each trial, the subject was asked to indicate verbally in which direction he or she thought his or her hand had actually moved. The results were twofold: first, the subjects were consistently able to trace lines that reached the target, that is, they accurately corrected for the deviation; second, they gave verbal responses indicating that they thought their hand had moved in the direction of the target, hence ignoring the actual movements they had performed. Thus, although they correctly performed the task, they were unable to consciously monitor the discordance between the different signals generated by their own movements, and falsely attributed the direction of their hand to that of the line. In other words, they tended to adhere to the visible aspect of their performance, and to ignore the way it had been achieved. In fact, the access to motor awareness seems to be a matter of threshold. When, in the above experiment, the deviation of the line was progressively increased, the subjects’ behavior suddenly changed. Their accuracy in reaching the target deteriorated and they became aware of the deviation: They could report that the movement of their hand erred in a direction different from that seen on the screen and that, in order to fulfill the instruction of reaching the target, they had to deliberately orient their hand
Neuroscience of Volition and Action
movement in a direction different from that of the target. The breakpoint at which the subjects shifted from an automatic compensation to a conscious adjustment to the deviation was around 14 on average. Thus, the awareness of a discordance between an action and its sensory consequences emerges when the magnitude of the discordance exceeds a certain amount. In another experiment based on the same principle and made by G. Knoblich and T. Kircher, the participants had to draw circles on a writing pad. As in the above experiments, they did not see their hand, but they saw an image of their movement, represented by a moving dot on a computer screen. The velocity of the moving dot was either the same as that of the subject’s movement, or it could be unexpectedly accelerated by a variable factor of up to 80%. To compensate for the change in velocity and to keep the dot move in a circle, as requested by the instruction, subjects had to decrease the velocity of their hand movement by a corresponding amount. They had to indicate any perceived change in velocity of the moving dot. Although they failed to detect the changes in velocity when they were of a small amplitude, their detection rate increased for faster velocity changes. Yet, subjects were found to be able to compensate for all changes in velocity, including for those that they did not consciously detect. These experimental situations have a counterpart in everyday life, for example, when an action which was unfolding automatically and outside the subject’s awareness fails because of some unexpected obstacle: The failure triggers a prise de conscience of the ongoing action. Such situations where an action is delayed, incompletely executed or blocked, typically create a mismatch between the desired motor output and its observed result. This mismatch would represent the ‘stimulus’ that triggers the conscious experience. Another set of experiments using neuroimaging techniques have directly examined the neural mechanism involved in this change in conscious awareness. They found that the neural activity of cortical areas located in the posterior and ventral parts of the right parietal lobe increases during a visuomotor conflict similar as those described in the above behavioral experiments. This increase appears to be a function of
113
the degree of the conflict. The changes in activity of these areas thus reflect the incongruence between normally congruent signals from central or sensory origin generated during a self-generated movement: The central command signals generated for moving the hand in a certain direction are contradicted by the visual signals showing a movement in a different direction. The role of the right parietal lobe in monitoring such signals is supported by clinical observations. Patients with lesion in this region show striking neuropsychological symptoms that testify to an alteration of self-consciousness and consciousness of action: neglect of contralateral space, neglect of the corresponding half of the body, denial of ownership, or even denial of the frequently associated hemiplegia (anosognosia).
Is Consciousness of Action an Afferent or an Efferent Phenomenon? Several sources of information are potentially available to make a conscious judgment about one’s own motor performance. Among the sensory sources are the visual and the kinesthetic signals. Visual signals are directly derived from vision of the moving limb, or indirectly from the effects of the movement on external objects; kinesthetic signals are derived from movement-related mechanical deformations of the limb, through receptors located in the skin, joints, and muscles. Nonsensory sources are mainly represented by central signals originating from various levels of the action generation system. These different types of signals do not have the same status. Visual cues are of an uncertain origin: They cannot differentiate a selfgenerated from an externally generated visual change. By contrast, the central cues and, to some extent the kinesthetic cues, clearly relate to a selfgenerated movement: They are ‘first-person’ cues in the sense that they can only conceivably arise from the self. Trying to examine the respective contributions of these different signals reminds the classical issue raised by the Two Williams debate alluded to in the historical section: Is consciousness of action an afferent or an efferent phenomenon? For a long time, experimenters have failed to resolve this
114
Neuroscience of Volition and Action
issue, mainly because of the methodological difficulty of isolating the two sources of information from one another. There are no reliable methods for suppressing kinesthetic information arising during the execution of a movement. Alternatively, it is possible to prevent muscular contractions in a subject who is attempting to move, for example, by using a curarizing agent (i.e., an agent that blocks neuromuscular transmission) for paralyzing one limb: If the subject reports sensations from his attempts to move his paralyzed limb, these sensations should result from outflow motor commands, not from proprioceptive inflow. The available evidence shows that no conscious perception of movement arises in this condition. However, experiments where an arm is only partially curarized (the arm is not paralyzed, but muscular force is weakened) suggest a more balanced conclusion: Subjects requested to estimate the heaviness of weights that they attempt to lift with their weakened arm report an increased perceived heaviness. This illusion was interpreted as reflecting the increase in motor outflow needed to lift the weights. This result provides indirect evidence as to the possibility for central signals to influence conscious experience. A more direct solution to this problem is to examine patients with a pathological loss of haptic sensations (sensations from skin, joints, and muscles). One such patient, patient GL suffering a long-term sensory neuropathy, has been extensively studied by several experimenters, including Y. Lamarre and J. Paillard. Patient GL has no haptic information about the movements she performs. Thus, provided visual feedback from her movements is suppressed, the only information on which she can rely to form a phenomenal experience about her own action must be derived from central signals during the action generation processes. This point was examined by using the apparatus already described for the Fourneret and Jeannerod experiment. GL had to draw a line with her unseen hand while the line was made to deviate to the right by an angle increasing from 1 to 20 over successive trials. Like a normal subject, GL performed the task without difficulty: She was able to compensate for the deviation and to reach the target. When asked, at the end of each trial, to estimate verbally the angle by which she thought her hand had
deviated to the left for bringing the line to the target, GL never explicitly reported a feeling of discordance between what she had seen and the movement that she thought she had made. Remember that, in this task, normal subjects become clearly aware of a displacement of their hand toward the left to compensate for the disturbance when the discordance exceeds a certain value. Instead, GL consistently gave responses indicating that she thought she had drawn the line in the direction of the target. In spite of expressing perplexity at the end of some trials, GL never became aware of the discordance and, consequently, of any strategy of correction she had to apply to correct for the deviation. When asked to describe her feelings, she only mentioned that she found the task ‘difficult’ and requiring an ‘effort of concentration.’ Another experiment with the same patient also addressed the question of a possible role of the efferent processes in motor consciousness, by exploring the production and the perception of muscular force. When muscular force is applied isometrically (with no change in muscle length), kinesthetic input is limited, because there is no displacement of the limb: Thus, this condition should maximize the role of the central commands in the conscious appreciation of the exerted force. When instructed, first to isometrically apply a certain degree of force with one hand, and then to match this degree of force with the other hand, GL performed with a close to normal accuracy. This result indicates that she was able to produce accurate central commands. Yet, she was unable to report any conscious feeling from her effort, neither did she experience fatigue when a high degree of muscular contraction had to be maintained. The central, nonsensory cues, which are still available in patient GL, appear to be of little use for consciously monitoring her own movements, except for the vague feelings of effort and difficulty that she reported in one of the tasks. However, a mere opposition between peripheral and central sources of information in providing cues for consciousness (which was the core of the Two Williams debate) may be misleading because it does not take into account the complete set of events arising during the voluntary execution of a movement. In the current model of selfgenerated actions, as designed by D. Wolpert and
Neuroscience of Volition and Action
his colleagues, the central signals are normally monitored and used as a reference for the desired action, and the reafferent sensory (e.g., kinesthetic) signals arising from the executed movement have to be matched with this reference. In patient GL, because no reafference resulted from her executed movement, this matching process could not take place. Here we propose that the conscious information about one’s movements is normally derived, not directly from the reafferent signals themselves, but from the output of the matching process, for which the presence of both central and peripheral reafferent signals is required. In the case of GL, it was the lack of kinesthetic input from movement execution or isometric force generation, which severely impaired the possibility to consciously monitor the efferent signals. The same explanation holds for the lack of conscious feelings during attempted movements in completely paralyzed subjects (e.g., produced by curarization). In spite of the generation of intense motor commands for fighting against the paralysis, the absence of corresponding kinesthetic reafferent signals prevents the matching process to take place. Conversely, when reafferent signals are present, but do not match the central signals, the matching process generates signals proportional to the degree of mismatch between the two. This would account for the conscious sensations of effort reported by subjects with an incomplete paralysis, where kinesthetic signals are still preserved and in normal subjects during lifting weights with incompletely paralyzed arms.
Can We Perceive Our Own Intentions? The picture of consciousness that arises when it is studied in its relation to voluntary action is that of a post hoc phenomenon. Consciousness of action is bound to signals resulting a posteriori from the completion of the action itself, not to central signals that arise prior to the action. Accordingly, one should not expect that events preceding the execution of an action can come to consciousness. Is this also the case for intentions? Intentions are part of the representation of actions; they contribute to the stream of processing that flows
115
from the early part of the representation down to action execution, when it occurs. In this section, we concentrate on experiments which have looked for the unfolding of intentions, their neural correlates, and the conditions for their conscious access. Benjamin Libet was the first to tackle this problem. He instructed subjects to perform simple hand movements ad libitum and to report the instant (W) at which they became aware of wanting to move. In order to do so, subjects verbally reported the clock position of a spot revolving on a screen. An electromyogram (EMG) was recorded from arm muscles for measuring the precise onset of the movement: The ‘W’ judgment was found to precede EMG onset by 206 ms. In addition, electroencephalogram (EEG) potentials were recorded from the subjects’ skull. The ‘readiness potential,’ a DC potential that appears during preparation to voluntary action, was found to anticipate ‘W’ by about 345 ms. This striking result shows that the intention (in the sense of ‘wanting to move’ or ‘feeling the urge to move’) can be perceived as distinct from execution itself; it also shows that the subject’s declarative awareness of this phenomenon does not correspond to the actual onset of movement preparation, which starts much earlier. The Libet’s paradigm revealed the possibility (to be discussed further in the next section) of attending to one’s own mental states preceding action. It also provided important information on the brain areas involved in this conscious process. A specific region of dorsal premotor cortex (the region corresponding to the supplementary motor area (SMA)) was consistently found to be activated in conjunction with the feeling of urging to move. First, the readiness potential, which in the experiments by Libet and his colleagues and by their followers, precedes the subjects’ report of the feeling (the ‘W’ judgment), is thought to originate from SMA. Second, electrical stimulation of a small area immediately anterior to the SMA (the pre-SMA) induces a feeling of urging to move, as reported by patients during direct low intensity electrical stimulation applied to this area during neurosurgical operations. The pre-SMA and the SMA are relatively closely connected to motor cortex: Indeed, increasing the intensity of the stimulus produces muscular contractions in the body part where the feeling was resented. Finally, a
116
Neuroscience of Volition and Action
functional magnetic resonance imaging (fMRI) study of normal subjects attending to their intention showed that the ‘W’ judgment made by these subjects was associated with activation of the preSMA and the dorsolateral prefrontal cortex. The dorsolateral cortex itself is known to be involved in willed actions. Its role in carrying intentions and motor plans, conscious or not, and in inhibiting concurrent activity has been demonstrated by neuropsychological experiments in normal subjects as well as by clinical observations (see below). Christopher Frith and his colleagues went further in monitoring brain activity in normal subjects during tasks involving a voluntary choice between two different actions (what they consider as the hallmark of free will). In one condition, the subjects were instructed to move, at their own choice, either their right index finger or their right thumb. In the other condition, they had to utter a word, for example, beginning with the letter s. Brain activity measured with positron-emission tomography (PET) was found to be specifically increased in an area corresponding to the lateral prefrontal cortex corresponding to Brodmann area 46: in the condition of finger movement, the activation was bilateral; in the condition of word choice, it was limited to the left side. The role of prefrontal cortex and other cortical regions in voluntary action is also supported by experiments measuring the time course of brain activity during conscious decisions. Recent fMRI experiments by Haynes and his colleagues in Berlin demonstrated that predictive information about the decision to move could be detected in prefrontal cortex (in the frontopolar area 10) and in a posterior parietal region up to 7–10 s prior to subject’s awareness of his or her decision. This suggests that, by the time the subject’s decision reaches awareness, it has been influenced by unconscious brain activity lasting several seconds.
How Do We Get the Sense of Being the Agent of Our Actions? Actions like aimlessly moving one’s finger or uttering a phoneme, such as those described in the previous section, have no real goal and little
impact on the external world. Still, they represent a minimal case for voluntary actions: They result from a deliberate decision or choice from the agent. In John Searle’s classification, they correspond to what he calls prior intentions, that is, intentions which are accompanied by the feeling of intending, wanting, wishing, choosing, or, in other words, of the feeling of conscious will. In classical philosophy, the concept of will is associated with the subjective impression of causing an action to appear. Thomas Reid, for example, in his Essays on the Active Power of the Human Mind (1788), wrote that ‘‘every man is conscious of a power to determine, in things which he conceives to depend upon his determination. To this power, we give the name of will.’’ For the psychologist, however, the problem is that of the relationship of this conscious sense of willing to act to the forthcoming action. As we saw earlier, execution of actions is automatic, and it is not preceded by a clear insight about what is going to happen. The brief flash of consciousness found by Libet in the 200–300 ms preceding muscular contraction is not likely to represent a significant cause for the action to appear. Indeed, Libet considered that his own results ‘‘lead to the conclusion that cerebral initiation. . . of a spontaneous voluntary act. . . can and usually does begin unconsciously’’. ‘‘The brain,’’ he said, ‘‘‘decides’ to initiate or, at least, to prepare to initiate the act before there is any reportable subjective awareness that such a decision has taken place.’’ The fact that consciousness follows, and does not precede, the process of action generation is hardly compatible with the notion of a causal role of the conscious sense of will. We are therefore faced with a paradox: whereas we feel and we strongly believe that our thoughts determine our behavior, we realize that conscious free choice and conscious will are consequences, not causes, of the brain activity that itself causes the actions to appear. The fact that brain activity is ahead of our mental states, and not the reverse, is not a surprise. The surprising thing, however, is the discordance between this very fact and our subjective experience of being the agent of the action (the sense of agency). Before discussing the issue of the causal will, we will examine experimentally what are the cues which determine the sense of agency. One way to answer this question is to examine the responses of
Neuroscience of Volition and Action
subjects placed in experimental situations where an uncertainty is artificially created about the origin of an action, and where they are explicitly requested to make conscious agency judgments. In these experiments, normal subjects are instructed to execute simple finger movements without direct visual control of their hand. Instead of seeing their hand during their movements, subjects see on a screen the image of two hands: Their own hand or an alien hand (the hand of an experimenter) executing the same or different movements. Successive trials involve different degrees of conflict between the seen and the felt position of the subject’s hand, and between the seen and the actually executed finger movements: The two hands can either make the same movements, different movements, or no movement at all. At the end of each trial, after the two hands have disappeared, a pointer is placed at the location of one of the two hands and the subject is asked to judge whether that hand was his own or that of the experimenter. In this condition, the subject’s errors can be in two different directions: in the other-to-self direction when the subject attributes to himself the hand of the experimenter (over-attribution errors) and in the self-to-other direction when he attributes his own hand to the experimenter (underattribution errors). According to E. van den Bos and M. Jeannerod who made this experiment, the overall pattern of results was that the subjects tended to make more over-attribution errors. This pattern was particularly striking in the condition where all the cues for discriminating between the two hands had been suppressed: In spite of responding at chance on average, the subjects still made significantly more errors by over-attribution than by under-attribution. The conclusion that can be drawn from the above results is twofold. First, the sense of agency can be subject to misinterpretation and to illusions (and possibly even to delusions, in psychotic states such as schizophrenia). Second, attribution judgments are biased: A subject placed in ambiguous situations like those created by the substitution experiments experiences alien actions as his own. This is indeed an example of an illusion of agency: Faced with an action of an uncertain origin, this subject will tend to believe that he is the agent who caused that action to appear.
117
The dissociation in normal subjects, for which we already saw several illustrations, between automatic processing of an action and conscious responses about that action raises the question of the relationship between their respective underlying mechanisms. Would the sense of agency be affected by perturbation of the automatic process like those we reported in the previous sections? Indications in the literature are very scarce on this point. Nielsen had reported that his subjects, when misattributing the alien hand to themselves, felt as if they had lost control of their movements, ‘as if driving on ice.’ Other authors reported that subjects observing a delayed presentation of their motor performance experienced the bizarre sensation of having an ‘anarchic’ hand. In both examples, subjects experienced the illusion of a disturbed sense of agency when the feedback from their actions did not correspond to the expected effect.
Disorders of Volition Pathological disorders affecting volition may shed light on the debate about the role of conscious will in behavior. The expression ‘disorders of volition’ refers to those pathological conditions where the ability to make choices, to express preferences, or possibly to experience pleasure and freedom in making these choices or expressing these preferences, are affected. The´odule Ribot, in his book on the diseases of the will, considered two opposite sides, the negative and the positive, of the alterations of the will. Several terms have been used to describe such pathological conditions. On the negative side, aboulia, a pure disease of the will according to Ribot, has been defined as an impairment to execute what is in mind. In this condition, there is no paralysis, no disorder of the muscular system, there is no lack of desire, but the transition from desire to execution becomes abnormally difficult. An extreme example of aboulia has been described clinically under the term of athymhormia: Such patients show loss of drive and of search for satisfaction, lack of curiosity, lack of taste and preferences, flattened affect. This condition, clearly distinct from depression, can be observed in schizophrenia and may also be caused by lesions in the
118
Neuroscience of Volition and Action
basal ganglia. Another example is the ‘chronic fatigue syndrome’ (CFS), defined by persistent or relapsing unexplained fatigue. CFS patients were tested in a task where they had to imagine themselves performing an action (a motor imagery task), while their brain activity was monitored with fMRI. They were found behaviorally to be slower than controls. In addition, an area in their anterior cingulate cortex, known to be involved in error monitoring, remained inactive when they made errors, which supports the notion that CFS may be associated with an inability to build action plans and to monitor one’s own performance according to these plans. On the positive side are found conditions where the deficient will cannot block impulses to act: The power to control and to inhibit, which Ribot considered as the highest level of the will, is impaired. A well-known example of this pathological condition is that of patients with frontal lobe lesions, who are unable to consciously monitor their performance and to override their automatic responses. These patients tend to compulsively imitate gestures or even complex actions performed in front of them by another agent; when presented with common graspable objects, they cannot refrain from using them. This striking behavior, termed ‘imitation’ or ‘utilization’ behavior by F. Lhermitte, may be explained by an impairment of the normal suppression, by the prefrontal areas, of inappropriate actions triggered by external stimuli. As a consequence of this impairment, the behavior of the patient is under the dependence of external events. One such patient exemplified this typical behavior by putting pairs of goggles on his nose each time he was presented a new pair, ending with several pairs superimposed upon each other! This type of behavior observed following frontal lesions suggests the existence of an imitative tendency, which would normally be inhibited in everyday life situations. Compulsive imitation in frontal patients would result from an impairment of this inhibitory mechanism. Following this idea, Marcel Brass and his colleagues have studied brain activity in normal subjects who were instructed to execute a simple predefined finger movement (e.g., lift one finger) in response to the onset of an observed movement executed by an experimenter. When the observed movement was congruent with the
movement the subject was instructed to perform, the response was given within a short latency; in contrast, when it was incongruent (e.g., the subject had to lift his finger in response to an observed tapping movement), the latency was increased. In addition, in incongruent trials, the lateral prefrontal cortex in the middle prefrontal gyri was strongly activated. It is therefore conceivable that a prefrontal lesion involving the lateral prefrontal cortex impairs the possibility to refrain from imitating other people. Lhermitte had used the term ‘environmental dependency syndrome’ to designate the tendency of these patients to stick to external events, at the detriment of their own free will.
Conclusion At this point, the problem of the consciousness of our own actions and intentions clearly merges with that of self-consciousness. The ability to identify oneself as the agent of a behavior or a thought – the sense of agency – is the way by which the self builds up as an entity independent from the external world and from other agents. The conscious sense of will arises from the belief that our thoughts have a causal influence on our behavior. While we tend to perceive ourselves as causal, we actually ignore the cause from which our actions originate. Because the conscious thought and the observed action are consistently associated, even though they may not be causally related, the narrative self tends to build a cause-and-effect story. Therefore, conscious free choice, like conscious will, may not be a direct perception of a causal relation between a thought and an action, but rather a feeling based on the causal inference one makes about the data that do become available to consciousness. In conclusion, according to Wolfgang Prinz, ‘‘There appears to be no support for the folk psychology notion that the act follows the will, in the sense that physical action is caused by mental events that precede them and to which we have privileged access.’’ ‘‘Experimental evidence suggests that two different pathways [. . .] may be involved in action – one for the generation of physical action and one for the mental awareness of its causal antecedents. If anything, the second follows the first, and not vice-versa [. . .].’’
Neuroscience of Volition and Action
The role of consciousness should rather be to ensure the continuity of subjective experience across actions which are – by necessity – executed automatically. Because it reads behavior rather than starting it, consciousness represents a background mechanism for the cognitive rearrangement after the action is completed, for justifying its results, or modifying the factors that have been at the origin of an action which turned out to be unsuccessful, as in learning a new skill, for example. This mechanism could have the role of establishing a declarative cognitive statement about one’s own preferences, beliefs, or desires. But the fact is that what we know about our mental content does not match its actual functioning, that which determines the causal relations of our behavior. When asked questions about these causal relations, we tend ‘to tell more than we can know’ and to build explanations based on incomplete information. This tendency unavoidably creates a dissociation between what we believe we do and what we actually do. See also: Brain Basis of Voluntary Control; Free Will; Habit, Action, and Consciousness; Perception, Action, and Consciousness.
Suggested Readings Berrios GE and Gili M (1995) Will and its disorders: A conceptual history. History of Psychiatry 6: 87–104.
119
Brass M, Zysset S, and Von Cramon Y (2001) The inhibition of imitative response tendencies. NeuroImage 14: 1416–1423. Fourneret P and Jeannerod M (1998) Limited conscious monitoring of motor performance in normal subjects. Neuropsychologia 36: 1133–1140. Gandevia SG and McCloskey DI (1977) Changes in motor commands, as shown by changes in perceived heaviness, during partial curarization and peripheral anaesthesia in man. Journal of Physiology 272: 673–689. Haynes J-D, Sakai K, Rees G, Gilbert S, Frith C, and Passingham RE (2007) Reading hidden intentions in the human brain. Current Biology 17: 323–328. Jeannerod M (2006) Motor Cognition: What Actions Tell the Self. Oxford: Oxford University Press. Knoblich G and Kircher TTJ (2004) Deceiving oneself about being in control: Conscious detection of changes in visuomotor coupling. Journal of Experimental Psychology. Human Perception and Performance 30: 657–666. Lhermitte F (1983) Utilisation behaviour and its relation to lesions of the frontal lobes. Brain 106: 237–255. Libet B, Gleason CA, Wright EW, and Pearl DK (1983) Time of conscious intention to act in relation to onset of cerebral activity (readiness-potential). The unconscious initiation of a freely voluntary act. Brain 106: 623–642. Liepmann H (1905) Ueber Sto¨rungen des Handelns bei Gehirnkranken. Berlin: S. Kargen. Nielsen TI (1963) Volition: A new experimental approach. Scandinavian Journal of Psychology 4: 225–230. Prinz W (2003) How do we know about our own action? In: Maasen S, Prinz W, and Roth G (eds.) Voluntary Action. Brains, Minds and Sociality, pp. 21–33. New York: Oxford University Press. Ribot T (1894) Les maladies de la volonte´. Paris: Alcan. Wegner D (2002) The Illusion of Conscious Will. Cambridge, MA: MIT Press. Wolpert DM, Ghahramani Z, and Jordan MI (1995) An internal model for sensorimotor integration. Science 269: 1880–1882.
120
Neuroscience of Volition and Action
Biographical Sketch
Marc Jeannerod was born in Lyon, France. He completed his degree in medicine at the University of Lyon, with a specialization in neurology and clinical neurophysiology. After a post doctoral stay at the University of California at Los Angeles, he became professor in physiology at the Claude Bernard University in Lyon, head of the INSERM Laboratory of Experimental Neuropsychology (1978–1996), and director of the Institute for Cognitive Science (1997–2004). Marc Jeannerod has published several books devoted to the control of action. His last book is Motor Cognition (OUP, 2006).
Perception, Action, and Consciousness M A Goodale, University of Western Ontario, London, ON, Canada A D Milner, Durham University, Durham, UK ã 2009 Elsevier Inc. All rights reserved.
Glossary Basal ganglia – Several large clusters of nerve cells (including the caudate nucleus, putamen, and the globus pallidus) located deep in the brain below the cerebral hemispheres. The basal ganglia play a role in a number of functions including motor control, cognition, and emotion. Blindsight – Residual visual abilities in individuals with large lesions of primary visual cortex who claim that they are blind in that part of the visual field in which they show evidence of visual sensitivity. For example, patients with unilateral lesions of primary visual cortex can sometimes point accurately toward stimuli in visual field contralateral to their lesion while at the same time denying any conscious experience of those stimuli. Brainstem – The major route by which the forebrain receives information from, and sends information to, the spinal cord and peripheral nerves. The chief divisions of the brainstem are the midbrain, pons, and medulla. Cortical blindness – Blindness that results from damage to the primary visual areas in the cerebral hemispheres. Individuals who have cortical blindness may nevertheless show blindsight. Dorsal stream – Visual pathway arising in early visual areas in the cerebral cortex and projecting to the posterior parietal cortex. This pathway is thought to subserve the visual control of actions, such as reaching and grasping. Extinction – In neuropsychology and neurology, extinction (or bilateral simultaneous extinction) refers to a phenomenon in which a patient with a unilateral lesion (typically in the inferior
regions of the posterior parietal cortex) fails to detect a visual (or tactile) stimulus presented contralateral to the lesion when another (similar) stimulus is presented simultaneously on the opposite side. When the contralateral stimulus is presented on its own, the patient is able to detect it. Hypoxia – A deficiency of oxygen reaching the tissues of the body. Inferotemporal cortex – The most ventral part of the temporal lobe. The inferotemporal cortex contains higher-order areas of the ventral stream of visual processing. Metacognition – Awareness of one’s own thinking and decision making; sometimes termed ‘knowing about knowing.’ Optic ataxia – An inability to guide the hand toward an object using vision. Optic ataxia is caused by damage to the posterior parietal cortex. Phototaxis – Movement of an organism toward (or away from) light. Posterior parietal cortex – Cortex in the parietal lobe behind the postcentral sulcus. A prominent sulcus in the posterior parietal cortex is the intraparietal sulcus (IPS). This region of the posterior parietal cortex contains visuomotor areas that make up the dorsal stream of visual processing. Premotor cortex – A complex mosaic of interconnected areas in the frontal lobe immediately anterior to primary motor cortex. Although there is not complete agreement about the function of different parts of premotor cortex, it has been suggested that areas in this region of cortex participate in motor planning and movement selection. Mirror cells are found in this region. Ventral stream – Visual pathway arising in early visual areas in the cerebral cortex of the
121
122
Perception, Action, and Consciousness
primate brain and projecting to the inferotemporal cortex. This pathway mediates visual perception, allowing the visual recognition of objects and events. Processing in this pathway is necessary, but not sufficient, for visual awareness. Visual agnosia – An inability to recognize visual stimuli despite spared low-level visual processing. There are several varieties of visual agnosia, all of which involve damage to some part of the ventral stream. In the case of associative agnosia, the patient is unable to recognize an object despite being able to draw a reasonably faithful representation of what he or she sees. In the case of apperceptive agnosia (or visual form agnosia), the deficit is more fundamental and the patient cannot recognize even simple shapes or discriminate between them – and is unable to copy line drawings.
Introduction Vision is the primary route to our conscious experience of the world beyond our bodies. Although we certainly hear and feel things in our immediate environment, neither of these sensory experiences is any match for the rich and detailed representation of the world provided by our sense of sight. The majesty of a distant mountain or the angry face of an approaching enemy can be appreciated only as visual experiences. But vision does more than provide us with our perception of the world. It also allows us to move around that world and to guide our goal-directed actions. Although it is tempting to think that these different functions of vision are mediated by one and the same visual representation in our brain, it has become increasingly clear over the last two decades that the visual pathways that underlie our perception of the world are quite distinct from those that underlie the control of our actions. Indeed, the distinction between ‘vision-for-perception’ and ‘vision-foraction’ has emerged as one of the major organizing principles of the visual brain, particularly with respect to the visual pathways in the cerebral
cortex. But before elaborating this distinction it is important to understand how vision began.
The Evolutionary Origins of Vision Visual systems first evolved not to enable animals to perceive the world, but to provide distal sensory control of their movements. One clear example of this is phototaxis, a behavior exhibited by many simple organisms whereby they move toward or away from light. Some bacteria, for example, use orange light as a source of energy for metabolic activity, but must avoid ultraviolet light, which can damage their DNA. As a consequence, these bacteria have developed a differential phototactic response, whereby the system measures light intensity at different wavelengths so that they end up moving toward orange light and away from UV light. To explain the bacteria’s light-sensitive behavior, it is not necessary to argue that these single-celled organisms ‘perceive’ the light or even that they have some sort of internal model of the outside world coded in their one or more of their organelles. One simply has to posit the existence of some sort of input–output device within the bacteria that links the intensity of ambient orange and UV light to the pattern of locomotion. As it turns out, exactly the same argument can be made about the visually guided behavior of more complex organisms, such as vertebrates. Indeed, as we shall see later, a broad range of human behavior can also be explained without reference to experiential perception or any ‘general-purpose’ representation of the outside world. In vertebrates, different classes of visually guided behavior have evolved as relatively independent neural systems. For example, in presentday amphibians, such as the frog, visually guided prey-catching and visually guided obstacle avoidance are mediated by separate neural pathways arising in the retina and projecting to distinct motor networks in the brain that produce the constituent movements of these two classes of behavior. In fact, evidence from several decades of work in both frog and toad suggests that there are at least five separate visuomotor modules, each responsible for a different kind of visually guided behavior and each having neural pathways from
Perception, Action, and Consciousness
a particular course of action has been chosen, the actual execution of the constituent movements of that action are typically carried out by dedicated visuomotor modules not dissimilar in principle from those found in frogs and toads. To summarize: vision in humans and other primates (and perhaps other animals as well) has two distinct but interacting functions: (1) the perception of objects and their relations, which provides a foundation for the organism’s cognitive life and its conscious experience of the world, and (2) the control of actions directed at (or with respect to) those objects, in which separate motor outputs are programmed and controlled online. These different demands on vision have shaped the organization of the visual pathways in the primate brain.
Two Visual Systems in the Primate Cerebral Cortex In the primate brain, two ‘streams of visual processing’ arise from early visual areas in the cerebral cortex and project to higher-order visual areas (Figure 1). One of these projection systems, the Posterior parietal cortex
Pulvinar or
D
Superior colliculus
Area V1+
Retina LGNd
Ve
nt
ra
ls
tre
am
am
tre
ls
sa
input to output. The outputs of these different modules certainly have to be coordinated, but in no sense are they guided by a single generalpurpose visual representation in the frog’s brain. There is evidence as well that the same kind of visuomotor modularity found in amphibians also exists in the mammalian and avian brain. Nevertheless, even though there is considerable evidence for visuomotor modularity in all classes of vertebrates, the very complexity of the day-today living in many mammals, particularly in higher primates, demands much more flexible organization of the circuitry. In monkeys (and thus presumably in humans as well), many of the visuomotor circuits that are shared with simpler vertebrates appear to be modulated by more recently evolved control systems in the cerebral cortex. Having this layer of cortical control over the more ancient subcortical networks makes it possible for primates to have much more flexible visually guided behavior. But even so, the behavior of primates, particularly with their conspecifics, is so complicated and subtle, that direct sensory control of action is often not enough. To handle these complexities, representational systems have emerged in the primate brain (and presumably in other mammals as well), from which internal models of the external world can be constructed. These representational systems allow primates such as ourselves to perceive a world beyond our bodies, to share that experience with other members of our species, and to plan a vast range of different actions with respect to objects and events that we have identified. This constellation of abilities is often identified with consciousness, particularly those aspects of consciousness that have to do with decision making and metacognition. It is important to emphasize that the perceptual machinery that has evolved to do this is not linked directly to specific motor outputs, but instead accesses these outputs via cognitive systems that rely on memory representations, semantics, spatial reasoning, planning, and communication. In other words, there are a series of cognitive ‘buffers’ between perceiving the world and acting on it, and the relationship between what is on the retina and the behavior of the organism cannot be understood without reference to other mental states, including those typically described as conscious. But once
123
Inferotemporal cortex Figure 1 Schematic representation of the two streams
of visual processing in human cerebral cortex. The retina sends projections to the dorsal part of the lateral geniculate nucleus in the thalamus (LGNd), which projects in turn to primary visual cortex ( V1). Within the cerebral cortex, the ventral stream arises from early visual areas ( V1þ) and projects to regions in the occipito-temporal cortex. The dorsal stream also arises from early visual areas but projects instead to the posterior parietal cortex. The posterior parietal cortex also receives visual input from the superior colliculus via the pulvinar. On the left, the approximate locations of the pathways are shown on an image of the brain. The routes indicated by the arrows involve a series of complex interconnections.
124
Perception, Action, and Consciousness
dorsal stream, projects from early visual areas to the posterior parietal cortex, a region of the brain that is reciprocally connected to motor areas in the frontal cortex and sends projections to the basal ganglia and other (older) motor nuclei in the brainstem. The dorsal stream also receives visual input from the midbrain (via the thalamus). The other projection system, the ventral stream, projects from early visual areas to inferotemporal cortex. The ventral stream has strong connections with medial temporal areas, including the amygdala and hippocampus, as well as the prefrontal cortex. As it turns out, the functions of the two streams map quite well onto the distinction between vision-for-action and vision-for-perception discussed above. Not surprisingly, given its interconnections with motor structures, it is the dorsal stream that plays the critical role in the programming and control of actions, transforming real-time information about the location and disposition of objects into the coordinate frames of the relevant motor systems. In contrast, it is the ventral stream (together with associated cognitive networks) that mediates the construction of the rich and detailed representations of the world that allow us to identify objects, events, and actions in others, attach meaning and significance to them, and infer their causal relations. In summary, processing in the ventral stream provides the conscious visual percepts that are essential for accumulating a knowledge base about the world, knowledge that we can access for cognitive operations, such as planning and decision making. Processing in the dorsal stream does not generate visual percepts; it generates skilled actions (as part of a network of structures involved in sensorimotor control). Of course, the two streams are not hermetically sealed from one another. Indeed, they work together in controlling our behavior as we live our complex lives – but they play separate and complementary roles in the production of adaptive behavior. Much of the evidence for this idea first came from work with neurological patients. The best known of these cases is the patient D.F., who developed a profound ‘visual form agnosia’ as a consequence of a hypoxic episode in which her brain was starved of oxygen. The nature of D.F.’s deficit in form vision can be understood to some extent by examining the drawings illustrated in
Figure 2. Not only was D.F. unable to identify the simple line drawings illustrated in left-hand column of this figure, but she was also unable to copy them, at least in a recognizable way. A preserved ability to see fine detail allowed her to depict some aspects of the drawings, such as the dots indicating the print in the line drawing of the open book. Nevertheless, she was unable to duplicate the overall shape or arrangement of the elements of the line drawings in her copies. D.F.’s inability to copy the drawings is not due to a problem in controlling the movements of the pen or pencil; when she was asked on a separate occasion to draw an object from memory, she was able to do so reasonably well, as the drawings on the right-hand side of Figure 2 illustrate. Needless to say, when D.F. was shown any of the drawings she had done herself, whether the ones retrieved from memory or those copied from another drawing, she had no idea what they were and commented that they all looked like ‘squiggles.’ It is important to emphasize that D.F. retains the ability to perceive and describe the colors and other surface properties of objects, such as their visual texture. The missing elements in her perceptual experience are restricted to shape and form, and
Model
Copy
Memory
Figure 2 The patient D.F.’s attempts to draw from
models and from memory. D.F. was unable to identify the line-drawings of the apple, open book, or the boat shown on the left. In addition, her copies were very poor. Note, however, that she did incorporate some elements of the line-drawing (e.g., the dots indicating the text in the book) into her copy. When she was asked on another occasion to draw these same items from memory, she produced a respectable representation of all three (right-hand column). When she was later shown her own drawings, she had no idea what they were.
Perception, Action, and Consciousness
thus her problems cannot be dismissed as a generalized inability to make perceptual reports. Remarkably, even though D.F. shows no perceptual awareness of the form or dimensions of objects, she automatically adjusts her hand to the size, shape, and orientation of an object as she reaches out to pick it up. For example, even though D.F. is unable to distinguish between rectangular blocks of different dimensions, when she reaches out to pick up one of the blocks, the aperture between her fingers and thumb is scaled in-flight to the width of the object, just as it is in people with normal vision (see Figure 3a). In other words, D.F. can scale her grip to the dimensions of an object in anticipation of picking it up, even though she is unable to ‘perceive’ the dimensions of that object. Similarly, she will rotate her hand correctly for objects placed in different orientations, and will direct her fingers to stable grasp points on the surface of the objects, even though in other tests she fails to indicate either verbally (or in a manual matching task) the orientation or shape of those same objects. She is also able to avoid other objects in the workspace as she reaches out toward a goal, even though she cannot judge their relative locations correctly in a more perceptual task. D.F. exhibits normal visuomotor control in other tasks
as well, including stepping over obstacles during locomotion, despite the fact that her perceptual judgments about the height of these obstacles are far from normal. To summarize: despite the fact that D.F. has lost all conscious perception of the form of objects, including their size, shape, and orientation, her visuomotor systems are able to make use of these same object properties to control skilled object-directed actions. So where is the damage in D.F.’s brain? Although D.F. shows some diffuse loss of tissue throughout her cerebral cortex (consistent with hypoxia), she has prominent focal lesions bilaterally in a region of the human ventral stream that has been shown to be involved in the visual recognition of objects. It is presumably the damage to these object-recognition areas that has disrupted her ability to perceive the form of objects. But clearly these ventral-stream lesions have not interfered with her ability to use visual information about form to shape her hand when she reaches out and grasp objects. The preservation of normal visually guided grasping in the face of ventral-stream damage suggests this ability is dependent on another visual pathway, the most likely candidate being the dorsal stream. This conclusion has been supported by neuroimaging evidence showing that when D.F. grasps
Patient D.F.: Visual form agnosia
Patient R.V.: Optic ataxia
110
100 90 Aperture (mm)
Aperture (mm)
100 90 80 70 0
(a)
125
80 70 60 50 40
2.5
5.0
Grasp calibration
2.5
0
5.0 cm
Manual estimation
(b)
2.5
5.0
Grasp calibration
2.5
5.0
cm
Manual estimation
Figure 3 Graphs showing the size of the aperture between the index finger and thumb during object-directed grasping and manual estimates of object width for D.F., a patient with visual form agnosia and R.V., a patient with optic ataxia. D.F. (a) showed excellent grip scaling, opening her hand wider for the 50 mm-wide object than for the 25 mm wide object. D.F.’s manual estimates of the width of the two objects, however, were grossly inaccurate and showed enormous variability from trial to trial. In contrast, R.V. (b) was able to indicate the size of the objects reasonably well (individual trials marked as open diamonds), but her maximum grip aperture in flight was not well-tuned. She simply opened her hand as wide as possible on every trial. Reproduced from Goodale MA, Milner AD, Jakobson LS, and Carey DP (1991) A neurological dissociation between perceiving objects and grasping them. Nature 349: 154–156, with permission from Nature Publishing Group.
126
Perception, Action, and Consciousness
objects that vary in size and orientation, she shows relatively normal activity in a small region of the dorsal stream that has been implicated in the visual control of grasping in healthy individuals. It should be emphasized that although patients like D.F. are rare, there are a number of other cases in the literature in which the same striking dissociation between visual perception and visuomotor control has been documented. But what about patients with damage to the dorsal stream? As it turns out, these patients exhibit a pattern of deficits and spared abilities that are complementary to that seen in patients with ventral-stream lesions. Thus, patients with dorsalstream lesions typically have problems reaching toward targets placed in different positions in the visual field, particularly the periphery. This deficit is referred to clinically as ‘optic ataxia.’ But the failure to locate an object with the hand cannot be construed as a problem in spatial representation: many optic ataxia patients, for example, can describe the relative position of the object in space quite accurately, even though they cannot direct their hand toward it. Also, the deficit is not purely motor: these patients usually have no difficulty using input from other sensory systems, such as proprioception or audition, to guide their movements. In addition to their deficits in reaching, many patients with damage in the dorsal stream are unable to use visual information to rotate their hand, scale their grip, or configure their fingers properly when reaching out to pick up objects, even though they are able to correctly report the orientation, size, and shape of those objects (see Figure 3b). In addition, they do not take into account the positions of potential obstacles when they are attempting to reach out toward goal objects even though they can indicate the relative location of the obstacles in other ways. In summary, patients with optic ataxia exhibit neither a purely visual nor a purely motor deficit, but instead a specific deficit in visuomotor control, confirming the critical role that the dorsal stream plays in the control of skilled actions. In addition to such work with neurological patients, there is a wealth of evidence from monkey neurophysiology and human neuroimaging supporting the idea of a ventral ‘perception’ stream and a dorsal ‘action’ stream.
Different Metrics and Frames of Reference for Perception and Action But why did two separate streams of visual processing evolve in the primate cerebral cortex in the first place? Or, to put it another way, why couldn’t one ‘general purpose’ visual system handle both vision-for-perception and vision-for-action? The answer to this question lies in the computational requirements of the two kinds of vision. To be able to grasp an object successfully, for example, the brain must compute the actual size of the object, and its orientation and position with respect to the grasping hand of the observer (i.e., in egocentric coordinates). The time at which these computations are performed is equally critical. Observers and goal objects rarely stay in a static relationship with one another and, as a consequence, the egocentric coordinates of a target object can often change radically from moment to moment. For these reasons, it is essential that the required coordinates for action be computed in an egocentric framework at the very moment the movements are to be performed. The computations underlying perception are quite different. Vision-for-perception does not deliver the absolute size of objects or their egocentric locations. In fact, such computations would be counterproductive for a recognition system precisely because we almost never stay fixed in one place in the world. The problem can be easily solved by the alternative strategy of encoding the size, orientation, and location of objects relative to each other. Such a scene-based frame of reference permits a perceptual representation of objects that transcends particular viewpoints, while preserving information about spatial relationships (as well as relative size and orientation) as the observer moves around. Indeed, it has been suggested that if the perceptual machinery had to deliver the real size and distance of all the objects in the visual array, the computational load would be prohibitive. The products of perception also need to be available over a much longer timescale than the visual information used in the control of action. We may need to recognize objects we have seen minutes, hours, days – or even years before. In short, the perceptual information is lodged in memory. To achieve this, the coding of the visual information has to be
Perception, Action, and Consciousness
somewhat abstract – transcending particular viewpoint and viewing conditions. By working with perceptual representations that are object- or scene-based, we are able to maintain the constancies of size, shape, color, lightness, and relative location, over time and across different viewing conditions. Although there is much debate about how this is achieved, it is clear that it is the identity of the object and its location within the scene, not its disposition with respect to the observer that is of primary concern to the perceptual system. Object recognition occurs when current perception concurs with stored information about previously encountered objects. Thus, the ventral stream provides the perceptual foundation for the off-line control of action, projecting action into the future, and incorporating stored information from the past into the control of current actions.
Perception, Action, and Consciousness As outlined in the section titled ‘Different metrics and frames of reference for perception and action,’ the ventral and dorsal streams play different but complementary roles in the service of behavior. The ventral stream (together with associated cognitive machinery) permits the brain to identify goals and plan appropriate actions; the dorsal stream (in conjunction with related circuits in premotor cortex, basal ganglia, and brainstem) programs and controls those actions. Ultimately then, both systems transform visual information into motor output. In the dorsal stream, the transformation is quite direct: visual input and motor output are essentially ‘isomorphic’ with one another. In the ventral stream, however, the transformation is indirect: the construction of a perceptual representation of the visual world enables a more ‘propositional’ relationship between input and output, taking into account previous knowledge and experience. The neuropsychological evidence from patients such as D.F. suggests that there is a close relationship between ventral-stream processing and consciousness. It is not that D.F. is simply unable to describe the form of objects; she seems to have absolutely no conscious appreciation of their
127
dimensions, shape, or orientation. For example, even though D.F.’s hand automatically conforms to the dimensions of an object when she reaches out to grab it, she cannot indicate the width of the object by using an explicit manual ‘matching’ response, using her index finger and thumb to show how wide she thinks it is. The loss of form perception is only one example of how ventralstream damage can affect visual consciousness. For example, patients with damage to other ventralstream areas can lose all conscious experience of color or visual texture. Such patients may continue to see the boundaries or edges between adjoining patches of color, even though they have no appreciation whatsoever of the colors determining those boundaries. Patients with damage to primary visual cortex, which is the major input to the ventral stream, typically report seeing nothing in the visual field contralateral to their lesion. They are cortically blind. But some of these patients show evidence of spared visuomotor control in their blind field, a phenomenon sometimes called blindsight. In other words, they can make eye movements or point to targets that they cannot consciously see, and may even show evidence of anticipatory shaping of the hand when they reach out to grasp objects placed in their blind field. It is thought that blindsight of this kind may rely on projections to phylogenetically ancient visuomotor structures in the midbrain and brainstem, which are in turn connected with the visuomotor systems in the dorsal stream. All of this evidence, as well as a wealth of other data, suggests that the ventral stream is necessary (though not sufficient) for visual consciousness. From an evolutionary perspective, the visual phenomenology that arises from ventral-stream processing must confer some kind of advantage to those that possess it. In other words, the capacity for conscious perception must give an organism an edge in natural selection. One can only speculate as to what that advantage might be. It is possible that conscious representations of the world are the only representations that can (eventually) enter our long-term visual knowledge. As already discussed, by retrieving information from long-term memory, we can manipulate this information in working memory (together with information from current percepts) for flexible ‘off-line’ control of
128
Perception, Action, and Consciousness
behavior. Moreover, when an animal is aware of what it sees, it can use this information to decide between different actions (including making no action), to plan future actions, and to communicate what it sees (or has seen) to its conspecifics. Being conscious of visual processing in the dorsal stream would confer no such advantages on the animal. Because the transformations that are carried out on visual input in the service of action involve ‘just-in-time’ computations based on the particular disposition of the goal object with respect to the actor, there would be no value and even a real cost in allowing this information to be accessible to conscious experience. Indeed, if such information were conscious, it would more often than not stand in real contradiction to the scene-based representations of the world offered by the ventral stream, which transcend particular viewpoints and thereby retain their utility over time. Far better then, that dorsal-stream computations should take place as automatically and unconsciously as those of the vestibular system in its efforts to help us maintain an upright and steady posture. In fact, there is now empirical evidence that the visual information that guides our movements is indeed unconscious. Normally, of course, we can consciously see our own arm as we reach toward a target, and we can consciously see other objects in the region of the target that could interfere with our reaching movements. But patients who show ‘extinction’ (typically following damage to the right parieto-temporal brain region) will often report not seeing a brief stimulus on the left when it is accompanied by a stimulus on the right (i.e., the presence of a stimulus on the right ‘extinguishes’ the perception of a similar stimulus on the left). Nevertheless, such patients will avoid colliding with an obstacle on the left when reaching toward a goal, even though they report seeing nothing in that part of space. Although it is the ventral stream that delivers the contents of our visual consciousness, this does not mean that the workings of the dorsal stream play no role at all in the determining our awareness. For one thing, by virtue of helping to direct our eyes (and shift our attention) between different objects and locations in the environment, the area in the dorsal stream that mediates the visual control of eye movements (and covert shifts of attention) causes changes
in the information that can be processed by the ventral stream, and thus the content of our visual awareness. The dorsal stream also participates in consciousness in other ways as well. People are typically aware of the actions they perform and, as has already been emphasized, the dorsal stream plays a critical role in the visual control of those actions. But there are some important distinctions to be made here. Although a person is often aware of the visual stimulus to which their action is directed, that visual awareness is mediated by processing in the ventral, not the dorsal, stream. The actual visual information used by the dorsal stream to specify and control the constituent movements of a goal-directed action (including an eye movement) is inaccessible to consciousness. Yet the compelling nature of visual consciousness makes it difficult to resist the intuition that it is one’s perception of the goal object that is guiding one’s action. The neuropsychological studies of D.F. and other patients provides compelling evidence for the idea that ventral-stream processing is necessary for visual awareness. There is now more direct evidence, however, for the role of the ventral stream in giving rise to visual awareness, deriving from single-unit and lesion studies in the monkey, and neuroimaging studies in humans. Although human subjects can readily report verbally on what they see, monkeys have to make their reports in other ways, having first been trained to do so. Monkeys, for example, can be trained in a detection task to press a touch screen to indicate whether or not a stimulus was presented somewhere on the screen. What is interesting is that even though monkeys with unilateral removal of primary visual cortex can point reasonably accurately toward a visual stimulus presented in their blind field, they will categorize the same stimulus as a blank (stimulus-absent) trial when faced with it in such a detection task. This result provides a compelling confirmation of the reality of blindsight, showing that visuomotor control can remain virtually intact despite the absence of visual perception. This dissociation provides strong empirical support for the thesis that visual perception depends on inputs to the ventral stream from primary visual cortex. But perhaps the most convincing evidence for the relationship between awareness of a visual
Perception, Action, and Consciousness
stimulus and neural activity in the ventral stream has come from studies using the phenomenon of binocular rivalry, in which different visual stimuli (say a face and a cloudburst pattern) are simultaneously and independently presented to the left and right eye. When human observers are presented with such stimuli, they typically report fluctuations in what they see, sometimes reporting seeing a face and sometime a cloudburst pattern. Only very rarely do they see a ‘blend’ of the two stimuli. This technique has been used with monkeys by training them to make a manual response to signal a switch between seeing stimulus A (the cloudburst pattern) and stimulus B (the face). In a sense, the monkey was being asked to report on the content of its visual experience. As it turns out, single neurons in the inferior temporal cortex, a higher-order area in the ventral stream, that were tuned to one stimulus (say the face) fired more rapidly when the monkey reported seeing that particular stimulus. This correlation between perceptual report and firing rate was not nearly so strong for neurons in earlier visual areas and was completely absent in primary visual cortex. These striking results demonstrated for the first time that there is a direct link between the activity of particular neurons in the ventral stream and what an animal perceives. More recent research with humans shows exactly the same thing. It has now become possible to record from single neurons in humans, typically in patients with epilepsy who have had electrodes implanted in their brains to localize the site where the seizure originates. Using a variant of the binocular rivalry paradigm, researchers have shown that visual neurons in the medial temporal lobe respond only to the perceived stimulus and never respond to the suppressed stimulus presented to the other eye. Similar effects have been observed in functional magnetic resonance imaging (fMRI) studies. For example, when volunteers in the fMRI scanner are presented with rivalrous images of a face and a building to the left and right eye, the activity in two areas in the ventral stream that are differentially selective for faces and scenes fluctuate in a reciprocal fashion with what the volunteers report seeing. Thus, when they report seeing a face, the ‘face area’ is more active and when they report seeing the building, the ‘place’ area is more active. The relationship between visual awareness
129
and brain activity is not seen in the dorsal stream; even when observers are unaware of visual stimuli because of interocular suppression, the fMRI activation elicited by those stimuli is just as great as when those stimuli are consciously perceived. Although all these studies (and many others) provide compelling evidence that visual consciousness is closely correlated with activity in the ventral stream, this does not mean that we are conscious of everything that is processed by this system. If this were the case, we would be overwhelmed with information, and consciousness could serve no function. In fact, even in the binocular rivalry experiments, activity was never completely abolished when the monkeys or humans reported not seeing the stimulus – it was just reduced. This may suggest that active inhibitory and/or facilitatory mechanisms are at work – and that for visual awareness to occur activity must exceed some sort of threshold. The neural origins of these inhibitory and/or facilitatory mechanisms are unknown, although regions in prefrontal cortex, medial temporal cortex, and the inferior parietal lobule (regions associated with the switching of visual attention) may play a significant role. But even though ventral-stream areas are subject to such modulatory effects from other brain regions, it is ventral-stream activity itself that determines the content of our visual awareness.
Perception, Action, and Illusions Although much of the evidence for the neural organization of the two visual systems model has come from human neuropsychology and neuroimaging, as well as from work with nonhuman primates, evidence for fundamental differences in the metrics and frames of reference used by vision-forperception and vision-for-action has also come from studies in normal human observers. These latter studies have also provided compelling evidence with respect to the dissociation between conscious visual experience and more ‘automatic’ visuomotor control. One of the most striking examples of dissociations between vision-for-perception and visionfor-action in normal observers has come from work with pictorial illusions. For example, when
130
Perception, Action, and Consciousness
people are asked to pick up a target in the context of a size-contrast illusion, such as the Ebbinghaus Illusion (see Figure 4), their grip aperture is typically scaled in flight to the real not the apparent size of the target. Although grip scaling escapes the influence of the illusion, the illusion does affect performance in a manual matching task in which people are asked to open their index finger and thumb to indicate the perceived size of a disk. Thus, the aperture between the finger and thumb is resistant to the illusion when the vision-foraction system is engaged (i.e., when the participant grasps the target) and sensitive to the illusion when the vision-for-perception system is engaged (i.e., when the participant estimates its size). This dissociation between what people say they see and what they do underscores once more the differences between vision-for-perception and vision-for-action. The obligatory size-contrast effects that give rise to the illusion (in which
different elements of the array are compared) normally play a crucial role in scene interpretation and object identification, a central function of vision-for-perception. In contrast, the execution of a goal-directed act, such as manual prehension, requires metrical computations that are centered on the target itself, rather than on the relations between the target and other elements in the scene. In fact, if our visually guided movements were based on the relative rather than the absolute size of objects, then many of our everyday actions, from driving a car to picking up a wine glass, would be subject to critical errors. As it turns out, the true size of a target (for grasping at least) can be computed from the retinal-image size of the object coupled with an accurate estimate of distance based on reliable cues such as vergence of the eyes. Such computations would be quite insensitive to the kinds of pictorial cues that distort perception when familiar illusions are presented.
(c) Perceptually different physically identical
Large disk Small disk
(a) Aperture (mm)
80
Perceptually identical physically different
(b)
60
40 20 1.0
2.0 Seconds
(d)
Figure 4 The effect of a size-contrast illusion on perception and action. (a) The traditional Ebbinghaus illusion in which the central circle in the annulus of larger circles is typically seen as smaller than the central circle in the annulus of smaller circles, even though both central circles are actually the same size. (b) The same display, except that the central circle in the annulus of larger circles has been made slightly larger. As a consequence, the two central circles now appear to be the same size. (c) A 3-D version of the Ebbinghaus illusion. People are instructed to pick up one of the two 3-D disks placed either on the display shown in panel A or the display shown in panel B. (d) Two trials with the display shown in panel B, in which the same person picked up the small disk on one trial and the large disk on another. Even though the two central disks were perceived as being the same size, the grip aperture in flight reflected the real not the apparent size of the disks. Reproduced from Aglioti S, DeSouza JFX, and Goodale MA (1995) Size-contrast illusions deceive the eye but not the hand. Current Biology 5: 679–685, with permission from Elsevier.
131
Perception, Action, and Consciousness
(a) Forehead
Perceptual task
Flicking task
Cheek
Types of face
There has been considerable debate in the literature about whether or not grasping and other visuomotor responses (such as saccadic eye movements) are refractory to pictorial illusions. In some instances, particularly with unpracticed or awkward movements (where there is a lot of cognitive supervision), grip scaling appears to fall victim to size-contrast illusions like the Ebbinghaus. But of course the fact that actions are sensitive to illusory displays under certain conditions can never by itself refute the idea of two visual systems, which is securely based on a much larger body of evidence ranging from neuroimaging to neurophysiology. Indeed, it is unsurprising that perception affects our motor behavior, even within the context of the two-visual-systems model. After all, there are a number of situations, such as picking up a hammer or a cup of coffee, where our perception of the goal object will determine the kind of grip posture we adopt. The real surprise (at least for monolithic accounts of vision) is that there are a number of situations where visually guided action is unaffected by pictorial illusions. The actions that fall into this category tend to be rapid skilled actions, usually with the preferred hand. Nevertheless, it is fair to say that the claim that actions can be resistant to pictorial illusions is still regarded as controversial, particularly amongst those who favor a more monolithic account of visual processing. Most of the pictorial illusions that have been used to dissociate vision-for-perception and vision-foraction distort perceived size by only a few millimeters. There are other illusions, however, which are not only much larger but which also show an actual reversal of depth. One particularly striking example is the hollow face illusion shown in Figure 5. In this illusion, knowledge about what faces look like impels observers to see the inside of a mask as if it were a normal protruding face, and the illusory face is perceived to be located several centimeters in front of the actual surface of the hollow mask. Despite the fact that observers cannot resist this compelling illusion, when they are asked to reach out quickly and flick off a small bug-like target stuck on the face, they unhesitatingly reach to the correct point in space (i.e., inside the mask). In other words, despite the presence of a strong hollow-face illusion, people direct rapid movements to the real, not the illusory positions of the targets. To
Illusory
Normal
−6 −4 −2 Inside
(b)
Illusory
Normal
0
2
4
Outside
Distance (cm)
6
−6 −4 −2 Inside
0
2
4
6
Outside
Distance (cm)
Figure 5 Perceptual judgments and visuomotor control with the hollow-face illusion. (a) A small magnet was placed on either the cheek or forehead of the normal face (left) or the hollow mask (right). Participants were required either to flick the magnet from the normal or illusory (actually hollow) face or to estimate its distance psychophysically. Inset shows a photograph of bottomlit hollow face, in which the illusion of a normal convex face is evident. (b) (Left) The mean psychophysical (perceptual) judgments of the apparent position of the magnets on the illusory and normal face with respect to the reference plate from which the two displays either protruded or receded. Note that participants perceived the hollow face as protruding forward like the normal face. (Right) The mean distance of the hand at the moment the participant attempted to flick the target off the cheek or forehead of the illusory (actually hollow) or the normal face. In the case of the illusory face, the end points of the flicking movements corresponded to the actual distances of the targets, not to consciously seen distances. Error bars indicate the standard error of the mean. Reproduced from Kro´liczak G, Heard P, Goodale MA, and Gregory RL (2006) Dissociation of perception and action unmasked by the hollow-face illusion. Brain Research 1080: 9–16, with permission from Elsevier.
do this, the visuomotor system must have access to a different source of visual information from that driving the illusion. In either case, people seem to be unaware of the veridical depth information they are using to control their flicking movements – and furthermore the use of this information does not ‘break’ the illusion. Again, this provides compelling evidence that one’s conscious perception of a visual
132
Perception, Action, and Consciousness
Summary
One is immediate and direct. For example, visual information about the size, shape, and disposition of an object with respect to the observer can be automatically transformed into the required egocentric coordinates for the programming and online control of a smoothly executed grasping movement. This kind of visual guidance, which is mediated by the dorsal stream of visual projections, needs to be quick and accurate, and evolution has ensured that it is. The visual information used by the dorsal stream is not accessible to consciousness – even though the actions controlled by that information clearly are. The other way in which vision can influence behavior is much less direct, and depends upon the construction and storage of visual representations that are initially processed in the ventral stream and reflect the structure and semantics of the scene facing the observer. The nature and intentions of subsequent actions will to varying degrees depend on the retrieval, and mental manipulation of, these representations. It is these representations that make up the visual contents of consciousness. The division of labor between the two streams is associated with fundamental differences in the metrics and frames of reference used by vision-for-action and visionfor-perception. Although both streams process information about the structure of objects and about their spatial locations, they use quite different modi operandi to do this. The operations carried out by the ventral stream use scene-based frames of reference and relational metrics; those carried out by the dorsal stream use egocentric frames of reference and absolute metrics. The two streams work together in the production of goal-directed behavior. The ventral stream (together with associated cognitive machinery) identifies goals and plans appropriate actions; the dorsal stream (in conjunction with related circuits in premotor cortex, basal ganglia, and brainstem) programs and controls those actions. Thus, a full understanding of the integrated nature of visually guided behavior will require that we specify the nature of the interactions and information exchange that occurs between the two streams of visual processing.
There appear to be two ways in which visual information can influence and guide behavior.
See also: The Neural Basis of Perceptual Awareness; Neuroscience of Volition and Action.
stimulus does not control visuomotor responses directed at that same stimulus.
Time and the Two Streams As was discussed earlier, the visuomotor systems in the dorsal stream works in real time, using information provided in a ‘bottom up’ fashion from the retina. Thus, movements directed to remembered objects (objects that were present, but are no longer visible) might be expected to differ from movements directed to objects in real time. In fact, this is exactly the case. For example, when people reach out to grasp objects that were visible only a few seconds earlier, their grip scaling is now susceptible to pictorial illusions that were present in the display. Such sensitivity is to be expected, of course, if the programming and control of a delayed grasping movement is dependent not on processing in the dorsal stream but on memories derived from perceptual processing in the ventral stream. This conclusion is supported by the fact that the visual agnosic patient, D.F., who has ventral-stream lesions, demonstrates extremely poor size scaling of her grip when she attempts to grasp a target object after it has been removed from view – even though she shows excellent real-time control of her grasping. In one experiment, when a 2 s delay was introduced between viewing the object and initiating the grasp, there was no correlation at all between the size of the object and the aperture of her grasp in flight. An even more surprising result has been obtained in experiments with patients with lesions of the dorsal stream. Even though these individuals have great difficulty scaling their grasp when reaching out to grasp visible objects immediately after the objects are presented, they show a paradoxical improvement in performance if they are required to wait for 5 s before initiating their movement. Again these findings support the idea that the programming and control of delayed actions depends on information derived from earlier perceptual processing in the ventral stream.
Perception, Action, and Consciousness
Suggested Readings Bruno N, Bernardis P, and Gentilucci M (2008) Visually guided pointing, the Mu¨ller-Lyer illusion, and the functional interpretation of the dorsal-ventral split: Conclusions from 33 independent studies. Neuroscience and Biobehavioral Reviews 32: 423–437. Carey DP (2001) Do action systems resist visual illusions? Trends in Cognitive Sciences 5: 109–113. Goodale MA (2008) Action without perception in human vision. Cognitive Neuropsychology 25: 891–919. Goodale MA and Milner AD (2004) Sight Unseen: An Exploration of Conscious and Unconscious Vision. Oxford: Oxford University Press.
133
Logothetis NK (1998) Single units and conscious vision. Philosophical Transactions of the Royal Society of London. B Biological Sciences 353: 1801–1818. Milner AD (2008) Visual awareness and human action. In: Weiskrantz L and Davies M (eds.) Frontiers in Consciousness Research, pp. 169–214. Oxford: Oxford University Press. Milner AD and Goodale MA (2006) The Visual Brain in Action. 2nd edn. Oxford: Oxford University Press. Milner AD and Goodale MA (2008) Two visual systems re-viewed. Neuropsychologia 46: 774–785.
Biographical Sketch
Professor Goodale was born in England but was educated entirely in Canada. After completing his PhD in psychology at the University of Western Ontario, he returned to the United Kingdom where he worked as a postdoctoral fellow at the University of Oxford with Professor Larry Weiskrantz. After two years at Oxford, Professor Goodale accepted a position in the School of Psychology at the University of St Andrews before returning to Canada in 1977, where he now holds the Canada Research Chair in visual neuroscience at the University of Western Ontario. Professor Goodale is best known for his work on the functional organization of the visual pathways in the cerebral cortex, and was a pioneer in the study of visuomotor control in neurological patients. His recent research uses functional magnetic resonance imaging (fMRI) to look at the activity in the normal human brain as it performs different kinds of visual tasks. Professor Goodale serves on the editorial board of a number of journals including Experimental Brain Research and Neuropsychologia, and is the past-President of the Association for the Scientific Study of Consciousness. He has been recognized for his distinguished scientific achievements by the Canadian Society for Brain, Behaviour, and Cognitive Science, receiving the D.O. Hebb Award in 1999. He was elected as a fellow of the Royal Society of Canada in 2001.
134
Perception, Action, and Consciousness
David Milner was born in Leeds, England. After taking his first degree at the University of Oxford, he undertook training in clinical psychology, and then a PhD in behavioral neuroscience, both at the Institute of Psychiatry at the University of London. In 1970, he moved to an academic position at the University of St Andrews, Scotland. It was there that Dr. Milner first collaborated with Dr. Goodale – at that time on research into midbrain visual function in animals. In 1988, he first came across the now-famous patient D.F., who had just suffered catastrophic brain damage due to carbon monoxide poisoning. It was the remarkable early results from testing D.F. that stimulated the development of the idea that the two main visual pathways in the primate brain might play complementary roles in visual perception and visuomotor control, respectively. These ideas were presented in detail in the monograph The Visual Brain in Action (Oxford University Press, 1995). In parallel with this work Dr. Milner has been carrying out a program of research into the clinical condition of visuospatial neglect over the past 20 years. David Milner was elected in a fellow of the Royal Society of Edinburgh in 1992, and served as Editor-in-chief of the journal Neuropsychologia from 2000 to 2005. He has been director of the cognitive Neuroscience Research unit at Durham University, England, since 2000.
Perception: Subliminal and Implicit M Snodgrass, University of Michigan Medical Center, Ann Arbor, MI, USA E S Winer, University of Illinois, Chicago, IL, USA N Kalaida, University of Michigan Medical Center, Ann Arbor, MI, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Conscious perception index – Task used to measure conscious perceptual influences; usually direct. Direct task – Requires discriminative judgments (e.g., identification) about relevant aspects of ostensibly unconscious stimuli themselves. Exclusiveness problem – Apparent requirement that objective threshold approaches make strong assumption that direct measures are insensitive to unconscious influences; else, unconscious effects at objective thresholds seem intrinsically weak or impossible. Exhaustiveness problem – Difficulty of ensuring that direct tasks actually assess all conscious perception relevant to putatively unconscious perceptual effects. Indirect task – Either assesses influences of ostensibly unconscious stimuli on other stimuli (e.g., priming effects) or on judgments regarding irrelevant aspects of the unconscious stimulus itself (e.g., stimulus preference when only prior exposure was manipulated). Null sensitivity problem – Difficulty of ensuring that all relevant conscious perception has indeed been eliminated; compounded by direct task measurement error. Objective threshold – Operationalizes consciousness as stimulus intensity where direct task performance is at chance; more stringent than subjective thresholds. Subjective threshold – Operationalizes consciousness as stimulus intensity where observers deny awareness; this occurs while direct performance still exceeds chance.
Unconscious perception index – Task used to measure unconscious perceptual influences; usually indirect, but not necessarily so.
Introduction Implicit and/or subliminal perception paradigms attempt to study the unconscious effects of sensory stimuli (usually visual), utilizing exposure conditions that preclude their conscious representation. The goal is to prevent the occurrence of any potentially confounding conscious perceptual influences altogether, so that any obtained effects can be said with confidence to reflect purely unconscious perceptual processes. In typical unconscious perception paradigms, two tasks are used: one intended to index conscious perceptual influences and a second to index unconscious influences. The conscious perception index is usually a direct discrimination task such as stimulus detection (e.g., ‘‘Has stimulus X been presented or not?’’) or identification (e.g., ‘‘Was stimulus X or Y presented?’’). Presumably, then, if stimuli are consciously perceivable, observers should perform above chance on such tasks, whereas demonstrating chance performance suggests the absence of any relevant conscious perception. Because these tasks straightforwardly request intentional reports on stimulus dimensions relevant to ostensibly unconscious effects, they are generally regarded as the most sensitive conscious perception indexes. Conversely, the unconscious perception index is frequently an indirect task such as priming, wherein the unintended influence of an initial stimulus on the processing of a later stimulus is examined. Because unconscious influences are frequently thought to not require conscious intention to 135
136
Perception: Subliminal and Implicit
manifest, indirect tasks are generally regarded as the best unconscious perception indexes. The usual objective is to obtain indirect effects despite chance performance on the direct task, thereby demonstrating unconscious perceptual influences on the former. If successful, unconscious perception paradigms can not only inform us about the nature and scope of unconscious mental processes, but moreover have fundamental implications for the nature of consciousness itself. For example, it is widely held that more complex and flexible mental processes are possible consciously than unconsciously. By investigating what happens when stimuli are consciously versus unconsciously perceived, we can empirically examine these and other crucial questions. To clearly establish the presence versus absence of consciousness turns out to be much harder than it might initially seem, however. In particular, it is surprisingly difficult to convincingly demonstrate that putatively unconscious effects are not, instead, merely weakly conscious effects in disguise. This fundamental problem is not limited to unconscious perception; indeed, it continues to bedevil all fields that attempt to distinguish conscious versus unconscious influences (e.g., implicit/unconscious memory and/or learning). Now, given that unconscious perception is currently widely accepted, one could easily assume that this core methodological dilemma had somehow been settled long ago. Despite the apparently sanguine state of the field, however, this is clearly not the case – which is particularly unfortunate, given that the 120-year plus history of unconscious perception essentially consists of cyclically alternating periods of skepticism versus acceptance. Accordingly, the current positive consensus is superficial and runs the risk of simply continuing the boom-and-bust cycle unless further progress can be made on these foundational issues. Perhaps even more importantly, careful consideration of the nagging problem of how to distinguish conscious and unconscious influences is not only methodologically essential but fundamentally informative in its own right, because it forces us to consider fundamental issues such as how to define consciousness, how conscious and unconscious processes interact, and the role of conscious control and intention. Indeed, there
remains essentially no consensus on these essential issues, even among unconscious perception proponents, let alone remaining skeptics. Moreover, these continuing disagreements have vital implications regarding which data are valid and how they should be interpreted. Accordingly, great caution is advisable when assessing virtually all unconscious perception research, as dramatically different conclusions would ensue depending on exactly how their ostensibly unconscious status is understood.
How Should Consciousness Be Defined? It is generally agreed that conscious perception covaries with stimulus intensity, usually manipulated by varying either stimulus exposure duration, the intensity of masking (initially presented stimuli become less consciously perceivable when immediately followed by a second, masking, stimulus), or both. Not surprisingly, strong stimuli are clearly visible, whereas conscious perception diminishes as stimulus strength is reduced. Finally, conscious perception appears to disappear entirely when stimuli are weak enough, and the point where this occurs is taken as the threshold for consciousness. But how exactly should this point be operationalized? Notably, there are two basic alternatives. Subjective Threshold Methods Perhaps the most straightforward approach is to simply ask people whether or not they can see the stimuli. These methods, derived from classical psychophysics, rely on self-reports of phenomenal states (i.e., subjective experience), and are now usually called subjective threshold methods. Such methods date from the very beginning of scientific psychology, and from their inception have yielded a striking effect: Performance on direct forcedchoice discrimination tasks (e.g., choosing which letter is on a distant card; selecting which of two weights is heavier) remains above chance, even when observers convincingly deny any relevant conscious experience. These are canonical subjective threshold effects, and reflects how they were originally discovered in the late nineteenth century. At the same time, such findings illustrate that unconscious perception indices can also be direct
Perception: Subliminal and Implicit
tasks, even though indirect tasks are typically used for this purpose. Subjective threshold methods continued to flourish throughout the 1940s and 1950s, particularly in so-called New Look paradigms, which often combined now-unfashionable psychoanalytic hypotheses with experimental psychological methods. For example, it was shown that taboo (i.e., socially unacceptable, at least at that time) words had higher perceptual thresholds than less objectionable words, apparently demonstrating a simple form of unconscious defense. Strikingly, however, virtually all unconscious perception research essentially halted as modern psychophysics emerged with the development of signal detection theory (SDT). Although psychophysicists had long known that self-reports might not exclusively reflect observers’ actual perceptions, but were perhaps also influenced by their willingness to respond in certain ways as well, there had heretofore been no clear method to disentangle these processes. Crucially, SDT finally succeeded in rigorously demonstrating that observers’ discrimination reports did not, as it might intuitively seem, simply convey the contents of their conscious perceptual states, but instead were indeed the joint product of independent, separable perceptual and decision processes. With this fundamental insight in mind, it became clear that subjective thresholds might not demarcate some qualitative boundary between conscious and unconscious perception, but perhaps instead reflect differing levels of confidence associated with varying degrees of conscious perception. In other words, denying awareness might simply indicate very low confidence, not the complete absence of awareness. After all, very weak stimuli clearly produce remarkably faint, indistinct, and unclear conscious perceptual experiences that, not surprisingly, individuals are quite uncertain about. Accordingly, SDT demonstrated that subjective threshold phenomena might well be simply very weakly conscious after all. With taboo words, for example, people might simply want to be a little more sure before reporting salacious or profane words than more innocuous material. In the language of SDT, then, this apparently dramatic effect could simply reflect different response criteria, rather than actual differences in how conscious the stimuli were.
137
It is important to emphasize that, perhaps contrary to initial appearances, the SDT criterion artifact critique does not somehow contradict or deny observers’ subjective reports. Rather, because SDT demonstrates that all reports are influenced by both perceptual and response criteria, alternative explanations in terms of the latter become straightforwardly possible. For example, even when neutral or innocuous stimuli are used, and hence there are no extraneous (e.g., embarrassment-related) influences on observers’ response criteria, a very weak conscious perception will plausibly resemble no stimulus at all much more closely than it does some positive response option. This could easily produce denials of awareness on a subjective detection task, but yet could clearly arise from weak, belowcriterion conscious perception. Taken together with the classic above-chance discrimination performance characteristic of subjective threshold effects, the SDT criterion artifact critique convinced most investigators that subjective threshold approaches were likely simply invalid, and unconscious perception research all but ceased for the next 15 years. Objective Threshold Methods With the advent of SDTand the clear problems with criterion-based methods, the revival of unconscious perception research in the mid-1970s and early 1980s instead arranged stimulus conditions such that observers not only denied awareness, but moreover could not perform above chance on direct discrimination tasks. Because such performance is behaviorally observable, these are called objective thresholds. Further, in contrast to subjective threshold approaches, requiring chance performance on direct tasks appeared to render them unusable as unconscious perception indexes. Accordingly, objective threshold paradigms rely heavily on indirect tasks (usually priming) to index unconscious perceptual influences. Notably, to arrange objective threshold conditions requires reducing stimulus intensity even more than when subjective thresholds are used; hence, objective thresholds are more rigorous. Unfortunately, this greater rigor seemed to yield much more mixed findings than those obtained using subjective threshold methods, leading many to conclude that genuine objective threshold results did not exist at all.
138
Perception: Subliminal and Implicit
In the late 1980s, further important methodological criticisms emerged. For example, it is very difficult to ensure that true discrimination performance does not exceed chance (the null sensitivity problem), because any obtained performance value inevitably contains measurement error. Accordingly, even when observed performance is at chance, the true underlying state of affairs might reflect some small, but nonetheless genuine, smidgen of conscious perception. To make matters even more difficult, it is essential that the particular conscious perception index in question indeed taps whatever conscious perception that could conceivably be responsible for the ostensibly unconscious effects (the exhaustiveness problem). In other words, the conscious perception index must validly assess the right type or kind of conscious perception; otherwise, demonstrating zero discrimination ability, even if it were possible, is useless. Unless both these problems are resolved, then, it is still conceivable that weakly conscious perceptual processes could still account for putatively unconscious effects, even using ostensibly more rigorous objective threshold methods. Importantly, however, these difficulties are just as problematic for subjective threshold methods. After all, they too must utilize exhaustively sensitive conscious perception indices. Moreover, as the SDT criterion artifact critique suggests, observers in subjective threshold paradigms might actually possess small but nonzero confidence, despite overall denials of awareness. Here again, then, true conscious perception might actually exceed zero, despite subjective reports. Quite recently, subjective methods have begun to take this problem much more seriously, and have increasingly turned to Type II SDT methods as a possible solution. In contrast to more typical Type I SDT tasks, wherein observers discriminate stimulus states in the external world (e.g., whether or not a stimulus has been presented), in Type II tasks observers make a further discrimination which arguably concerns their internal phenomenal state – namely, how confident they are in the correctness of their response. Although Type II methods are important, a deeper grasp of them requires technical analysis beyond the scope of this article.
So Which Methods Are Valid? It seems that unconscious perception researchers face a Goldilocks dilemma: Perhaps subjective methods are too lax, whereas objective methods may be too stringent. Even more importantly, however, a closer look at subjective and objective methods suggests that they embody fundamentally different models of unconscious perceptual influences. Recall, for example, that the canonical subjective threshold effect is above chance performance on direct discrimination tasks, even when awareness is denied. Obviously, in subjective models, this above chance performance must be due to unconscious perceptual influences. But if this is so, requiring that direct discriminations not exceed chance, as objective methods do, would apparently reduce or even completely eliminate all influences – not only conscious, but unconscious too. Subjective threshold models inherently predict, then, that objective threshold effects should simply not occur or, at best, be weaker and less interesting versions of their more robust subjective brethren. In contrast, for objective threshold methods to be viable, they must apparently assume that direct discrimination tasks exclusively tap only conscious, not unconscious, perceptual influences – and, in turn, that unconscious perceptual influences manifest more freely on indirect methods such as priming. If so, this would apparently indicate that direct tasks are indeed not influenced by unconscious perceptual processes. This would, in turn, crucially imply that subjective threshold effects on direct tasks must actually be due to conscious influences after all, and hence that subjective methods are invalid. Consequently, subjective and objective methods are not just more or less stringent, as many assume, but actually imply fundamentally incompatible models, and hence cannot both be valid.
Modern Unconscious Perception Models Given the above, any adequate unconscious perception model must not only address these formidable methodological difficulties, but deal meaningfully with their theoretical implications
Perception: Subliminal and Implicit
degree and complexity of conscious perceptual processes are positively related to stimulus intensity. See Figure 1(a). This alternative null hypothesis model is kept in mind as the three recently proposed unconscious perception models are described.
Unconscious perception index (e.g., semantic priming)
SIT
Supraliminal
Conscious perception index
(a)
0
ODT (c)
OIT
OIT
SIT
Supraliminal
Conscious perception index
As with all subjective threshold approaches, this model holds that subjective effects are indeed
Unconscious perception index (e.g., semantic priming)
ODT
Philip Merikle and Associates’ Subjective Threshold Model
Unconscious/ conscious boundary
Unconscious
(d)
Conscious
0 ODT
(b)
Unconscious perception index (e.g., semantic priming)
Unconscious perception index (e.g., semantic priming)
as well. To date, three major attempts have been made to address these issues. Before considering these modern unconscious perception models, however, it is helpful to recall that any such model must, as its ultimate objective, falsify the null hypothesis, that is, the single process conscious-perceptiononly model. Moreover, because this model is simpler, it is preferable unless clearly refuted. Accordingly, greater specification of this model is important. Fortunately, certain properties of the conscious-perception-only model are widely agreed upon – namely, as noted earlier, that the
0
139
OIT
SIT
Supraliminal
Conscious perception index
0
ODT
OIT
SIT
Supraliminal
Conscious perception index
Figure 1 (a) The single-process conscious perception model’s monotonic prediction for the relationship between the
conscious and unconscious perception indices. Unconscious index performance remains flat and does not exceed zero until both the objective detection (ODT) and objective identification (OIT) are surpassed; then, the relationship becomes positive; (b) The subjective threshold model’s monotonic prediction for the conscious versus unconscious perception index relationship. Although identical to the single-process conscious perception model’s prediction, the subjective threshold model nonetheless holds that stimuli below the subjective identification threshold (SIT) are unconscious, whereas stimuli above this threshold are conscious; (c) The objective threshold/rapid decay model’s monotonic prediction for the conscious versus unconscious perception index relationship. Here, unconscious index performance exceeds zero but remains flat throughout the entire stimulus intensity range, provided very rapid responses are obtained; (d) The objective threshold/nonmonotonic model’s prediction for the conscious versus unconscious perception index relationship. This relationship is negative as stimulus intensity increases from the ODT to the OIT, reaching zero at the OIT. Beyond the OIT, the relationship becomes positive. Adapted by permission from Snodgrass M, Bernat E, and Shevrin H (2004) Unconscious perception: A model-based approach to method and evidence. Perception & Psychophysics 66: 846–867. Original: Figure 3, p. 856.
140
Perception: Subliminal and Implicit
unconscious, and that alleged objective threshold effects are ultimately artifactual – thus explaining the latter’s apparent empirical unreliability. The very considerable challenge, then, for this model is to overcome the SDT criterion artifact critique and convincingly rule out alternative weakly conscious explanations in some more definitive fashion. This is particularly difficult because, similar to conscious perceptual influences, subjective threshold effects are positively related to stimulus intensity (see Figure 1(b)). To meet this burden, these researchers have persuasively argued that subjective (and, for that matter, objective) threshold models must additionally demonstrate qualitative differences between putatively unconscious and conscious perceptual processes to provide essential convergent evidence for the claimed distinction. Further, requiring demonstration of qualitative differences seems especially reasonable given that they presumably do differ in important ways. After all, if not, why bother to distinguish them at all? Qualitative differences in subjective paradigms. Notably, a variety of qualitative differences between subjective threshold and clearly conscious stimuli have indeed been reliably demonstrated. These follow a central and theoretically important theme – namely, illustrating various ways in which consciously controlled strategies are indeed applied when stimuli are clearly consciously perceived, but not when subjectively unconscious. Perhaps most compellingly, it has been repeatedly demonstrated that subjective threshold stimuli produce exclusion failure (i.e., responding with just-presented stimuli despite being instructed not to), whereas observers have little difficulty successfully excluding responding with stimuli that are clearly consciously perceived. This qualitative difference is particularly convincing, given the widely popular belief that complex cognitive processes, such as selective control over responding, can only occur when stimuli are consciously perceived. Despite its intuitive appeal, however, exclusion failure and its cognates can also be criticized as a possible SDT criterion artifact. For example, exclusion is plausibly a criterion-dependent decision process; that is, observers may wish to be reasonably confident that they perceived the initial
stimulus in order to exclude it as a response. If so, just as in the standard criterion artifact account, exclusion failure could simply reflect low confidence rather than genuine unconscious perceptual processes. Accordingly, these undeniably impressive qualitative differences might simply indicate two plausibly conscious processes, that is, conscious perception proper and criterion-dependent strategic exclusion processes. If the SDT-based critique of exclusion-related paradigms is correct, subjective threshold models need to demonstrate additional, more definitive qualitative differences; otherwise, the single-process conscious perception model is preferable. Perhaps in part due to their strong intuitive appeal, however, exclusion-related qualitative differences are still currently widely accepted. Further, this alternative account illustrates a little-noticed but important weakness of the otherwise appealing qualitative differences approach: Demonstrating qualitative differences alone does not demonstrate that one or the other process is unconscious, but rather merely that they are distinct. With this in mind, it is probably necessary to independently demonstrate that the relevant stimuli are in fact unconsciously perceived in order for qualitative differences to truly carry any inferential weight. Despite this underappreciated problem, however, the qualitative differences approach is important and generative, and has proved to be widely influential. Indeed, in some fundamental sense it is necessary, because such differences must exist in some important sense for conscious and unconscious processes to really be distinct at all. Blindsight and subjective threshold models. Blindsight is a striking neuropsychological syndrome in which rare individuals suffering from certain forms of brain damage perform quite well on simple direct discrimination tasks (e.g., position discrimination) despite vigorous denials of visual awareness, thus yielding a particularly dramatic subjective threshold effect. Further, blindsight effects differ qualitatively from what simply very weak conscious perception would produce. On the other hand, confidence ratings in blindsighted observers predict performance, in line with weak conscious accounts. Overall, the status of blindsight is unclear.
Perception: Subliminal and Implicit
Although many researchers seem convinced that blindsight is genuinely unconscious, they may instead be qualitative different but nonetheless weakly conscious. After all, given the brain damage intrinsic to blindsight, one would expect any residual visual capacities to qualitatively differ – whether conscious or unconscious. Anthony Greenwald and Associates’ Objective Threshold/Rapid Decay Model Apparently convinced by the SDT criterion artifact critique, these researchers assume that subjective threshold effects are indeed weakly conscious and hence that more stringent objective threshold approaches are necessary. They further posit that objective threshold phenomena are inherently very short lived; if so, this could explain why objective paradigms seemingly produce mixed results, alleviating this important concern. Moreover, if objective effects are indeed intrinsically very short lived, then their reliable measurement requires paradigms that engender very rapid responding – such as Greenwald and associates’ response window procedure, typically implemented in conjunction with response conflict priming paradigms. Here, there are typically two stimulus categories, and the primes and targets either match or mismatch on each trial. Using such approaches, large and readily repeatable objective threshold effects have been obtained, apparently overcoming their alleged unreliability and supporting the hypothesis that such effects are indeed quite short lived. Strikingly, in these paradigms, typically no relationship is found between the direct and indirect measures (see Figure 1(c)). This pattern, which clearly differs from the single-process conscious perception model’s (cf. Figure 1(a)), may indicate that conscious and unconscious perceptual influences are not only distinct but essentially independent. This conclusion, however, requires strong assumptions that the conscious and unconscious perception indexes are exclusively sensitive to conscious and unconscious perceptual influences, respectively. Alternatively, perhaps only conscious influences are involved, and the lack of relationship between the two indexes might result simply from severe time pressure on
141
one task but not the other – thus accounting for the lack of a relationship. Notably, response window effects seem largely primitive and unsophisticated (i.e., nonsemantic), instead largely driven by part-word analysis. This feature may not be intrinsic to unconscious perceptual processes, however, but instead possibly result from the severely limited processing time imposed by the response window procedure itself. Further, because part-word effects could be caused by partial stimulus identification, it may be that classification tasks (the usual conscious perception index in these paradigms, e.g., ‘‘Was the word positive or negative?’’), which require considerably more information, may not be exhaustively sensitive to such less sophisticated information. If so, these priming effects may be weakly conscious. The regression approach and the null sensitivity problem. Methodologically speaking, Greenwald and associates’ regression approach has been widely influential. Here, regression analysis techniques are used to model the unconscious perception index (e.g., priming effects) as a function of the conscious perception index (e.g., word valence classification performance), with the former on the y axis and the latter on the x-axis. Analogous to regression analysis generally, here, the y-intercept estimates the value of the putatively unconscious effect when performance on the conscious perception index is zero (i.e., chance). Accordingly, obtaining significant y-intercepts apparently demonstrates unconscious perception; if instead only conscious perceptual processes were present, the y-intercept should equal zero. Because y-intercepts can be generally obtained regardless of the overall mean on the x-axis variable (here, direct discrimination performance), the regression approach appears to avoid the null sensitivity problem. Specifically, demonstrating true overall chance performance now seems unnecessary, because one can simply ascertain the y-intercept. Notwithstanding this intuitively appealing framework, however, unfortunately the direct measure – here, the predictor on the x-axis still contains measurement error, the source of the null sensitivity problem to begin with. Such error in the predictor violates fundamental regression assumptions, and in single-predictor situations is
142
Perception: Subliminal and Implicit
known to artifactually flatten regression slopes. In many common situations, such flattened slopes will in turn produce invalidly inflated y-intercepts. Although corrective adjustments have been proposed by Greenwald and associates, it is thus far unclear whether these modified procedures definitively solve the problem. Moreover, examining y-intercepts is a special case of the general use of regression equations to predict one variable given knowledge of another. When the x and y variables are unrelated as in response window paradigms, using the regression equation to predict any value of y given x, including y-intercepts (i.e., predicting y when x ¼ 0), may simply not be meaningful. Despite these potentially serious problems, however, the regression approach is an important advance which moreover draws valuable attention to the relationship between conscious and unconscious influences, a theoretically crucial but often ignored matter. The regression approach has been widely influential, and much recent research routinely presents such analyses. Valence classification and other response conflict paradigms. Finally, in an important convergence, it is becoming increasingly clear that various other widely employed objective threshold priming paradigms, such as number classification (e.g., primes and targets are numbers above and below five) and motor activation (e.g., primes and targets are left- and right-pointing arrows) are response conflict paradigms, as is Greenwald and associates’ valence classification paradigm. Notably, these kinds of priming tasks yield similar processing characteristics such as the lack of relationship between the conscious and unconscious perception indices, producing the typically flat regression slopes noted above. Moreover, these other priming paradigms, similar to valence classification, have importantly demonstrated that consciously intended response strategies (e.g., to classify targets in the instructed way) further affect how the unconsciously presented primes are processed as well. This is important theoretically, as it contradicts long-standing assumptions that unconscious processes work completely independent of, and are unaffected by, conscious intentions. Instead, such evidence suggests that conscious intentions indeed moderate unconscious
perceptual influences, although not in the optional, flexible way generally observed with clearly consciously perceivable stimuli. Snodgrass and Associates’ Objective Threshold/Nonmonotonic Model Snodgrass and associates suggest that objective threshold effects’ apparent unreliability is in fact moderated by stimulus intensity, which in turn depends on the particular discrimination task used to determine the objective threshold. In particular, because detection tasks require less information than identification tasks, objective detection thresholds require more stringent exposure conditions than those necessary for objective identification thresholds. Because either task is plausibly exhaustively sensitive given typical unconscious perception indices, however, these task type differences in stimulus intensity have heretofore gone relatively unnoticed. Surprisingly, however, it turns out that stimulus intensity strongly moderates these effects: Under objective detection threshold conditions, unconscious effects are both sizeable and reliable; in contrast, under objective identification conditions, such effects are frequently absent. In contrast, as stimulus intensities increase still more, beyond the objective identification threshold and into subjective threshold regions, the relationship becomes positive, then matching the familiar pattern characteristic of subjective and/or clearly conscious stimuli. It has been repeatedly demonstrated, for example, that direct identification and/ or classification performance is strongly and positively related to priming effects when objective thresholds are clearly exceeded. Putting it all together, this constitutes a nonmonotonic relationship (first negative, then positive), see Figure 1(d). Methodological implications. Collectively, these findings suggest a methodologically powerful negative relationship between the conscious and unconscious perception indices when stimulus intensities are between the objective detection and identification thresholds. This directly contradicts the positive (or at least zero) relationship predicted by the single process conscious perception model, providing considerable warrant for inferring a second, unconscious perceptual process. Along the same lines, the negative relationship
Perception: Subliminal and Implicit
rebuts analogous predictions made by the exhaustiveness and null sensitivity problems. Further, the regression approach benefits when negative relationships are obtained, because slope flattening now underestimates rather than overestimates y-intercepts. Consequently, the presence of measurement error with negative relationships actually increases, rather than decreases, confidence in any obtained results. Moreover, the presence (vs. absence) of a relationship between the conscious and unconscious perception indices allows clearly valid use of the regression equation to estimate the y-intercept, whereas such procedures are dubious when the indices are unrelated. Finally, negative versus positive relationships have important implications for evaluating the probative force of various qualitative differences. In particular, if ostensibly unconscious effects are weaker, less complex, or less controlled than conscious effects, such qualitative differences – however plausible – are questionable because one would expect this same pattern with weak versus strong conscious perception as well. Unfortunately, many qualitative differences are vulnerable to this alternative interpretation (e.g., Greenwald and associates’ partial word effects; see above). On the other hand, obtaining the opposite pattern provides considerably stronger evidence for unconscious perception. For example, finding more complex effects under objective than subjective conditions (e.g., semantic vs. structural) would be harder to explain with solely conscious influences, given the latter’s stronger stimulus conditions. Process implications. The nonmonotonic relationship further suggests that conscious and unconscious influences may be functionally exclusive, such that when conscious perception is present, it typically overrides unconscious perceptual influences. At first, when only barely detectable, weak conscious perception is useless and hence interferes with unconscious effects, producing the negative relationship region. As the objective identification threshold is exceeded, however, conscious perception complex enough to drive typical unconscious perception index effects begins to emerge, and the relationship becomes positive. In contrast, the subjective and objective/rapid decay models assume that conscious and unconscious perceptual influences are either positively related
143
or unrelated, respectively. Finally, the nonmonotonic model further implies that subjective and objective threshold effects index qualitatively distinct processes, rather than the latter simply being weaker versions of the former, as is often assumed. Crucially, however, for the nonmonotonic relationship to be genuine, objective detection thresholds must indeed actually be below objective identification thresholds. If, instead, these thresholds were the same or even reversed, the nonmonotonic model would be weakened or refuted outright. Although available evidence is supportive thus far, further tests are clearly necessary. Moreover, it is important to note that objective detection threshold studies in general are relatively uncommon in recent years, probably because such thresholds are difficult to attain using typical computer monitors. Overall then, at present, the objective threshold/nonmonotonic model’s empirical base is significantly less extensive than that of the subjective and objective threshold/rapid decay models, and must be seen as relatively provisional.
Unresolved Issues and Current Controversies Various other fundamental issues are currently unresolved; these also have important methodological and theoretical implications. The Role of Direct versus Indirect Tasks In the vast majority unconscious perception research, the conscious perception index is a direct task, whereas the unconscious perception index is typically indirect. Although rarely discussed, this practice implicitly assumes that unconscious influences manifest more readily on indirect than direct measures – regardless of whether objective or subjective threshold methods are used. If reliable results in such designs are indeed obtainable, it implies that direct tasks indeed tap mostly or entirely conscious influences (cf. the exclusiveness problem). On the other hand, recent findings reliably suggest that unconscious perceptual processes do influence direct task performance after all, but in an unexpected fashion: Individual differences, sometimes interacting with task set, moderate the
144
Perception: Subliminal and Implicit
actual direction of unconscious influences such that some individuals perform above chance (i.e., facilitate) while others actually perform below chance (i.e., inhibit). Such bidirectional effects do not affect the overall mean; hence, they are easily missed unless individual differences are examined, leading to the erroneous inference that unconscious influences are simply absent on direct measures. In contrast, indirect measures seem more frequently to produce primarily unidirectional, facilitative effects – with certain important exceptions (e.g., response conflict priming paradigms in certain situations). It is currently unknown why direct versus indirect task effects would differ in this way. Is Attention Necessary for Consciousness? Some findings suggest that similar findings are obtained when stimuli are presented outside of focal attention, on the one hand, and when they are attended but rendered unconscious using typical masking techniques. This has led some to suggest that attention is necessary for consciousness – although not sufficient, because other findings suggest that attention can be directed to stimuli that nonetheless remain unconscious. Crucially, however, these parallels have been obtained using subjective threshold methods; whether unattended versus objectively unconscious stimuli would also produce similar findings is not yet known. Consequently, the unattended/subjectively unconscious parallels could alternatively mean that both reflect weakly conscious, not unconscious, processes. Unconscious Perception and Brain Processes Not surprisingly, neuroimaging and other physiological measures are increasingly employed to investigate the brain correlates of unconscious perceptual processes. Findings from such investigations, typically indicating that unconscious processes are less complex and/or relatively localized, are consistent with currently influential Global Workspace theories which link consciousness to widespread frontoparietal brain activations. However, as with other qualitative differences, it is
unclear whether this truly reflects differences between conscious and unconscious processes or may instead simply contrast weak versus strong conscious processes. Implicit, Unconscious, and Subliminal: Synonymous or Distinct? Frequently, these three terms are used interchangeably to indicate unconscious perceptual processing. Increasingly, however, it is clear that they differ in at least one important sense. For example, in implicit effects, the relevant stimuli themselves are clearly conscious – it is their influence on some subsequent task that is ostensibly unconscious. For example, in implicit memory paradigms, observers are perfectly conscious of initially presented word lists, but may not realize that these previously seen words influence how they complete word fragments later on. In contrast, unconscious perceptual paradigms – even when using implicit terminology – invariably attempt to prevent conscious perception of the critical stimuli proper, not just their subsequent influences on other tasks. Accordingly, unconscious and implicit perception may index distinct processes. For example, the former may assess phenomenal consciousness itself (e.g., visual qualia), whereas the latter may assess reflective (i.e., metacognitive) consciousness. Finally, although subliminal is universally used synonymously with unconscious, its original derivation can create erroneous impressions. Literally, subliminal means below the limen, which in classical psychophysics reflected the stimulus intensity at which subjects would report a stimulus 50% of the time. If assessed, discrimination performance easily exceeds chance at such intensities. Indeed, from the perspective of SDT, the limen is closely related to the criterion, and hence subjective threshold approaches. In modern usage, however, subliminal terminology is often used in objective threshold paradigms, which do not have the limen implication. Does Masking Technique Make a Difference? Different masking methods are frequently used in unsystematic ways. Usually, some form of pattern
Perception: Subliminal and Implicit
masking is used, where a particular kind of nonsense shape either immediately follows or precedes the primary stimulus, yielding backward or forward masking, respectively. Alternatively, other important techniques such as metacontrast masking and binocular rivalry are sometimes used. Notably, although masking theory suggests that the various techniques should produce importantly different effects, such differences are not yet completely clear, and certain techniques formerly believed to differentially limit unconscious processes apparently do not. Even so, as further investigations attempt to specify key features of unconscious processes, possible masking type effects should be carefully explored, both as potential confounds and as substantively interesting moderators in their own right.
Conclusions Notwithstanding the longstanding and perhaps especially difficult methodological issues in unconscious perception research, in recent years important advances (including qualitative differences, the regression approach, and the potential usefulness of negative relationships) have emerged which may allow substantial progress in this notoriously controversial area. Nonetheless, there remains sufficient uncertainty that a distinguished minority is still unconvinced that genuinely unconscious perception occurs in any form, especially involving complex semantic processing. With this in mind, considerable additional work is essential to finally enable definitive progress. Along the way, however, it is likely vital to distinguish subjective and objective methods, because they may index fundamentally distinct processes, rather than simply being more or less stringent. See also: Implicit Learning and Implicit Memory; Intuition, Creativity, and Unconscious Aspects of Problem Solving; Perception: Unconscious Influences
145
on Perceptual Interpretation; Psychoactive Drugs and Alterations to Consciousness; Unconscious Cognition.
Suggested Readings Abrams R and Greenwald A (2000) Parts outweigh the whole (word) in unconscious analysis of meaning. Psychological Science 11: 118–124. Cheesman J and Merikle PM (1984) Priming with and without awareness. Perception & Psychophysics 36: 387–395. Draine S and Greenwald A (1998) Replicable unconscious semantic priming. Journal of Experimental Psychology: General 127: 286–303. Dulany D (1997) Consciousness in the explicit (deliberative) and implicit (evocative). In: Cohen J and Schooler W (eds.) Scientific Approaches to Consciousness, pp. 179–212. Mahwah, NJ: Erlbaum. Holender D (1986) Semantic activation without conscious identification in dichotic listening, parafoveal vision, and visual masking: A survey and appraisal. Behavioral & Brain Sciences 9: 1–23. Marcel A (1983) Conscious and unconscious perception: Experiments in visual masking and word recognition. Cognitive Psychology 15: 197–237. Merikle P and Joordens S (1997) Parallels between perception without attention and perception without awareness. Consciousness and Cognition 6: 219–236. Merikle P, Smilek D, and Eastwood J (2001) Perception without awareness: Perspectives from cognitive psychology. Cognition 79: 115–134. Naccache L and Dehaene S (2001) Unconscious semantic priming extends to novel unseen stimuli. Cognition 80: 223–237. Reingold E and Merikle P (1988) Using direct and indirect measures to study perception without awareness. Perception & Psychophysics 44: 563–575. Reingold E and Merikle P (1990) On the inter-relatedness of theory and measurement in the study of unconscious processes. Mind & Language 5: 9–28. Snodgrass M (2002) Disambiguating conscious and unconscious influences: Do exclusion paradigms demonstrate unconscious perception? American Journal of Psychology 115: 545–580. Snodgrass M, Bernat E, and Shevrin H (2004a) Unconscious perception: A model-based approach to method and evidence. Perception & Psychophysics 66: 846–867. Snodgrass M, Bernat E, and Shevrin H (2004b) Unconscious perception at the objective detection threshold exists. Perception & Psychophysics 66: 846–867. Snodgrass M and Shevrin H (2006) Unconscious inhibition and facilitation at the objective detection threshold: Replicable and qualitatively different unconscious perceptual effects. Cognition 101: 43–79.
146
Perception: Subliminal and Implicit
Biographical Sketch
Michael Snodgrass is a senior research associate and a codirector of the Laboratory for Conscious and Unconscious Processes in the Department of Psychiatry at the University of Michigan Medical Center. Dr. Snodgrass’ primary research focus is unconscious perception, especially theoretical and methodological issues.
Natasha Kalaida is a project coordinator at the Laboratory for Conscious and Unconscious Processes in the Department of Psychiatry at the University of Michigan Medical Center. Her research interests include unconscious perception, especially translational applications investigating the causes and improved treatment of anxiety disorders.
E. Samuel Winer is a graduate student at the University of Illinois at Chicago, where he is completing his PhD in clinical psychology. His primary research focus is on unconscious perception in relation to individual differences.
Perception: The Binding Problem and the Coherence of Perception T Schmidt, University of Giessen, Giessen, Germany ã 2009 Elsevier Inc. All rights reserved.
Glossary Ba´lint–Holmes syndrome – A clinical condition following bilateral parietal brain damage, characterized by eye movement and reaching problems, severe spatial disorientation, and simultanagnosia. Binding problem – The task of bringing together individually coded features so that they can be integrated into a single coherent object representation. Sometimes binding has to occur across space (grouping), time (object constancy), or sensory modalities (crossmodal binding). Solving the binding problem requires explaining how binding is established, signaled, and read out by the cognitive system. Cardinal cell – A single cell signaling a distinct feature, object, or scene. Feature conjunction – A stimulus that is defined by more than a single feature. Sometimes, illusory feature conjunctions are perceived. Feature Integration Theory – Proposes that visual attention is used to link independently represented features to a particular spatial location in order to form an object representation. Neuronal synchronization – Two or more cells are synchronizing when they tend to fire their action potentials at the same time. Population coding – Groups of cells jointly signaling a feature, object, or scene. More efficient than coding by cardinal cells. Recurrent processing – Occurs when information going out from one area keeps reentering that area by feedback loops. In contrast, the fast feedforward sweep is the first one-way pass of activity through the
visual system before feedback information becomes effective. Simultanagnosia – A clinical condition that can be described as an inability to perceive more than one object at a time. Synesthesia – A benign variant of normal perception where the perception of certain objects is reliably accompanied by additional sensations, possibly in a different sensory modality. Visual search – Psychophysical task where participants have to search a target item among a set of distractor items. Often search is assumed to be serial when target positions have to be searched one at a time, and parallel when search times do not depend on the number of distractors. In the latter case, the target seems to ‘pop out’ immediately.
Introduction Most visual objects we encounter possess a multitude of properties. Looking out at a parking lot, you can see a number of cars, each one defined by its particular color, shape, and its unique place on the lot. Visually, these objects can be characterized as collections of visual features, which must be tied together in the correct way. ‘Binding’ refers to the process whereby separately analyzed features are combined to form a perceptual object. The ‘binding problem’ arises when features have to be assigned unambiguously to different objects without creating false combinations. Figure 1 gives an example. The stimulus that the system tries to encode here consists of a
147
148
Perception: The Binding Problem and the Coherence of Perception Stimulus
Features present: Red
Green
Blue
Circle
Square
Triangle
(a)
(b)
(c)
(d)
Without additional information which feature belongs where, most feature combinations are false:
etc. Figure 1 A binding problem occurs whenever a
stimulus is dismantled into separate features that have to be recombined at a later stage. If the stimulus consists of a red square, a blue triangle, and a green circle, simple feature detectors are only able to signal that a square, a triangle, and a circle must be present, and that there are the colors red, blue, and green. Correctly conjoining these features to the correct stimulus objects and locations requires a binding mechanism. Without binding, the interpretation of the stimulus remains ambiguous.
red square, a blue triangle, and a green circle, each shape at a particular position. Assuming that the system is clear about the relative positions of all objects, there are still many ways to combine color and shape features with those positions: Without prior knowledge which feature belongs where, there are no less than 36 possibilities here, even with only three colors, shapes, and positions. Considering larger collections of objects, such as those cars in the parking lot, immediately shows that the number of false combinations is typically immense. If our visual system was only able to check which features are present, but not how they combine, there would be little chance of picking the correct combination. This suggests that the binding problem must be solved before accurate perception is possible.
Figure 2 In these visual search displays, the target
element is a vertical black bar embedded in a context of distractor elements. The participants’ task would be to signal as quickly as possible whether the target is contained in a display. In panels (a) and (b), the target is easy to find because it is defined by a single feature (blackness) that is not shared by any distractor. As a result, the target seems to ‘pop out’ and can be found immediately, independent of how many distractors are present. In panels (c) and (d), the target is more difficult to find because it differs from the distractors by a conjunction of features (blackness and verticality), while each of these two features can occur in the distractors. FIT holds that in such a conjunction search, visual attention has to scan the display element by element, leading to search times that sharply increase with the number of distractors.
Empirical Evidence for Binding Problems in Visual Perception Psychophysical Evidence for Binding Problems in Vision There is solid evidence from psychophysical research that the system does face a binding problem, and sometimes fails to solve it. Much of this evidence comes from Anne Treisman’s paradigm of ‘visual search.’ In visual search tasks, participants try to find a target item amongst a set of distractor items, and indicate as quickly as possible whether the target is present or absent. To get a
Perception: The Binding Problem and the Coherence of Perception
feeling for this type of task, take a moment to search for the black vertical bar in each of the panels of Figure 2. You may have noticed that the target is very easy to find in panels (a) and (b). This is because the target differs from all the distractors in one distinctive feature: It is the only black item present and seems to ‘pop out’ of the display. Because search time is constant, no matter how large the number of distractors, this type of search is often called ‘parallel search.’ In contrast, search is difficult in panel (c) and especially in panel (d). The reason for this is that you have to search for a black vertical bar among black horizontal and white vertical bars. Because of this, you cannot use a single feature to distinguish the target from the distractors; instead, the target is defined by a ‘conjunction’ of features. Search times for conjunction targets are not only longer than pop-out searches, they also increase strongly with the number of distractors. This finding suggests that observers have to engage in a ‘serial search,’ where items are scanned one at a time until the target is found. This also agrees with the finding that search times increase twice as steeply with the numbers of distractors when the target is absent from the display. When the target is present, it is on average found halfway through the search, but when it is absent, all positions have to be searched before the search can be called off. This is not the case for parallel search: Here, observers are able to decide very quickly that ‘nothing pops out,’ and to declare the target absent. The fact that conjunction targets are more difficult to find than single-feature targets suggests that binding features into conjunctions is difficult for the visual system. As we will see below, Treisman’s Feature Integration Theory (FIT) postulates that visual attention is needed to combine single features into a coherent object representation. Treisman assumed that single features are represented independently of their spatial position, so that in the absence of binding, the features might become disattached from any specific position. Evidence for such ‘free-floating’ features comes from experiments where observers view brief displays of various objects under conditions of diverted or reduced attention. Various experiments showed that under such conditions, participants often report feature combinations that were
149
not actually there, which suggests that the features of the presented objects had been combined erroneously. For instance, after a brief presentation of red circles, green circles, and green squares, subjects may report a red square; and after presentation of a collection of letters ‘O’ and ‘R,’ they may report having seen a ‘P’ or a ‘Q ,’ falsely conjoining the R’s squiggle to one of the O’s. These mistakes are called ‘illusory conjunctions’; they are a clear indication that the visual system sometimes fails to bind features together in the correct way. Illusory conjunctions are frequent in patients with hemineglect (XXX), further indicating that attentional deficits may lead to binding problems. However, it is not entirely clear whether this failure occurs during actual perception or during the reconstruction from memory of what has been seen. The Big Challenge: Binding Over Space and Time The binding problem as classically formulated involves cases where a single, multi-featured object occupies a specific location in the visual field. In such cases, binding might be comparatively easy to explain because all the object features refer to the same spatial position. More difficult problems occur when binding has to be extended over space or time: these problems are known as ‘grouping’ and ‘object constancy’ and can be considered the toughest challenges in binding research. For example, imagine an octopus passing by an unsuspecting diver. The octopus might undergo marked changes in shape and color in addition to its change in position; yet these changing features must continue to be bound to the same animal. In this case, location does no longer serve as an unambiguous cue to object coherence: The different parts of the octopus’s body have to be assigned to the same animal, even though they occupy different spatial positions. When the octopus passes behind some partly occluding object, or some parts of its body occlude other parts, we still have to group the visible parts together. And when the moving animal is completely occluded for a stretch of time, we have to use memory to bind together the visual experiences of the animal before and after it passed the occluder. Young infants do not master this feat of object constancy over time. It also seems to
150
Perception: The Binding Problem and the Coherence of Perception
challenge the enemies of octopuses, whose most effective defensive weapon is an occluding ink cloud. Single-Case Evidence for Binding Problems: Synesthesia and the Ba´lint–Holmes Syndrome One instructive approach to the binding problem is to study populations of people where binding goes awry. ‘Synesthesia’ is a not-too-uncommon condition where people’s perception of certain events or objects is reliably accompanied by the experience of additional sensations, sometimes in a different modality. The most common type of synesthesia seems to be grapheme-color synesthesia, where the viewing of letters or digits is accompanied by the experience of colors. These combinations are usually one-to-one, so that each grapheme goes with a specific color, and are stable over time. In other synesthetes, digits might be associated with certain sounds; or the spoken names of certain persons might go with different tactile sensations. Synesthetes differ in the way these concomitant experiences intrude into ongoing perception: Some of them perceive digits or letters as downright colored, some only see a weak colored halo around them, and others report ‘knowing’ which colors belong to the digits but do not report experiencing them phenomenally. Synesthesia is a benign variant of human perception that bears no obvious disadvantages for the observer. To the contrary, some artists have used synesthetic experiences to inspire their art (e.g., color-sound synesthesia in painter Wassily Kandinsky, composers Olivier Messiaen and Duke Ellington, and electronic music avantgardist Aphex Twin). Some synesthetes with digit-color synesthesia actually use the additional color information when calculating numbers. Color-digit synesthesia can aid in finding a target digit in a visual search display, even though it is disputed whether synesthesia can elicit a full-fledged pop-out effect. Many studies show that synesthetic experiences can produce perceptual conflicts: For example, digits printed in their idiosyncratic synesthetic colors can be identified faster than digits printed in incongruous colors. Such results indicate that at least some types of synesthesia involve early visual processing. It has also been shown that synesthetically
experienced colors activate brain areas implicated in color processing. Even though it is still unclear why this happens, the binding of extra features to certain perceptual objects is stable over time and seems to be hard-wired, perhaps by permanent connections between cells that lead to reliable coactivations between different brain areas. In contrast to synesthesia, the Ba´lint–Holmes syndrome is a severely disabling clinical condition that is often seen after bilateral damage to the posterior parietal cortex (at the junction between the parietal and occipital lobes). It is disputed whether Ba´lint’s syndrome really qualifies as a distinct diagnostic category, because it is a rather loose collection of symptoms that may occur after different types of lesion, not necessarily involving the parietal cortex. Patients with Ba´lint–Holmes syndrome have difficulty directing their gaze from one visual object to the next, and are impaired in performing visually directed hand and arm movements (optic ataxia) even though they are not afflicted with paralysis of the affected limbs. Some patients suffer from severe spatial disorientation: In extreme cases, they may be unable to find their way from their bed to the sink in the same room. Importantly in our context, many of these patients also suffer from a condition called ‘simultanagnosia,’ which can be described as an inability to perceive more than one object at a time. For instance, when looking at the overlaid shapes in Figure 3, patients would only be able to consciously perceive one of them at a time. Other patients have difficulty reading because they tend to skip words and lines, and generally have to read sentences in a word-by-word
Figure 3 In this example, several contour shapes are
superimposed, and we have to trace the contours across space to tell the shapes apart. The spatial position of object parts is not sufficient for assigning them to the correct objects here. Patients with simultanagnosia, a condition often associated with the Ba´lint–Holmes syndrome, have difficulty distinguishing more than one object at a time in displays like this.
Perception: The Binding Problem and the Coherence of Perception
fashion. Many of these problems may stem from the fact that the patients have difficulty localizing different objects or object parts, changing their direction of gaze, or shifting their focus of attention from one aspect of the scene to another. In sum, synesthesia can be conceived as a bias to bind certain features to objects independently of whether this is required by the stimulus. In contrast, simultanagnosia is a condition where binding breaks down to a degree where only one coherent object can be maintained at a time. Both conditions show that binding is not a trivial feat for the visual system. A Binding Problem in Early Vision? The binding problem was first recognized in computer science and became a prominent research topic in psychology and neuroscience during the 1980s. One major force that gave the binding problem some urgency in cognitive neuroscience was the increasingly popular notion that the visual system disassembles incoming stimuli into different processing streams, each dealing with a separate feature such as color, orientation, or motion, in specialized regions of the cortex. The most influential model of this sort was put forward by Margaret Livingstone and David Hubel in 1988. These authors proposed that different visual features (specifically, color, motion, orientation, and retinal disparity) were analyzed in anatomically separate processing streams. These streams would originate in different types of ganglion cells in the retina, run through different types of layers in the visual part of the thalamus, and enter different layers of the primary visual cortex (V1). Staining patterns for cytochrome oxidase (CO), an enzyme whose presence is associated with increased metabolic demand, were supposed to provide the anatomical scaffold for further segregation in V1 and the adjacent area V2. For instance, color would be processed in the CO-rich ‘blobs’ of V1 and then passed on to the ‘thin stripes’ of V2 and finally the putative ‘color area’ V4. Similarly, the motion stream would run through the ‘thick stripes’ of area V2 and on to the putative ‘motion area’ MT. Early single-cell studies largely supported this model. In the Livingstone–Hubel model, different visual features are segregated into different anatomical
151
compartments, and each feature is processed by a specialized subset of cells. The visual system is thus viewed as disassembling incoming visual information into their component features. This model has been massively popular, and its central tenets are still taken for granted by many researchers. Accordingly, many reviews of the binding problem use the disassembly metaphor as a point of departure: If the system dismantles incoming information into its component features, how does it later reassemble the processed features into perceptual objects? However, more recent research has undermined many of the central tenets of the Livingstone–Hubel model. Careful anatomical studies indicate that signals from different thalamic layers do not remain segregated while filtering through the different layers of V1, but mix early and extensively. While it has been confirmed that CO-rich compartments in V1 tend to be linked to those in V2 (albeit not as exclusively as originally proposed), the evidence for a strict segregation of specialized cells into these compartments is mixed. For example, even though color-selective cells are more prominent in the ‘thin’ stripes of V2 than in the ‘thick’ or ‘interstripes,’ most of them are tuned to stimulus orientation as well. The principal conclusion from this evidence is that neurons in V1 and V2 should not be regarded as simple feature detectors at all: It is unlikely that simple visual features are ever encoded completely independently of each other in early visual processing. Instead, cells are typically broadly tuned to more than one feature without being neatly segregated into different cortical compartments, which raises doubts about the disassembly metaphor in its strong form.
Other Types of Binding Multisensory Integration As noted above, the processing of different visual features is not strictly segregated in early cortical areas. In contrast, the processing of different sensory modalities, like vision, audition, and touch, clearly is, with each one originating in a welldefined primary sensory area of the cortex. To bring different modalities into register so that they refer to a single multisensory object, such as a honking car or an insect you see tickling your
152
Perception: The Binding Problem and the Coherence of Perception
foot, some type of binding has to occur. This integration of information from different modalities is called ‘crossmodal binding.’ There are many behavioral indicators of crossmodal binding. For example, people can respond faster to simultaneous stimuli in two different modalities than to either stimulus alone (the ‘redundant target effect’). Facilitating effects from redundant stimuli in different modalities are especially strong if these stimuli occur in close temporal and spatial proximity to the stimulus in the target modality. For instance, sensitivity in the detection of a visual stimulus in one modality can be enhanced by simultaneously presenting an auditory stimulus at the same location. Some effects of crossmodal binding are so compelling that they may lead to crossmodal illusions: For example, a ventriloquist’s voice may appear to come from the puppet rather than the actual speaker, even though the two are clearly spatially separate. Anatomically, widespread areas of the brain are involved in the analysis of sensory stimuli without being devoted to any sensory modality in particular. These include the premotor, cingulate, and prefrontal cortex. Other regions are especially important for combining input from two or more sensory modalities, for example, the superior temporal gyrus (vision and audition), many areas in the parietal cortex (vision, audition, and somatosensory information) as well as many subcortical structures (e.g., the midbrain’s superior colliculus, which contains different retinotopic maps for vision, audition, and motor output, all superimposed in different layers). Binding Problems in Action Control and Monitoring Other types of binding problems occur in the control and monitoring of motor actions. The great majority of motor tasks require an accurate mapping of stimuli to responses. For example, when driving a car, you are expected to respond reliably to special visual cues like red traffic lights or stop signs. Laboratory experiments often involve arbitrary mappings between stimuli and responses. Some of these stimulus–response mappings are more difficult to perform than others, which can be due to so-called ‘stimulus–response
compatibility’ effects. For example, responding with a left keypress to a square and with a right keypress to a circle is easier when the square is presented on the left-hand side of the display while the circle is presented on the right-hand side (this particular compatibility phenomenon is known as the ‘Simon effect’). Other binding problems occur when simple motor actions must be integrated into more complex actions or action sequences. Binding effects might also play a role in the monitoring of self-generated actions. Patrick Haggard and coworkers showed that when participants performed keypresses at times of their own choosing to elicit tone signals, they tended to perceive the time of action and the time of the action’s effect as closer than they really were. The authors argue that this misjudgment reflects a binding together of actions and action effects.
Possible Solutions to the Binding Problem in Visual Perception Crossmodal, motor, and sensorimotor binding effects pose intriguing challenges for any theory of binding, but they have been rarely addressed so far. Most theories of binding focus on visual processing. In what follows, we will assume that the visual system is confronted with an essentially infinite set of possible objects or scenes, and has to deal with this diversity given a finite number of visual cells that can encode various stimulus features. The system’s task is to use information from the feature detectors to establish a representation of the visual scene where each detected feature is assigned unambiguously to the correct object. (As a caveat, remember our conclusion that most cells in early visual areas do not qualify as simple feature detectors, but are somewhat selective for several features.) A number of different approaches to solving the binding problem have been proposed. To evaluate these proposals, it must be noted that a complete theory of binding has to account for three types of problems. First, there is the ‘establishment problem,’ which is essentially the binding problem proper: How does the system determine which features belong together and should be bound
Perception: The Binding Problem and the Coherence of Perception
into a single object? Second, there is the problem how binding is signaled in the brain: How does the representation of bound stimulus features differ from that of unbound features? Third, it must be explained how this information is read out by the rest of the system: How does the brain recognize and retrieve an integrated stimulus representation where features have been bound together? In addition, a convincing theory of binding should be able to explain grouping and object constancy in addition to the binding of stimulus features at a single location (none of the current theories has been designed to deal with binding problems in multisensory integration or action control). Table 1 reviews to what extent the different theoretical approaches address these issues. Binding by Convergence The simplest way to bind features together is to have all the detectors of the simple features converge on a single ‘cardinal cell.’ This cell would then serve as a detector of the compound object. This solution has been proposed in the 1950s by Horace Barlow and is known as the ‘neuron doctrine.’ It was propelled forward by David Hubel’s and Torsten Wiesel’s proposal that cells with complex stimulus selectivity (e.g., for motion in a specific direction) could be formed by convergence of cells with simpler properties, and by their continuous discovery of such cells since the 1960s. More recently, neurons in the inferotemporal cortex have been discovered that are amazingly selective for even very complex objects: They are especially sensitive to complex shapes or faces, or even respond selectively to pictures containing Bill Clinton (that’s no joke!). Table 1
153
However, the notion of cardinal cells was soon caricaturized as postulating ‘‘grandmother cells’’ or ‘‘yellow Volkswagen cells’’, which would only fire when your grandma or her old car was in your field of view. Indeed, a convergence model of this sort runs into plausibility problems very quickly. The number of distinct objects or visual scenes that can be encountered is essentially unlimited, and while the number of distinguishable scenes is likely to be much smaller, it is still vast. Now the number of visually responsive neurons is vast as well. But even if the number of neurons was sufficient to represent each distinguishable scene on a one-toone basis, it would still be unclear how this system could ever classify a completely novel object, say, a blue camel. If you encountered a blue camel for the first time, how could you expect to possess the appropriate blue-camel neuron to recognize it? And if you had such a detector, wouldn’t that imply that you likely had green, red, and yellow detectors for other animals as well? Obviously, it seems implausible that the brain would harbor specialized detectors for objects it was likely to never encounter. A further problem occurs at read-out: If a given cardinal cell fires, how is the rest of the system to know which stimulus it is encoding, and in general which cells have to be accessed to extract some information of interest? While some grandmother (or Bill Clinton) neurons do seem to exist in the visual system, coding by cardinal cells cannot serve as a general coding strategy. Binding by Population Coding Even though some researchers continue to argue for a simple convergence theory of binding, most
Aspects of the binding problem addressed by different theories
Convergence coding Population coding Synchrony Attention Recurrent processing
Addresses establishment problem
Addresses signaling problem
Addresses readout problem
Addresses grouping or object constancy
No
Yes
No
Yes
No
Yes
No
Yes
No Yes Yes
Yes Yes Yes
No Yes Yes
Yes No Yes
154
Perception: The Binding Problem and the Coherence of Perception
authors dismiss this approach as implausible because it requires a vast number of detector cells to encode a combinatorial nightmare of possible feature conjunctions. A simple way to work around this difficulty is to consider the possibility that different cells jointly signal different objects. For example, consider a network of cells where each cell can be in one of two states, ON (= 1) or OFF (= 0). To represent 1024 objects, we could use 1024 cardinal cells, only one of which would be ON for any given object presented. But alternatively, we could assign one object to each possible state of the entire network. In that case, only 10 ON/OFF cells would be sufficient to represent the objects because they allow for 210 = 1024 states of the network (e.g., 1001011100 could represent one object and 0111010001 another). This scheme is known as ‘population coding.’ Of course, it is not confined to cells with binary states. However, problems remain even if cardinal-cell coding is replaced by population coding. Just like the cardinal-cell model, the theory is able to tell us how binding is signaled, but not how it is established (i.e., how these cell properties were generated in the first place) nor how the relevant information can be read out (i.e., how the rest of the system knows which cell or population state is signaling what). Thus, both the convergence and the population coding theories remain confined to explaining object representation rather than being a solution to the actual binding problem. On the other hand, they are able to deal with binding across space and time, because cells coding different object parts or positions could readily converge onto cells that would signal an object independent of its particular state or position. Binding by Synchrony In the early 1980s, Christoph von der Malsburg proposed a different way how neurons could signal that they are encoding the same visual object. He proposed that cells encoding properties of the same object could fire in ‘synchrony,’ that is, emit their spikes at about the same time (say, within a time window of 5–10 ms). In a coding scheme like this, a small number of objects could be represented at the same time, each one signaled by a distinct firing frequency of the neurons encoding
it. This theory has been playfully likened to the working of a transistor radio, which tunes into different stations by switching from one frequency band to the next. Much of the empirical work on neural synchronization comes from Wolf Singer and his coworkers. In their studies, these researchers could convincingly show that synchronization does occur in the brains of cats and macaque monkeys. In a particularly striking study, recordings were made from pairs of cells in area MT of macaque monkeys, a visual area that contains many cells sensitive to motion in a specific direction. The receptive fields of the two cells were partly overlapping, so that a single moving bar could cover both receptive fields. Cells synchronized their spiking activity when a single bar stimulated both receptive fields, but not when two unconnected bars stimulated the receptive fields simultaneously – even if the unconnected bars were actually more effective than the single bar in evoking responses from the two cells. Results like this are consistent with the idea that synchrony between two cells serves to signal that both cells are dealing with the same object. Even though the evidence for neuronal synchronization is solid, its interpretation has been controversial. For example, it is still unclear how synchronization comes to pass: Are cells actively synchronized by a top-down signal, is synchronization elicited by the visual stimulus, or do cells somehow synchronize each other? Obviously, if there was some such mechanism that could bind single-neuron feature detectors together by synchrony, it would be that synchronizing mechanism, not the synchronization per se, that solves the binding problem. A similar shortcoming exists on the read-out side: How could neurons detect that some of their inputs are synchronous? Incoming electrical potentials sum more effectively when they arrive together, which means that inputs from synchronous cells might have more impact on the target neuron than inputs from asynchronous cells. But typical visual neurons receive input from several thousand cells, and it takes at least hundreds of incoming spikes to elicit a single spike from the target cell. This basically means that each incoming spike will always be simultaneous to many other incoming
Perception: The Binding Problem and the Coherence of Perception
spikes, even without synchronization, and that the advantage gained from active synchronization of some of these spikes might be small. The synchrony model might be described as a population-coding model with the possibility of neuronal synchronization added. For this reason, it is at least as powerful as the populationcoding approach, including its potential capacity to deal with binding across space and time. However, it also inherits some of its shortcomings: Just like the cardinal-cell and population-code models, the synchrony model is a theory of how binding is signaled, not how it is established or read out. The mechanisms that generate and decode synchrony have remained elusive so far, and it is still unclear to what extent neuronal synchrony plays a role in signaling binding. Binding by Attention Anne Treisman’s Feature Integration Theory (FIT), developed in the context of visual search tasks, postulates that the correct binding of object features requires visual attention. Treisman starts from the assumption that the visual system is equipped with a set of ‘feature maps,’ with a different map for every conceivable feature (e.g., ‘red,’ ‘green,’ ‘square,’ ‘circular’). These feature maps are connected to a ‘master map of locations,’ which encodes the exact positions of objects in visual space (Figure 4). When the search target is defined by a single feature, for example, ‘red,’ one only has to monitor the respective feature map and trace its activation to the target’s position on the master map. This is what occurs in parallel search when the target seems to ‘pop out’ of the display (see ‘Empirical evidence for binding problems in visual perception’ section). However, when the target is defined by a conjunction of features, it is necessary to employ visual attention. Each object location on the master map has to be scanned by a ‘spotlight’ of attention, and for each of these positions the single features have to be collected from the feature maps and bound into an ‘object file’ that can be compared with a target template stored in working memory. As soon as the current object file matches the target template, the object has been found, and the search is terminated. Because the sequential scanning of object positions is a timeconsuming process that depends on the number of
155
Feature maps: White
Master map of locations:
Search display:
Black vertical
Focus of attention
Horizontal
Figure 4 In FIT, stimuli are disassembled into their component features and represented in multiple ‘feature maps,’ which all feed into a ‘master map of locations.’ Visual attention must be directed to a position in the master map to extract all the features corresponding to this position and bind them together in a single object representation. In the original theory proposed by Anne Treisman and G. Gelade, location information was exclusively confined to the master map. Later variants and extensions of the theory differ in the way they allow positional information to be encoded within the feature maps themselves, making them more plausible physiologically.
positions to be covered, this model can explain serial search. Subsequent research has produced some challenges for Treisman’s theory that led to refinements and variants. More recent theories extending Treisman’s work (e.g., Jeremy Wolfe’s ‘guided search theory’) view the master map as a map of object ‘salience,’ that is, the degree to which an object differs in its features from the remaining objects. They stress the possibility of controlled preactivation of feature maps, which allows searching for single features as well as feature conjunctions. Consistent with that idea, many studies showed that the distinction between parallel and serial search was not all-or-none: Sometimes, singlefeature targets would be difficult to detect; sometimes conjunction targets would be surprisingly easy. Furthermore, it is unlikely that all the stimulus distinctions that can be used in visual search experiments are independently coded in the brain in separate feature maps: While this is conceivably so for the color or orientation of simple stimuli, it is less likely for features such as stimulus size, ‘circularity,’ or ‘squareness.’ Finally, the original notion that
156
Perception: The Binding Problem and the Coherence of Perception
features are encoded independently of stimulus location has become physiologically implausible, because many identified feature maps in the brain either show some retinotopy themselves, or are tightly linked and cross-activated by retinotopic maps. In FIT, the deployment of visual attention is necessary to bind single features into coherent objects. A major advantage of FIT is that it can explain all stages of the binding process, not just the way binding is signaled. The trick is to use the ‘spotlight’ of visual attention to solve the binding problem sequentially at small portions of the image, where all the assembled features are simply assumed to belong to only one local object – in such a small, local image patch, there is simply no binding problem left to solve. The flipside of this strategy is that the theory cannot solve binding problems where the different features are at different locations, as is the case in grouping and image segmentation. Binding by Recurrent Processing Generally, visual areas receiving input from earlier areas also send massive feedback connections back to the originating areas. These feedback connections allow for ‘recurrent activity’ between different areas in the visual hierarchy. Victor Lamme and Pieter Roelfsema argue that recurrent processing is critical for the development of visual attention, visual awareness, binding, and grouping. Analyzing the response latencies of various cortical areas to a sudden visual stimulus, these authors argued that each new stimulus creates a wave of activation traveling from posterior to anterior areas, reaching most cortical areas within about 150 ms, including prefrontal and primary motor cortices. The authors estimated that this leaves cells with only about 10 ms time to pass their own activation on to later areas, which is about the duration of a typical interspike interval. Therefore, if most cells have to pass on their activation with the next spike fired, there is little or no time for them to integrate feedback from other cells. Based on this, Lamme and Roelfsema suggested that the first wave of visual activation travels through the system as a ‘fast feedforward sweep’ whose wave front is essentially free of intracortical feedback information. Naturally, the only binding processes possible in this feedforward activation phase would be
binding by convergence or population coding. However, Lamme and Roelfsema argue that feedforward processing is generally not sufficient for solving the binding problem or for generating visual awareness. Along with several other authors, they propose that conscious perception is possible only with recurrent processing of stimuli. Evidence for this view comes from studies indicating that visual awareness of a stimulus is suppressed if feedback loops from higher visual areas through primary visual cortex are disrupted at critical points in time, for instance, by a visual masking stimulus or by transient magnetic stimulation. Recurrent processing is not only viewed as a precondition for visual awareness here, but also for fully solving the binding problem. In particular, the theory assumes that feature detectors (or populations) activated by the feedforward sweep send reentrant projections back to earlier areas, so that activation in any area can be readily linked to earlier topographic maps. In this way, features of a single object can be processed in separate areas but would all the time refer back to a single spatial location by recurrent interactions with a common spatial map. The analogy to Treisman’s FIT is obvious; and indeed Lamme regards visual selective attention as a special form of recurrent activity. Based on the distinction between feedforward and recurrent processing, Pieter Roelfsema has suggested a model that is also suitable for grouping and image segmentation. He suggests that grouping occurs in two steps: At first, there is only ‘base grouping,’ which is confined to the stimulus conjunctions that can be encoded by convergent or population coding during the feedforward phase of processing. This step is followed by ‘incremental grouping,’ whereby neurons encoding the same object are labeled with enhanced activity. Incremental grouping depends on recurrent processing and is able to actively bind different parts of an object together across space. As an example, let’s take the problem of telling apart the three shapes in Figure 3. Incremental grouping would start as an activity enhancement somewhere on an outline contour and then spread along the contour until the entire shape was labeled. On its way around the contour, the process would take advantage of the fact that visual cells selective for similar contour orientations tend to wire together and coactivate each other.
Perception: The Binding Problem and the Coherence of Perception
Binding and Awareness What is the relationship between visual binding and visual awareness? Some authors are convinced that solving the binding problem is the key to solving the consciousness problem. Indeed, we have the subjective impression of a world where all visual objects are seamlessly coherent, and we have seen that it takes sophisticated laboratory experiments to demonstrate that a visual binding problem exists at all. Also, some of the theories of binding also address visual awareness (e.g., the recurrent activation model). However, the relationship between binding and awareness is less than clear. One major question is whether awareness starts before or after binding has been accomplished. For example, reports of illusory conjunctions suggest that we are sometimes phenomenally aware of features that have not yet been bound into a single object. Consistent with that observation, Victor Lamme and Ronald Rensink have both proposed theories where vision starts from a transient, phenomenally conscious representation where binding might still be incomplete, and is transformed by visual attention or recurrent processing into a more durable representation where tightly bound objects can be accessed for further cognitive processing. Note that in those theories, binding is not a necessary condition for awareness. At least for the moment, it therefore seems best to treat binding and awareness as separate problems.
Conclusions ‘Binding’ refers to a process whereby separately analyzed features or object parts are combined to form a single perceptual object. The ‘binding problem’ arises when features have to be assigned unambiguously to different objects without creating false combinations. Although it is unclear to what extent binding problems already arise in early neural coding, there is solid psychophysical evidence that they can occur in vision, multisensory integration, and action control. Moreover, binding breaks down in Ba´lint’s syndrome, and can occur in unexpected ways in synesthesia. Proposed solutions include
157
convergent cell specialization, population coding, temporal synchronization, attentional selection, and recurrent activation. While all of these approaches can explain how binding is signaled in the brain, some of them have difficulty explaining how it is detected, or how it is established in the first place. Because the relationship between visual binding and visual awareness is still unclear, it currently seems best to treat both problems separately. See also: The Neural Basis of Perceptual Awareness; Neurobiological Theories of Consciousness.
Suggested Readings Calvert GA and Thesen T (2004) Multisensory integration: Methodological approaches and emerging principles in the human brain. Journal of Physiology 98: 191–205. Ghose GM and Maunsell J (1999) Specialized representations in visual cortex: A role for binding? Neuron 24: 79–85. Gray CM (1999) The temporal correlation hypothesis of visual feature integration: Still alive and well. Neuron 24: 31–47. Haggard P and Cole J (2007) Intention, attention and the temporal experience of action. Consciousness & Cognition 16: 211–220. Hubbard EM and Ramachandran VS (2005) Neurocognitive mechanisms of synesthesia. Neuron 48: 509–520. Lamme VAF (2003) Why visual attention and awareness are different. Trends in Cognitive Sciences 7: 12–18. Lamme VAF and Roelfsema PR (2000) The distinct modes of vision offered by feedforward and recurrent processing. Trends in Neurosciences 23: 571–579. Rensink RA (2000) The dynamic representation of scenes. Visual Cognition 7: 17–42. Rizzo M and Vecera SP (2002) Psychoanatomical substrates of Ba´lint’s syndrome. Journal of Neurology, Neurosurgery, and Psychiatry 72: 162–178. Roelfsema PR (2006) Cortical algorithms for perceptual grouping. Annual Review of Neuroscience 29: 203–227. Shadlen MN and Movshon JA (1999) Synchrony unbound: A critical evaluation of the temporal binding hypothesis. Neuron 24: 67–77. Sincich LC and Horton JC (2005) The circuitry of V1 and V2: Integration of color, form, and motion. Annual Review of Neuroscience 28: 303–326. Wolfe JM and Cave KR (1999) The psychophysical evidence for a binding problem in human vision. Neuron 24: 11–17.
Suggested Listening Gentle Giant (1972) Octopus. Hamburg, Germany: Repertoire Records.
158
Perception: The Binding Problem and the Coherence of Perception
Biographical Sketch
Thomas Schmidt graduated in psychology at the University of Braunschweig, Germany, in 1997. He received his doctoral degree in natural sciences in 2002 and his habilitation degree in 2007, both from the University of Goettingen. Since 2005, he is working as a senior researcher at the Department of General and Experimental Psychology at the University of Giessen, Germany. He has recently accepted a call for a professorship in Psychology at the Technical University of Kaiserslautern, Germany. His major research interests include unconscious perception and visuomotor priming effects, basic visual perception and attention, visual short-term memory, and spatial cognition. He is currently running a project funded by the German Research Foundation, entitled ‘Response priming, attention and awareness.’ His website: www.allpsych.uni-giessen.de/thomas.
Perception: Unconscious Influences on Perceptual Interpretation B G Breitmeyer, University of Houston, Houston, TX, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Blindsight – A neurological visual deficit, produced by destruction of the primary visual cortex, in which a patient reports no phenomenal awareness of stimuli that nonetheless can affect his/her behavior. Extrastriate cortex – Also known as peri- or circumstriate cortex, the secondary and tertiary areas of cortical visual processing. Masked priming – The ability of a stimulus whose visibility is suppressed or masked to prime the perceptual response to a probe stimulus. Primary visual cortex – Also known as striate cortex, the initial site of cortical visual processing. Qualia – The subjectively experienced aspects of the contents of visual cognition, for example, the color, shape, or motion of a stimulus. Subcortical visual processing – The processing of information anywhere below the level of the visual cortex. This includes areas such as the lateral geniculate in the thalamus and the superior colliculus and pulvinar in the midbrain. Visibility suppression – The suppression of the visibility of a stimulus via experimental procedures that include the use of various psychophysical techniques as well as the application of magnetic pulses at cranial points overlying specific areas of the visual cortex.
Introduction Prior to the 1980s, few if any studies published in scientific journals or books dealt explicitly with
consciousness. The scientific study of consciousness and of its companion, the unconscious, began during the latter half of the nineteenth century. In vogue in these early years of academic psychology and psychoanalysis, it subsequently fell into disfavor, especially in the Anglo-American academy, with the advent of behaviorism and its emphasis on overtly measurable or observable variables. It remained in this unfavorable state until the latter part of the twentieth century when the study of consciousness accelerated noticeably, especially in the past 25 years. However, historical events predating and anticipating this resurgence of interest were already evident in the decade following World War II. Theoretical developments in the study of information and communication, artificial intelligence, and neuroscience coincided and converged in the 1950s with what can be called the cognitive revolution in experimental psychology. One of the key features of this revolution was the study of human information processing and its control via attention, a phenomenon that since William James’s The Principles of Psychology was viewed as intimately related to consciousness. For that reason, from the 1950s until the 1980s, very many investigations of psychological and cognitive phenomena implicitly focused on topics that currently are at the forefront of consciousness research. Since the 1980s and especially the 1990s, the Decade of the Brain, the number of scientific and scholarly studies and publications on consciousness has run into hundreds and perhaps thousands, most of them published in recently established journals dealing explicitly with consciousness. Moreover, these journals are affiliated with several professional organizations devoted to the scientific and scholarly study of consciousness. The flipside of this resurgent interest in the scientific and scholarly study of consciousness has been a renewed interest in the study of
159
160
Perception: Unconscious Influences on Perceptual Interpretation
unconscious information processing, sometimes known by its misnomer subliminal perception. Subliminal (or unconscious) perception has been considered a misnomer for two reasons. First, along with the taboo of consciousness studies occasioned by the ascendancy of behaviorism, the very existence of subliminal perception and its putative influence on human action has been viewed extremely skeptically, if not denied outright, on theoretical, methodological, and empirical grounds. Initially much of the interest in subliminal perception was motivated by psychodynamic approaches. While they advocated a theory-driven hermeneutics which allowed interpretation of manifest behaviors or symptoms in terms of a hidden unconscious, for the more positivist-minded scientists such interpretations appeared much too facile and post hoc, reflecting the lack in these approaches of testable hypotheses clearly linking theoretical concepts to experimental outcomes. The collateral damage of this severe skepticism prevailed at least until the 1980s. Second, according to the definition adopted by some leading cognitive psychologists perception is the conscious registration of sensory activity. Thus one can make a distinction between vision and visual perception, audition and auditory perception, and so on, with the former of each pair of terms being more inclusive than the latter. This definition or perception is adopted in the following discussions and perforce excludes talk of subliminal or unconscious perception. Such exclusion of course leaves as a remainder a good portion of visual information processing that is not perceptual and that can influence human cognition and action, so to speak, below the conscious radar. By definition, we cannot be consciously aware of the contents of these unconscious processes. Hence, any unconscious influences on perceptual interpretation cannot be assessed by conscious introspection. Fortunately they can be inferred from the results of studies that use a variety of objective procedures, including psychophysical, behavioral, electrophysiological, and brain-imaging techniques. The inferences drawn from these findings have been subject to ongoing theoretical, methodological, and empirical controversies. Although the following deals exclusively with unconscious visual information processing, the
major issues discussed also apply to the study of unconscious processing in other sensory modalities.
Conceptual, Empirical, and Methodological Issues in Studying Unconscious Information Processing Defining Unconscious Information Processing in Terms of Phenomenal Consciousness Any discussion of unconscious influences on perceptual interpretation presupposes some understanding of unconscious information processing. Establishing such a conceptual handle requires arriving at workable and acceptable definitions of consciousness. Through our individual private experiences, all of us know what consciousness is; however, most of us have only vague or conceptually confused ideas about how to define it. Arriving at clear conceptual distinctions and definitions is therefore a preliminary but necessary exercise. As a starter, a most useful distinction is that between consciousness as a state and consciousness as a trait. Consciousness as a state globally characterizes our waking hours, when we are (more or less) alert, and depends primarily on the overall level of arousal or activation of the central nervous system. It therefore is something that we lack entirely when we are in a deep, dreamless sleep, in deep anesthesia, or in a coma. When awake there can be various gradations in the global state of consciousness that depend on effects of diurnal activation cycles, fatigue, disease, or drugs. It is possible that unconscious information processing occurs during deep dreamless sleep or anesthesia; however, it is difficult to assess such processing. The definition of consciousness as a state may arouse little controversy, since it can be tied to quantifiable indices of activation in the central nervous system. Consciousness as state can interact with cognitive contents. For example, a driver who hears a siren briefly becomes more aroused and alert to the ongoing traffic pattern. Cognitive contents can be characterized by a number of features. For instance, some cognitive contents differ according to their specific modality such as audition, touch, or smell. Similarly
Perception: Unconscious Influences on Perceptual Interpretation
consciousness can be viewed as a trait or feature of cognitive contents. For that reason its definition is a bit more controversial. When cognitive contents are conscious they are deemed to possess qualia that confer on them a feely aspect, which is like knowing or having something from a first-person subjective perspective. One can be in a highly alert state of consciousness yet, due to a neurological disorder or an experimental manipulation, the experiential field can be lacking some qualia characterizing a particular object of cognition, for example, the shape or the color of an object and so on. Most current research in unconscious information processing, since it uses human observers who are in a fully alert state, explores this trait aspect of consciousness. Thus one can, by attenuating or suppressing certain qualia from experience, explore the remaining types and levels of information processing. When qualia are suppressed, we can speak of types and levels of unconscious information processing. Even for the fully alert state, one can distinguish among various types of consciousness. Distinctions abound; and because they are highly controversial and debated among cognitive scientists, they cannot be covered exhaustively. To illustrate a sample of the controversies and problems that have arisen, we can focus on four categories of consciousness that recently have captured the attention of cognitive scientists and philosophers: phenomenal, access, reflective, and narrative consciousness. Each of these presupposes that one is in a conscious state.
161
such as forced-choice discrimination of the location of objects or their direction of motion in cases of blindsight, where phenomenal consciousness (of qualia) is lacking, could qualify as access consciousness. On the other hand, it has been noted since the beginnings of experimental psychology, that the phenomenal visual field at any one time can contain much greater content than can be accessed through verbal (or other means of) report. For example, numerous studies of short-term visual store, also known as iconic memory, have demonstrated that one can be (brief ly) aware of the extensive contents of the visual field but have only limited access to it (imagine having to report what is seen when a lightning flash briefly illuminates a richly varied landscape at night). Hence, phenomenal awareness is neither a necessary condition (e.g., blindsight) nor a sufficient one (iconic memory) for defining access consciousness. These characteristics of access consciousness, while presumably establishing its independence from phenomenal consciousness, are puzzling to many cognitive scientists who find it hard to talk of consciousness without phenomenal awareness. Below it will become apparent why the positing of access consciousness without a phenomenal counterpart would seriously question an entire research program on unconscious information processing. For that reason, most investigators in this program subsume access consciousness under phenomenal consciousness rather than vice versa.
Phenomenal consciousness This type of consciousness can be considered as the totality of the subjective contents, that is, of experienced qualia.
Narrative consciousness As long as we are in a conscious state, we experience a flow or stream of contents passing through our awareness. These changing contents are viewed as the contents of a temporally evolving narrative.
Access consciousness A second type of consciousness has been defined more inclusively in terms of control of cognitive behavior expressed through reasoning, recognition, identification, verbal reporting, or motor response. Hence, access consciousness, at least in limited form, can be considered as a consciousness that is separable from phenomenal consciousness or simply includes it as one of its particular forms. On the one hand, the control of some behaviors
Reflective consciousness This type of consciousness is defined in terms of a monitored self-awareness of phenomenal consciousness. Such monitoring is most often the exception than the rule (although it may be a highly polished skill among practiced meditators). For example, a person may be phenomenally aware at any one time of a particular cognitive content, and his or her cognitive system may have access to this content for control of behavior, but s/he
162
Perception: Unconscious Influences on Perceptual Interpretation
may be unaware of the fact that s/he is having these experienced contents or that they change over time. In this sense, reflective consciousness can be regarded as a distinct form of, or separable from, phenomenal, access, and narrative consciousness. Yet, at the moment that the reflective person becomes aware of the changing contents – that is, flow or stream – of his or her narrative consciousness, this awareness is itself a new content of narrative consciousness and, having been recognized as a new content, is also an expression of access and of phenomenal consciousness. For these reasons and until (1) debates among cognitive scientists converge on a settlement regarding controversies surrounding these (and other) concepts of consciousness and (2) distinctions among consciousnesses are sufficiently refined, the following schema is adopted: All types of consciousness are regarded as specific variants of phenomenal consciousness. Another important reason is that if we allow behaviorally relevant information somehow accessed in, say, blindsight to qualify as a form of consciousness, we have every reason to reject many, if not all, influences on human action or perceptual interpretation that are phenomenally unconscious. Thus, in the following discussions, lack of phenomenal awareness, that is, lack of experienced qualia, is taken as the sine qua non of unconscious information processing. Methodological Issues Concerning the Study of Unconscious Information Processing Given this specific definition of unconscious information processing, establishing or exploiting conditions which adequately comply with this definition poses several problems. In studies of unconscious visual information processing, either (1) an observer with a known neurological deficit, for example, blindsight, is used to study effects of a (total or partial) deficit-induced suppression of the visibility of stimuli or else (2) a normal observer is subjected to experimentally controlled conditions that (totally or partially) eliminate the visibility or phenomenal awareness of stimuli. In both cases, the problem is how to establish that phenomenal awareness of the stimulus is absent. Before dealing
with the specific methodological rationales adopted in such studies, it pays to make some important distinctions among the types of stimulus information an observer has and lacks in her visual awareness. The issue of criterion content Criterion content refers to the information used by an observer to indicate a specific visual property of a stimulus. For instance, a stimulus is either present or absent in display; if present, it has a particular location in the display, a shape or form, luminance-contrast or lightness, color, direction of motion, speed of motion, and so on. Given the many dimensions or attributes that can specify a visual object, it is not surprising that an observer can be blind to, that is, phenomenally unconscious of, some stimulus dimensions without being blind to others. For example, a neurological patient suffering from achromatopsia or cortical color blindness would be phenomenally lacking hue distinctions but would retain phenomenal distinctions of luminance contrast, form, motion speed, motion direction, and so on. In other words, a stimulus may be visible in some parts or respects and invisible in others. Another example is that of a normal observer who, participating in a visual masking study, is unable to report the shape or the color of a masked target stimulus, but is able to report its presence or location in the visual field. Both examples illustrate that a stimulus may be visible according to one criterion content but not according to another. Thus, one can specify the existence of unconscious visual information processing along one stimulus dimension, for example, shape or color, while retaining conscious information processing of another dimension, for example, the location of the stimulus in the visual field. This point needs to be made repeatedly, since many investigators have ruled out and continue to rule out the presence of unconscious information processing simply because some information can be used either (1) to detect the mere presence or location of a stimulus even when an observer is phenomenally blind to its other qualia such as color, shape, speed of motion, depth, and so on, or (2) to correctly infer the presence of a stimulus feature that is not phenomenally experienced. For example, in studies of backward visual masking, in
Perception: Unconscious Influences on Perceptual Interpretation
which a briefly flashed target is followed by a briefly flashed mask, the shape of the target, which is suppressed from phenomenal awareness by the aftercoming mask, nevertheless can be determined indirectly by attending to subtle apparent-motion cues that arise from the spatiotemporal target-mask sequence. Based strictly on behavioral report, the observer did make correct shape discriminations and thus can be said to have seen the shapes of the targets. However, based on phenomenal report, the observer did not at all see the shape of the target, but was able to cleverly infer it on the basis of another, actually irrelevant, phenomenal dimension of motion. This illustrates that, in studies of unconscious visual information processing, appropriate weight be ascribed to subjective phenomenal reports as well as to objective behavioral responses. Indexing the absence of phenomenal consciousness Research into unconscious information processing rests on specific assumptions about what constitutes conscious and unconscious processing. One approach relies on the notion of a threshold for conscious awareness. The assumption here is that each observer has some internal level of (neural) activation, t, above which processed information registers in consciousness and below which it does not. The idea is that conscious experience simply relies on more of the same activity that is present in unconscious processing. The concept of a threshold has a long history, and threshold measurements are ubiquitous in psychophysical and perceptual research. Since consciousness is characterized by subjective, first-person experience, one way to determine threshold for consciousness is to ask observers to report on their subjective experiential state. For example, when a stimulus is presented in the affected part of the visual field of a blindsight patient, s/he reports whether or not the stimulus is experienced; or when, in a normal observer, the visibility of a target stimulus is suppressed by a mask stimulus, the observer reports whether or not s/he has seen the target. However, measurements of subjective thresholds pose problems, since, as noted, it is hard to determine what criterion an observer uses to render his binary yes/no report. To deal with this problem,
163
researchers can adopt what are presumably criterion-free or objective threshold measuring procedures. One way to accomplish this, for example, is to use a multiple-alternative (usually a twoalternative) forced-choice procedure, in which an observer is required to report, and guess if need be, during which of several possible cued time intervals the target stimulus was presented. Another example is to require the observer to indicate which of two (or more) equally likely shapes characterize the target stimulus. Other stimulus attributes such as location, color, size, luminance contrast, and so on, could also be used. Although objective threshold measures may be preferable to subjective ones, other approaches to measuring stimulus detection reject the entire notion of fixed internal thresholds. The theory of signal detection is such an approach. The notion here is that an observer’s sensory system processing a stimulus event is characterized not by some internal threshold but rather by the following assumed features: (1) the existence of (intrinsic or externally introduced) noise in the detecting system, (2) the level of noise activation varies randomly (with a Gaussian or normal probability distribution), (3) a stimulus, that is, a signal, adds a fixed amount of activity, indexed by a sensitivity parameter known as d 0 (d prime), to the noise response of the system, (4) the existence of an adjustable criterion level of activity, indexed by a parameter known as b. When an activation level produced by the signal (plus the ever present noise) is below b, the system decides that a stimulus has not been presented; when above b, it decides that a stimulus has been presented. Finally, (5) the criterion level b can be changed by contingencies based on the relative frequency of a signal occurrence or on payoffs and penalties regarding correct and incorrect decisions and the expectancies. Within this framework, conscious processing reflects a detecting system in which a stimulus produces a d 0 >> 0; whereas unconscious processing reflects a detecting system in which a stimulus produces a d 0 0. Two problems characterize this approach. One is statistical in nature. Determining that d 0 0 is equivalent to confirming the null hypothesis, a procedure that demands very strong statistical power so as to maximally reduce the probability
164
Perception: Unconscious Influences on Perceptual Interpretation
of a Type 2 error, that is, of falsely concluding that a failure to report a stimulus indicates absence of its conscious registration. The other problem is that whether or not a stimulus event is detected depends not only on d 0 and b but also, as noted above, on the criterion content; that is, on the particular stimulus information used by the observer to make his responses. For instance, the information content of interest may be the shape of a stimulus. While its shape may not register at all in consciousness, its location or presence may. Here one could wrongly conclude on the basis of one stimulus attribute or criterion content, for example, its location, that another stimulus feature, its shape, was visible when in fact it was not. This is equivalent to committing a Type 1 statistical error, that is, rejecting the valid null hypothesis that no shape was seen, when in fact it was not. The solution to this problem is to note that detection of each stimulus attribute, A, can in principle be characterized by its own sensitivity and criterion level, dA0 and bA. This can be illustrated by considering the shape or form attribute of a stimulus. As a particular example, two forms, a diamond and a square, serve as signal presented on half the trials (signal trials), and an eight-sided star (combination of diamond and square) serves as a neutral nonsignal in the other half of the trials (noise trials). On each trial an observer is asked to decide if the stimulus is one of the signals (a diamond or a square) or else the neutral star. Here, on each trial an observer may be able to see that some stimulus was presented but nonetheless not be able to see which stimulus shape it was. Thus, the d 0 for shape would be close to 0. The problem of exhaustive exclusion The above also addresses a related problem relevant to both the threshold-determining and the signal detection approaches. In order to assure that a stimulus is not consciously processed, it is deemed essential according to some current methodological accounts that measures used to assess stimulus detectability be exhaustive. That is to say, according to the exhaustiveness criterion, ideally only if all possible measures of detectability fail to yield evidence of conscious registration of a stimulus can one conclude that its processing is
unconscious. The problem here is that one can never be assured that all such measures have in fact been used. One can circumvent this problem by noting that one need not eliminate the visibility of all aspects of a stimulus but rather focus only on those aspects that are relevant. For example, recalling that if one is interested in studying unconscious processing of color, say, in an achromatopsic patient, conscious detectability of the stimulus along another attribute such as shape, surface contrast, or location need not be eliminated, unless such an irrelevant attribute provides information that can be used to infer the (unseen) color of a stimulus. Indirectly indexing the presence of unconscious information processing While the presence of conscious processing can be assessed directly by subjective report, the presence of unconscious processing, by definition, cannot. Hence, indirection is a key methodological strategy for studying unconscious processing. To realize this strategy one relies on what are known as dissociation phenomena. As an example, an achromatopsic patient who has no subjective experience of color nonetheless can, under appropriate experimental conditions, render betterthan-chance wavelength discriminations. In other words, the objectively measured presence of wavelength-discrimination abilities is dissociated from a lack of subjective color experience, indicated by his inability to name the color of a stimulus. Similarly, in a normal observer a stimulus or one of its properties can be rendered invisible by some experimental procedure; nonetheless, the unconscious processing of that stimulus or property can affect the response to a subsequently presented visible probe stimulus. Here, there is dissociation between the lack of visibility of a stimulus or of one of its properties and the presence of an influence it has on the processing of another stimulus. Under some special circumstances one can create what are known as double dissociations between the visibility of one, the inducing stimulus and its effects on another, the probe stimulus. Here, on the one hand varying the visibility of an inducing stimulus is not associated with concomitant variations of its ability to influence the visibility of, or response to, the other, probe stimulus.
Perception: Unconscious Influences on Perceptual Interpretation
That is, changes of the inducing stimulus’s visibility are dissociated from its ability to influence the processing of the probe stimulus. On the other hand, it is possible to obtain variations of the visibility of the probe stimulus without also varying the visibility of the other, inducing stimulus. That is, changes of an inducing stimulus’s influence on the visibility of the probe stimulus are dissociated from the visibility of the inducing stimulus. For instance, the physical luminance contrast of the inducing stimulus can be varied and thus produce variable effects on the visibility of, or response to, the probe stimulus, even though the visibility of the inducing stimulus, regardless of its physical luminance contrast, is suppressed by a mask. The use of normal versus neurologically impaired observers The influences of unconsciously processed visual information on perceptual interpretation can be measured by studying either neurologically impaired or normal observers. When using the former observers it is, of course, important to diagnose as thoroughly as possible the nature of the neurological impairment so as to assess which visual functions are impaired and the extent to which they are impaired. Although there may be overlap of cortical damage and impairments that manifest among different classes of patients, the classes can nonetheless be distinguished. For example, blindsight patients suffer from cortical damage that differs from that of achromatopsic (color-blind) patients, whose damage in turn differs from that of akinetopsic (motion-blind) patients, whose damage in turn differs from form-agnosic patients, and so on. Correspondingly, blindsight patients manifest blindness symptoms ranging over an extensive set of stimulus properties, whereas achromatopsic patients are blind only or primarily to color, akinetopsic patients to motion, form-agnosic patients to shapes of objects, and so on. It should be obvious that such patients can be very useful to studies of unconscious and conscious visual information processing. However, one need not rely solely on study of neurologically impaired patients. Normal, neurologically intact observers also can be used. Through application of clever experimental techniques or
165
paradigms developed and tested over the years, one can temporarily render the visual system blind to a stimulus or to one or more of its properties. Most of these techniques, some of which have been in use for and refined over decades, rely on psychophysical procedures in which one stimulus renders another one invisible. Other, more recent techniques render the cortical visual system transiently blind by producing magnetic pulses that pass through the cranium and briefly disrupt the underlying cortical visual processes. The methods are described below. Methods of rendering stimuli invisible Psychophysical methods rely on the use of experimental procedures that promote the perceptual disappearance of a stimulus, resulting in its being processed unconsciously. The influence of the suppressed stimulus on the perception of visible probe stimuli can then be assessed. Binocular rivalry
When two similar images are presented separately to the two eyes the observer typically reports having a single fused percept that is characterized by a vivid sensation of stereo depth. However, when the two images are too different or disparate, they do not allow perceptual fusion. Although this can results in diplopia, commonly known as double vision, the result is that the visibility of one of the images often is suppressed. This phenomenon can be amplified by presenting very disparate images to the two eyes, for example, a vertical grating to the left eye and a horizontal grating to the right eye, as shown in the first, top panel of Figure 1. The resulting percept is not a fused grid but rather rivalrous perceptual fluctuations. Here, the image presented to, say, the left eye dominates perceptually for a short period while simultaneously the image to the right eye is perceptually suppressed; then the image to the right eye dominates while that to the left is suppressed, and so on. While the image to one eye is suppressed, a test stimulus presented to that eye will also be perceptually suppressed. Perceptual fading of semistabilized images
In order to maintain visibility of a stimulus, its image on the retina must regularly shift over the
166
Perception: Unconscious Influences on Perceptual Interpretation
Binocular rivalry Stimulus presentation Left eye
Perception
Right eye or else
Semi-stabilized images +
Visual (two-transient) masking Stimulus presentation Target Mask
Perception
or
C
or
Time Visual crowding +
K
+
X N KF H
Figure 1 First or top panel: on the left is a schematic representation of two typical stimuli presented to the separate eyes to induce binocular rivalry; on the right: depiction of alternating eye-dependent percepts produced during binocular rivalry. Second panel: depiction of a visual display used to produce semistabilized retinal images. The observer should fixate the cross as rigidly as possible and, without moving his or her eyes, note what happens to the two faint stimuli to the left and right of fixation. Third panel: schematic depiction of typical visual stimuli and their time sequence used to produce visual masking. The first, target stimulus, a letter C, can be followed by the second, mask stimulus which can either be a surrounding stimulus such as a larger letter O or a spatially overlapping stimulus composed of random – in this case – arc-like elements. Fourth or bottom panel: on the left a depiction of a visual display consisting of a fixation cross and an isolated target letter, here a K, to the right of fixation, and on the right the same display but now with four other letters crowding the target letter.
retinal receptors; otherwise it fades and disappears from sight. One can exploit this phenomenon by, for instance, having the observer maintain his visual fixation as constantly as possible on, say, a tiny cross in the center of his visual field while two other stimuli are present at some distance to the right and left from the fixation cross, as shown in the second panel of Figure 1. What typically occurs when one fixates as rigidly as possible on the cross is that eventually one or both of the two eccentric stimuli will perceptually fade from view for a while, then reappear, then fade from view again, and so on. This type of fading as well as the aforementioned binocular-rivalry suppression can be amplified or prolonged by several additional methods listed below.
Motion-induced blindness
When, an array of, say, three stationary yellow dots, located at the vertices of a notional triangle is presented within in a larger array of coherently moving blue dots, the three yellow dots will, after a small while fade either singly, pairwise, or as a unit. Flash suppression
As with binocular-rivalry suppression, the visibility of semistabilized images fades and reappears according to a more or less random, stochastic schedule. For that reason the exact timing of the fading cannot be determined. The suppression of the visibility of semistabilized as well as rivalrous images can, however, be temporally triggered and
Perception: Unconscious Influences on Perceptual Interpretation
prolonged by transiently flashed stimuli. For instance, in procedure known as continuous flash suppression, the rivalrous stimulus presented to one eye can be suppressed continuously by a stimulus flashed repeatedly to the other eye. Similarly, the fading of an eccentric semistabilized image can be triggered and prolonged by a briefly presented nearby stimulus. The visual stimulus transients presented either as a single flash or as a train of repeated flashes act as visual masks that suppress the visibility of a rivalrous or a semistabilized image that is physically present for a fairly long duration, sometimes on the order of a minute or so. Visual (two-transient) masking
However, there also is a way to suppress the visibility of a stimulus presented briefly, say, on the order of a few tens of milliseconds. The typical procedure, as shown in the third panel of Figure 1, is to briefly present one stimulus, designated as the target, which is followed at brief intervals by a second stimulus, known as the mask. This procedure is known as backward masking. The mask can either spatially overlap the target or surround it. In either case, by choosing appropriate stimulus parameters, the effect of the second, mask stimulus is to suppress the visibility of the first, target stimulus. Crowding
Another masking phenomenon is produced by visual crowding. This is illustrated in the fourth panel of Figure 1. Here a target stimulus is presented, usually for a relatively long duration of several seconds, at some distance from central fixation. When the stimulus is presented in isolation, that is, without being crowded by nearby, adjacent stimuli, it is clearly recognizable. However, when surrounded by other visual stimuli it is not easily or at all recognizable. It is important to note, however, that its visibility as such is not suppressed. Although some thing is seen, its identity cannot be determined. Attentional limits
It is known that the effectiveness of visual (twotransient) masking and crowding is modulated
167
by attentional factors. The visibility or the recognizable identity of a target stimulus is more readily masked when attention is withdrawn from it. Inattentional blindness demonstrates that an otherwise easily detectable target fails to be detected when attention is diverted to another, simultaneous task. Several other procedures that rely on the withdrawal or diversion of attentional resources can be used to suppress the identification or the visibility of a stimulus. One exploits a phenomenon known as the attentional blink. Here, several stimuli such as alphanumeric characters are presented at the same display location one at time in a rapid sequence, each stimulus flashed on the order of 100 ms. Embedded within a sequence of numeric characters are two letters of the alphabet that serve as targets. The observer is required to report the identity of the two letter targets. What typically happens is that if the first letter is identified correctly, the second letter, when presented, say, 200–400 ms later, tends not to be identified. Presumably, the attention devoted to the processing of the first stimulus after it is presented is thereby effectively withdrawn for a couple of 100 ms from the processing of the second target. Another expression of attentional limits manifests in a phenomenon known as change blindness. For example, imagine viewing a series of two alternating depictions of a scene, such as an urban street scene, in which there is a noticeable difference between the two depictions. The change or difference can be a deletion of some significant object in, or some portion of, the scene, an addition to it, or both. As long as the scene is fairly complex, that is, contains many visual elements and features, attention cannot be focused continuously on all of its contents. Hence changes introduced into the scene, such as a series of substitutions or deletions and reinsertions of elements, will go undetected in a large proportion of observers. Transcranial magnetic stimulation
Finally a recent technique for rendering stimuli invisible that is similar to the two-transient visual masking technique or to continuous flash suppression, is to apply a magnetic pulse or a longer lasting series of magnetic pulses to the cranium overlying a particular part of the visual cortex.
168
Perception: Unconscious Influences on Perceptual Interpretation
This technique, known a transcranial magnetic stimulation (TMS), causes a disruption of neural activity in cortical tissue located directly below the cranial point of application. For instance, when applied to the midoccipital region of the cranium at the back of the head, TMS can produce a blindness near the foveal part of the visual field by disrupting neural activity in that area of the underlying primary (or secondary) visual cortex responsible for processing information received from the near-foveal region of each eye’s retina.
Findings and Their Theoretical Interpretations It is generally agreed among neuroscientists that conscious perception of a stimulus requires that some informational content about it be processed at relatively high cortical levels. For instance, to perceive its shape, activity in the lateral occipital cortex, which supports object recognition, must be present; while to perceive its motion, activity in the medial temporal region of the visual cortex must be present. Both of these areas of visual cortex are separated from the primary visual cortex by one or more synaptic junctures. This does not mean that lower levels of visual processing are unimportant. As noted above, blindsight, the absence of visual object qualia in the phenomenal field, results from damage to the primary visual cortex, the first anatomic site of cortical visual processing. By extension, elimination of conscious vision also results from damage to sites, such as the lateral geniculate nucleus in the thalamus, anatomically located even lower in the visual system. As a consequence there are many loci in the human visual system where unconscious information processing occurs. As an obvious and trivial example, the retinas of the two eyes also are sites where unconscious processing occurs. However, there the unconscious neural activity as such (i.e., without activating higher levels of processing) has no effect on perception or behavior, and thus cannot be assessed. The trick then is to find levels of processing in the human visual system that are unconscious and that can affect perception or behavior.
Delimiting the Sites of Postretinal Unconscious Processing in the Human Visual System It is noted above that there is no one and only way to render a stimulus or a part of it inaccessible to consciousness. Perhaps more than coincidentally, at postretinal levels of processing there is no single site of unconscious processing of stimuli. However, in some cases the relation between method of rendering stimuli invisible and the nature of unconscious processing may be specified. Thus, it is possible, at least in principle, that two (or more) different ways of rendering a stimulus or its attributes inaccessible to consciousness can be associated with two (or more) different types or levels of postretinal processing. The following, besides giving examples of such associations between method and finding, explores the different levels and types of unconscious visual processing. Unconscious visual processing in neurological patients Blindsight Patients with damage to the primary visual cortex, also known as striate cortex, fail to spontaneously report the phenomenal presence of most visual stimuli in the defective part of the visual field. Yet, when studied carefully they exhibit two methodologically distinct types of visual function. In the first type, one visual stimulus presented in the blind area can prime the response to a second stimulus presented in the sighted area of the visual field. For instance, a colored stimulus presented in the blind field can prime the response to a second colored stimulus presented in the sighted part of the field. In the second type, significantly betterthan-chance performance is obtained when stimuli are presented within the visual field defect and the patients are required to guess whether a stimulus was presented, where a stimulus was presented, or which of several stimuli was presented. Although they lack experiences of qualia (except occasionally a coarse phenomenal sense of motion), when pressed to make discriminative responses these patients are capable of residual unconscious vision. Among indices of preserved visual functions are (1) discriminating the presence of a briefly flashed stimulus from its absence,
Perception: Unconscious Influences on Perceptual Interpretation
169
Form blindness
have residual visual function. An example is the aforementioned achromatopsic patient who fails to report any phenomenal awareness of color, yet can, if prompted, discriminate among different wavelengths of light. Cases of blindsight, form blindness, achromatopsia, and so on, indicate why, as suggested above, the notion of access consciousness is problematic. These patients, although phenomenally blind to a stimulus attribute, have access to residual visual information about that stimulus dimension that can control their behavioral responses. Thus, by definition they have access consciousness of the stimuli. However, it remains puzzling as to the way they can be conscious of the accessed stimulus information without being phenomenally conscious of it. Most researchers in visual cognition fail to stretch the definition of consciousness beyond phenomenal awareness, and they regard access consciousness of the type described above as examples of unconscious processing.
Some patients have damage at poststriate or extrastriate stages of cortical visual processing. In particular, a number of patients have damage to an area of extrastriate cortex called the lateral occipital complex. This area is crucial to the perception of the form or shape of objects. Although these patients fail to report phenomenal awareness of the form of these stimuli, the form can nevertheless control motor behavior. While they fail to report the orientation of a rectangular elongated slot or aperture, they can adjust their hand appropriately when asked to insert a card held in their hand into the slot. For instance, if the length of the slot has a vertical orientation, the patient will attempt to insert the card by holding it vertically; whereas, if the orientation of the slot is changed to, say, a left oblique one, the orientation of the card held in the hand will be adjusted accordingly. Such patients also can adjust the size of their hand’s grip in anticipation of grabbing an object that fails to appear in phenomenal awareness. For instance, grabbing a pencil will result in a narrower grip than grabbing a dictionary. Many more such residual unconscious controls of behavior have been documented in such patients. There are many other types of neurological patients – too many to exhaustively describe – who lack specific phenomenal qualia yet who
Unconscious visual processing in normal observers Effects of binocularly rivalry suppression Despite reports of global or high-level perceptualgrouping factors modulating binocular rivalry suppression and amplification of rivalry suppression along the ascending cortical processing pathways, increasing evidence indicates that binocular rivalry is by and large eye-specific and is initiated powerfully at relatively low levels of cortical processing. This suggests that the visual information that survives the suppressive effects is processed unconsciously at low levels. Thus, although some information giving rise to the detection of apparent motion, to orientation-specific adaptation effects, and to the production of afterimages survives binocularly rivalry suppression, no semantic word priming, picture priming, or color priming can be produced by stimuli presented to an eye during its suppressed phase. On the other hand, responses in the emotional centers of the brain can be obtained when pictures of emotionally expressive faces are presented to the suppressed eye. In these respects binocular rivalry has effects similar to blindsight, as would be expected since blindsight also is caused by disruption of neural activity at low cortical levels of processing.
(2) localizing, via saccades or hand pointing, a stimulus presented at variable locations in the blind area of the visual field, (3) discriminating among speeds and directions of moving stimuli, (4) rudimentary wavelength discrimination, (5) the ability of an unseen stimulus to cue temporal and spatial shifts of attention, (6) coarse orientation and low-spatial frequency pattern discrimination, (7) appropriate responses to a variety of facial emotional expressions, and (8) report of phenomenal awareness of chromatic and figural properties of the afterimage of stimuli presented into the visual field defect. Exactly where and how in the visual system these residual abilities are expressed is not certain, although it appears that direct pathways from the lateral geniculate nucleus to cortical extrastriate cortex, thus circumventing the primary visual cortex, and the subcortical pathways including the superior colliculus and pulvinar are involved.
170
Perception: Unconscious Influences on Perceptual Interpretation
Effects of motion-induced blindness
The unconsciously processed information of stimuli that are rendered invisible during motioninduced blindness can produce several perceptual effects. For one, the visually suppressed stimuli are still able to produce visible afterimages. Moreover, when changes are introduced to invisible stimuli during motion-induced blindness, these unseen changes can nevertheless exploit perceptual Gestalt grouping factors which update subsequently seen object representations by influencing not only how stimuli reenter into phenomenal awareness but also what objects reenter. Regarding the former effect of afterimage formation, the residual effects of unseen stimuli during motioninduced blindness are similar to the effects of unseen stimuli in blindsight. Effects of continuous flash suppression
Continuous flash suppression can yield effects distinct from standard binocular-rivalry suppression. For instance, whereas standard methods producing alternating periods of binocular-rivalry dominance and suppression had little or no effect on a suppressed stimulus’s ability to produce afterimages, continuous flash suppression, perhaps by deepening and prolonging the effects of binocular-rivalry suppression, does measurably affect the suppressed stimulus’s ability to produce perceivable afterimages. Effects of visual masking
A stimulus, called the prime, whose visibility is suppressed by a subsequent stimulus, called the mask (backward masking), can, despite its being not seen, affect the perceptual response to a subsequently presented probe stimulus. In particular, a masked, unconsciously processed prime can affect the response to a probe’s color, form, and meaning. These results are interesting when considered in relation to the unconscious processing during binocular-rivalry suppression, since a stimulus that is suppressed during binocular rivalry fails to yield priming by its shape, color, or meaning. This suggests that the suppressive effects of masking are realized at a level of processing that is subsequent to, or higher than, than the level of processing at which the suppressive effects binocular-rivalry suppression occur. This is consistent
Target
15–30 ms
Mask
200 ms
Stimuli
Percept Figure 2 A schematic of typical stimuli and their
temporal sequence used in studies of feature inheritance.
with the finding that the mechanism responsible for the suppressive effects of backward masking can in turn be suppressed during binocular rivalry. An interesting phenomenon that can accompany backward masking is what is known as feature inheritance. Here, as shown in Figure 2, the visibility of a target stimulus composed, say, of two misaligned vertical elements is suppressed by an array or grating composed of aligned vertical elements presented about 30 ms later. However, although the grating contains aligned vertical elements, it is perceived as if the elements in it had been misaligned in accordance with the actual misalignment of the suppressed target. Here, the unconsciously processed misalignment feature of the target is perceptually inherited by the aligned mask grating. It is possible that such feature inheritance may contribute to the priming effects produced by a masked prime. Effects produced by TMS
Magnetic stimulation applied to the occipital pole locally disrupts neural activity in the underlying primary visual cortex. In this respect, it can be regarded as transiently inducing a state similar to blindsight. In fact, its effects are also similar in that the wavelength and orientation of a stimulus whose visibility is suppressed can still be discriminated. Moreover and also similar to blindsight, stimuli whose visibility is suppressed by TMS can nonetheless direct saccadic eye movements to their unseen location in the visual field. It is possible that other effects of unconsciously processed information obtained in blindsight patients will also be obtained with suppression produced by TMS.
Perception: Unconscious Influences on Perceptual Interpretation
Effects of attentional limits
As noted above, limits on attentional resources can be experimentally imposed in a variety of ways. It has been shown that stimuli that fail to access conscious report during the attentional blink can nevertheless prime the perceptual processing of subsequently presented probe stimuli. Priming effects can also be produced during change blindness by both the prechange stimuli and the postchange stimuli that are unavailable to conscious report. Moreover, blindness to the prechange stimuli is associated with the limited duration of iconic memory and thus, as noted above, may reflect a failure of access consciousness. The Role of Attention in Unconscious Visual Processing It was noted above that blindsight patients can direct attention to the location of a stimulus presented in the visual field defect. It is also becoming clear that attention can modulate the effects produced by unconsciously processed stimuli. Such attentional modulation has been found in the unconscious processing of stimuli whose visibility is suppressed during backward masking and during visual crowding. Nonetheless some visual stimuli, such as natural scenes or faces and their emotional expressions, whose visibility is suppressed by an aftercoming mask seem to require little if any attention in order to be processed unconsciously. Natural scenes and faces occur very often in our normal everyday commerce with the visual world. In addition faces conveying emotional expressions are particularly salient social stimuli. For any of these reasons such stimuli might be processed automatically with little or no need of attentional modulation. On the other hand, the unconscious processing of abstract and nonecological stimuli typically used in laboratory experiments may be subject to attentional modulation since they are both less familiar and less salient. The Role of Feedforward and Reentrant Feedback Connections In current approaches to visual cognition and neuroscience, prominence is given to the distinction between feedforward and reentrant or feedback
171
connections in cortical neural networks. According to some theoretical approaches, the multisynaptic feedforward transmission and transformation of visual information along successive levels of cortical processing is not sufficient to give rise to a conscious percept. Also deemed necessary are the feedback connections from higher levels of processing to lower ones. Accordingly, much or most of the unconscious visual information processing can theoretically be accomplished during the feedforward sweep of cortical neural activity. This can give rise to, for example, the unconscious priming of the processing of stimulus attributes such as form, color, and even semantic content. Suppression of the feedforward flow of information at low levels, such as the suppression produced during binocular rivalry, would therefore leave less information processing intact than suppression produced at higher levels of the feedforward flow, such as the suppression produced by visual backward masking. Moreover, if the feedforward sweep of cortical processing is indeed largely responsible for unconscious processing, it follows that neural activity at higher levels of cortical processing will code more complex, relational stimulus properties than do the lower levels. In addition, some unconsciously processed information may proceed along an alternate, subcortical feedforward route, since there are direct connections from the retina to the superior colliculus of the midbrain and hence, via the pulvinar, to cortical extrastriate regions and to the emotional centers in the limbic system including the amygdala. Such feedforward pathways could be responsible not only for the residual motion perception in blindsight patients but also the activation of emotional centers by stimuli rendered invisible either in blindsight or by psychophysical methods.
Conclusions Much evidence has accumulated over the past three decades indicating the existence of unconscious information processing in the human visual system. The evidence, however, must be weighed in light of important methodological and conceptual distinctions. The most important methodological issues concern the adoption of appropriate
172
Perception: Unconscious Influences on Perceptual Interpretation
criterion contents and of adequate rationales and procedures used to assure the suppression of the conscious registration of visual stimuli. Different neurological conditions and different experimental methods tap into different types and levels of unconscious information processing, and one of the tasks of future research is to work out the relationship among these different conditions and methods and the correspondingly different types and levels of unconscious processing. See also: Attention: Change Blindness and Inattentional Blindness; Attention: Selective Attention and Consciousness; Concepts and Definitions of Consciousness; Implicit Learning and Implicit Memory; Perception, Action, and Consciousness; Perception: Subliminal and Implicit; Phenomenology of Consciousness; Unconscious Cognition.
Suggested Readings Alais D and Blake R (2005) Binocular Rivalry. Cambridge, MA: MIT Press. Breitmeyer BG and O¨g˘men H (2006) Visual Masking: Time Slices through Conscious and Unconscious Vision. Oxford: Oxford University Press. Breitmeyer BG, Ro T, and O¨g˘men H (2004) A comparison of masking by visual and transcranial magnetic stimulation: Implications for the study of conscious and unconscious visual processing. Conscious and Cognition 13: 829–843. Holender D (1986) Semantic activation without conscious identification in dichotic listening, parafoveal vision, and
visual masking: A survey and appraisal. Behavioral and Brain Sciences 9: 1–66. Kihlstrom JF (1996) Perception without awareness of what is perceived, learning without awareness of what is learned. In: Velman V (ed.) The Science of Consciousness, pp. 23–46. London: Routledge. Kim C-Y and Blake R (2005) Psychophysical magic: Rendering the visible ‘invisible.’ Trends in Cognitive Neuroscience 9: 381–388. Koch C (2004) The Quest for Consciousness. Englewood, CO: Roberts & Co. Kouider S and Dehaene S (2007) Levels of processing during non-conscious perception: A critical review of visual masking. Philosophical Transactions of the Royal Society B 362: 857–875. Mack A and Rock I (1998) Inattentional Blindness. Cambridge, MA: MIT Press. Merikle PM and Daneman M (1998) Psychological investigations of unconscious perception. Journal of Consciousness Studies 5: 5–18. Neumann O and Klotz W (1994) Motor responses to nonreportable, masked stimuli. Where is the limit of directparameter specification. In: Umilta C and Moscovitch M (eds.) Attention and Performance XV, pp. 123–150. Cambridge, MA: MIT Press. O¨g˘men H and Breitmeyer BG (2005) The First Half Second: The Microgenesis and Temporal Dynamics of Unconscious and Conscious Visual Processes. Cambridge, MA: MIT Press. Simons DJ and Rensink RA (2005) Change blindness: Past, present, and future. Trends in Cognitive Sciences 9: 16–20. Stoerig P (1996) Varieties of vision: From blind responses to conscious recognition. Trends in Cognitive Neurosciences 19: 401–406. Weiskrantz L (1997) Consciousness Lost and Found: A Neuropsychological exploration. Oxford: Oxford University Press.
Perception: Unconscious Influences on Perceptual Interpretation
173
Biographical Sketch
Bruno G Breitmeyer is a professor of psychology at the University of Houston. He studied mathematics at the University of Illinois, Champaign-Urbana, and received his PhD in psychology from Stanford University. He did his postdoctoral research at the Bell Laboratories, Murray Hill, NJ and, as an Alexander von Humboldt fellow, at the Neurological Clinic of Freiburg University, Germany. His research interests are in visual cognition and neuroscience. Primary research activities include investigations of the temporal dynamics of visual masking, object perception, and attention. In recent years, his research, in collaboration with his ¨ g˘men, has focused on the use a variety of masking techniques to study the types and University of Houston colleague, Haluk O levels of conscious and unconscious visual information processing. He has published numerous scientific journal articles and book chapters and coauthored three books dealing with these topics. He has also organized major workshops, served on various journal editorial boards and has received several awards, among them the Science Citation Award from the Institute of Scientific Information, the Alexander von Humboldt Fellowship, and a Distinguished Alumnus Award from the University of Illinois.
Phenomenology of Consciousness D Zahavi, University of Copenhagen, Copenhagen, Denmark ã 2009 Elsevier Inc. All rights reserved.
Glossary Externalism – The view that what is thought and experienced is essentially dependent on aspects of the world external to the mind of the subject. Intentionality – The directedness or of-ness or aboutness of consciousness; when one perceives or judges or feels or thinks, one’s experience is about or of something. Internalism – The view that a subject’s beliefs and experiences are wholly constituted by what goes on inside the mind of that subject. Intersubjectivity – The relation between experiencing subjects. Lived body – The body understood as an embodied first-person perspective. Prereflective self-consciousness – The ongoing first-personal manifestation of experiential life. Transcendental – A philosophical reflection on the conditions of possibility of experience and cognition.
Introduction Until recently, many scientists considered the study of consciousness to be inherently unreliable due to its subjective nature and thus unsuitable for scientific research. Since the early 1990s, however, a marked change has taken place, one occasionally described in terms of an ongoing ‘consciousness boom.’ Many new journals devoted to the study of consciousness have been established and currently questions pertaining to the nature of phenomenal consciousness, the structure of the first-person perspective, and the status of the self are once again considered crucial for a scientific understanding of consciousness.
Given the recent interest in the subjective dimension of consciousness, it is no wonder that many have started to emphasize the importance of phenomenology. Thus, when studying consciousness rather than, say, deep-sea ecology, we should provide a phenomenologically sensitive account of the various structures of experience, since an important and nonnegligible feature of consciousness is the way in which it is experienced by the subject. The aim of the following contribution is to outline and discuss some central features of a phenomenological account of consciousness. In this context, however, phenomenology will not refer to a form of psychological self-observation, but rather to a specific philosophical tradition – a tradition inaugurated by Husserl (1859–1938), and comprising among its most well-known champions, philosophers like Max Scheler, Martin Heidegger, Aaron Gurwitsch, Maurice MerleauPonty, and Jean-Paul Sartre. Phenomenology counts as one of the important philosophical traditions in the twentieth century, and has made significant contributions to most areas of philosophy, including philosophy of mind, social philosophy, philosophical anthropology, aesthetics, ethics, philosophy of science, epistemology, theory of meaning, and formal ontology. Within the study of consciousness, phenomenology has in particular provided analyses of such topics as intentionality, perception, embodiment, self-consciousness, sociality, and temporality. Like any other tradition, phenomenology spans many differences, and has developed as a heterogeneous movement with many branches. For the same reason, it will be impossible to do justice to all aspects of phenomenology in a short overview like the present. Rather, the focus will be on a few selected topics; topics deemed to be of particular importance for contemporary discussions in philosophy of mind and cognitive science. Furthermore, rather than spending time articulating the
175
176
Phenomenology of Consciousness
differences between the various figures in phenomenology – differences that have frequently been overstated – the emphasis will be on some of the overarching concerns and common themes that have united its proponents.
Methodological Considerations Edmund Husserl’s maxim for phenomenology was ‘Back to the things themselves!’ By this he meant that phenomenology should base its considerations on the way things are experienced rather than by various extraneous concerns which might simply obscure and distort that which is to be understood. In short, phenomenology should not start with theory; it should be critical and nondogmatic, shunning metaphysical and theoretical prejudices, as much as possible, and seeks to be guided by that which is actually experienced, rather than by what we expect to find given our theoretical commitments. It asks us not to let preconceived theories determine our experience, but to let our experience inform and guide our theories. To the extent that phenomenology stays with experience it is said to take a first-person approach. That is, the phenomenologist is concerned to understand the experience in terms of the meaning it has for the subject. It does not seek to disclose the subpersonal mechanisms that might enable us to experience the way we do. To put it differently, phenomenology is concerned with attaining an understanding and proper description of the experiential structure of our experiential life; it does not attempt to develop a naturalistic explanation of consciousness, nor does it seek to uncover its biological genesis, neurological basis, psychological motivation, or the like. To fully appreciate the philosophical import of the phenomenological account of consciousness, it is, however, important to understand that the goal of phenomenology is not to describe each and every idiosyncratic experience – ‘‘here and now, this is what I experience’’ – rather, it aims to capture invariant and universal structures of experience; structures that are intersubjectively valid and accessible. For the very same reason, its analyses are open for corrections and control by any (phenomenologically tuned) subject. Furthermore,
it is important to realize that phenomenology – notwithstanding various other differences to said tradition – is in some respects firmly situated within a certain Kantian or post-Kantian framework. One way to interpret Kant’s revolutionary Copernican turn in epistemology is by seeing it as amounting to the claim that our cognitive apprehension of reality is more than a mere mirroring of a preexisting world, for which reason a philosophical analysis of reality, a reflection on what conditions something must satisfy in order to count as ‘real,’ cannot ignore the contribution of consciousness. Thus, and this pinpoints a main difference to at least a good part of recent analytic philosophy’s preoccupation with consciousness, the phenomenological interest in the first-person perspective is not primarily motivated by the relatively trivial insight that we need to include the first-person perspective if we wish to understand mental phenomena. Rather, the phenomenologists’ focus on the first-person perspective is as much motivated by an attempt to understand the nature of objectivity, as by an interest in the subjectivity of consciousness. Indeed, rather than taking the objective world as the point of departure, phenomenology asks how something like objectivity is possible in the first place. What are the primitive modes of understanding that precede our scientific understanding? How is objectivity constituted? In phenomenological texts the term ‘constitution’ is a technical one. Constitution must be understood as a process that allows for the manifestation or appearance of objects and their signification, that is, it is a process that permits that which is constituted to appear, to manifest and present itself as what it is. And this process is something that in significant ways involves the contribution of consciousness. Objects are constituted, that is, experienced and disclosed in the ways they are thanks to the way consciousness is structured. Phenomenologists consequently reject the suggestion that consciousness is merely one object among others in the world, on par with – though possibly more complex than – volcanoes, waterfalls, ice crystals, gold nuggets, rhododendrons, or black holes, since they would consider it to be a necessary (though not sufficient) condition of possibility for any entity to appear as an object in the way it does and with the meaning it has.
Phenomenology of Consciousness
Husserl occasionally speaks of the phenomenological attitude as consisting in a reflective move that takes a step back from our naı¨ve and unexamined immersion in the world in order to focus on the way in which the world manifests itself to us. In other words, once we adopt the phenomenological attitude, we are no longer primarily interested in what things are – in their weight, size, chemical composition, etc. – but rather in how they appear, and thus as correlates of our experience. Indeed to put it slightly paradoxically, phenomenologists are not interested in consciousness per se. They are interested in consciousness because they consider consciousness to be our only access to the world. They are interested in consciousness because it is world-disclosing. Contrary to certain misunderstandings, the phenomenological reflection does not amount to an inward turn, as if its aim is to make us ignore the world in favor of consciousness. On the contrary, it permits us to investigate the world we live in from a new reflective attitude, namely in its significance and manifestation for consciousness. Although this reflective investigation differs from a straightforward exploration of the world, it remains an investigation of reality; it is not an investigation of some otherworldly, mental realm. If we wish to describe the experiential difference between tasting wine and tasting water, between perceiving and imagining a rhinoceros, we do not sever our intentional link with the world. Rather, we discover these differences, and we analyze them descriptively by paying attention to how worldly objects and states of affairs appear to us. When we perceive, judge, or evaluate objects, a thorough phenomenological examination will lead us to the experiential structures and modes of understanding to which these types of appearance are correlated. We are led to the acts of presentation – the perception, judgment, or valuation – and thereby to the experiencing subject (or subjects) in relation to whom the object as appearing must necessarily be understood. By adopting the phenomenological attitude we pay attention to how public objects (trees, planets, paintings, symphonies, numbers, states of affairs, social relations, etc.) appear. But we do not simply focus on the objects precisely as they appear; we also focus on the subjective side of consciousness,
177
thereby becoming aware of our subjective accomplishments and of the intentionality that is at play. In short, phenomenology must be understood as a philosophical analysis of the different types of world-disclosure (perceptual, imaginative, recollective, etc.), and in connection with this as a reflective investigation of those structures of experience and understanding that permit different types of beings to show themselves as what they are.
Consciousness and Self-Consciousness When it comes to the relation between consciousness and self-consciousness, literally all of the major figures in phenomenology defend the view that a minimal form of self-consciousness is a constant structural feature of conscious experience. Experience happens for the experiencing subject in an immediate way and as part of this immediacy, it is implicitly marked as my experience. For the phenomenologists, this immediate and first-personal givenness of experiential phenomena must be accounted for in terms of a ‘prereflective’ self-consciousness. Self-consciousness on this account is not merely something that comes about the moment one scrutinizes one’s experiences attentively (let alone something that only comes about the moment one recognizes one’s own mirror image, refers to oneself using the first-person pronoun, or is in possession of a theory of mind, or an identifying knowledge of one’s own life story). Rather, selfconsciousness comes in many forms and degrees. It makes perfect sense to speak of self-consciousness whenever I consciously perceive an external object – an oak, a table, or the moon – because to consciously perceive something is not simply to be conscious of the perceptual object, but also to be acquainted with the perception of the object. The difference between a nonconscious perception of a sunset, and a conscious perception of the sunset, is that there is something it is like to consciously perceive a sunset. Although my attention is on the object, the experience itself remains conscious. Not in the sense that I am thematically aware of it, but in the sense that there is something it is like to be in that state, it has phenomenal
178
Phenomenology of Consciousness
qualities. The notion of prereflective selfconsciousness is consequently related to the idea that experiences have a subjective ‘feel’ to them, a certain (phenomenal) quality of ‘what it is like’ or what it ‘feels’ like to have them. As it is usually expressed in analytical philosophy of mind, to undergo a conscious experience necessarily means that there is something it is like for the subject to have that experience. One reason experiences are said to be subjective is because they are characterized by a subjective mode of existence in the sense that they necessarily feel like something for somebody. Our experiential life can consequently be said to entail a primitive form of for-me-ness. This for-me-ness is not a quality like yellow, salty, or spongy. It does not refer to a specific content of experience, to a specific what, but to the unique mode of givenness or how of experience. It refers to the first-personal givenness of experience. In its most primitive and fundamental form, self-consciousness is a question of the ongoing first-personal manifestation of experiential life. This claim should not be misunderstood. The phenomenologists are not advocating a strong thesis concerning total and infallible self-knowledge; rather they are calling attention to the constitutive link between experiential phenomena and the firstperson perspective. When speaking of a first-person perspective it is important to be clear about the distinction between having or embodying such a perspective and being able to articulate it linguistically. Whereas the latter presupposes mastery of the first-person pronoun the former is simply a question of the subjective manifestation of one’s own experiential life. Although both capabilities deserve to be investigated, phenomenologists have mainly been accentuating the significance of the former. To emphasize the importance of the firstperson perspective is consequently simply to stress that there is a distinctive way experiential episodes present themselves to the subject whose episodes they are. Phenomenologists explicitly deny that the self-consciousness that is present the moment I consciously experience something is to be understood in terms of some kind of reflection, or introspection, or higher-order monitoring. That is, the self-consciousness in question is not a new consciousness. It is not something added to the experience,
an additional mental state, but rather an intrinsic feature of the primary experience. Thus, when phenomenologists speak of self-consciousness as a permanent feature of consciousness, they are not referring to what is usually called reflective selfconsciousness. Reflection (or higher-order monitoring) is the process whereby consciousness directs its intentional aim at itself, thereby taking itself as its own object. But on the phenomenological account such a form of objectifying self-consciousness is derived. It presupposes the presence of the more fundamental prereflective self-consciousness. The latter form of selfconsciousness is not thematic or attentive or voluntarily brought about; rather it is tacit, and very importantly, thoroughly nonobservational (i.e., it is not a kind of introspective observation of myself) and nonobjectifying (i.e., it does not turn my experience into a perceived or observed object). I can, of course, reflect on and attend to my experience, I can make it the theme or object of my attention, but prior to reflecting on it, I was not mind- or self-blind. The experience was already something for me, and in that sense it counts as being prereflectively self-conscious. It is because of their conviction that consciousness is as such characterized by a primitive, tacit, self-consciousness, that many phenomenologists have also found it appropriate to ascribe a primitive sense of self to the experiential phenomena. The basic idea is that an understanding of what it means to be a self calls for an examination of the structure of experience, and vice versa. Thus, on this minimal definition the self is not something that stands opposed to the stream of consciousness, it is not an ineffable hypothesized precondition, nor is it a mere social construct that evolves through time; it is taken to be an integral part of the structure of our conscious life. More precisely, they argue that the self possesses experiential reality, and link it to the first-personal givenness of the experiential phenomena. In short, the self is conceived as the ubiquitous dimension of first-personal givenness in the multitude of changing experiences. Incidentally, this view makes it clear that selfexperience is not to be understood as an experience of an isolated, worldless self; nor is the self located and hidden in the head. To have a selfexperience is not to interrupt the experiential
Phenomenology of Consciousness
interaction with the world in order to turn one’s gaze inward; on the contrary, at its most primitive self-experience is simply a question of being prereflectively aware of one’s own intentional consciousness. It is – as we shall see in a moment – always the self-experience of an embodied and embedded mind. It would consequently be a mistake to interpret the phenomenological notion of a minimal self as a Cartesian-style mental residuum, that is, as some kind of self-enclosed and selfsufficient interiority. The phenomenological notion of self is fully compatible with a strong emphasis on the fundamental intentionality, or being-in-theworld, of consciousness.
Intentionality The concept of intentionality has a long history that stretches back at least as far as Aristotle. It played a central role in medieval scholastic epistemology, but the modern revival of the term ‘intentionality’ is due to Brentano. Intentionality has to do with the directedness or of-ness or aboutness of consciousness, that is, with the fact that when one perceives or judges or feels or thinks, one’s experience is about or of something. The phenomenologists have primarily been interested in intentionality as a decisive feature of consciousness. Moreover, they have specifically focused on an account of intentionality from the first-person perspective, that is, from the subject’s point of view. In fact, none of the phenomenologists have been engaged in a naturalization of intentionality, if that is understood as an attempt to explain intentionality reductively by an appeal to nonintentional mechanisms and processes. If one thinks that a theory of intentionality must issue in a reductive account one will be bound to find the phenomenological treatment of intentionality disappointing. What is the aim of the phenomenological account of intentionality then? First and foremost, to provide a descriptive analysis of intentional the structures of consciousness. In doing so, however, the phenomenologists also seek to clarify the relation between mind and world (rather than the relation between mind and brain). This latter investigation, which basically intends to demonstrate the world-involving character of the mind and
179
rejects the view that consciousness is a subjective sphere that exists independently from the world that is revealed through it, has some more overarching philosophical implications. Husserl’s ‘Logical Investigations’ contains the first proper phenomenological investigation of intentionality. Like Brentano, Husserl argues that one does not merely love, fear, see, or judge, one loves a beloved, fears something fearful, sees an object, and judges a state of affairs. Regardless of whether we are talking of a perception, thought, judgment, fantasy, doubt, expectation, or recollection, all of these diverse forms of consciousness are characterized by intending objects, and cannot be analyzed properly without a look at their objective correlate, that is, the perceived, doubted, expected object. The converse is also true: The intentional object cannot be analyzed properly without a look at its subjective correlate, the intentional act. Neither the intentional object nor the experience that intends it can be understood apart from the other. Acts of consciousness and objects of consciousness are essentially interdependent: the relation between them is an internal rather than an external one. That is to say, from the perspective of phenomenology one cannot first identify the items related and then explore the relation between them. Rather one can identify each item in the relation only by reference to the other item to which it is related. It is customary to speak of intentional ‘relations’ as being aspectual or perspectival. One is not simply conscious of an object, one is always conscious of an object in a particular way. One always has a certain perspective or point of view on the object; the object is always presented in a certain way or under a certain aspect for the subject. More specifically, however, we need to distinguish the intentional object in ‘the how of its determinations’ and in ‘the how of its givenness.’ To take a simple example, let us consider a perception of an iPod. I can see the iPod from one perspective or another, I never see it in its totality all at once. Furthermore, the iPod always appears to me in a certain illumination and with a certain background. Moreover, it also appears in a certain context with a determinate meaning. Depending on my previous experiences and current interests, I might see it as a present I got from a good friend, as an efficient
180
Phenomenology of Consciousness
way to storage my music collection, as a headache and source of irritation (because it is not working properly), etc. But apart from intending different properties of the object, apart from varying what the object I am intending is presented as, I might also vary the very form of presentation itself. Instead of perceiving the iPod, I can also imagine it, judge about it, remember it, etc. Husserl, for instance, typically distinguishes between signitive, imaginative (pictorial), and perceptual ways of intending an object or state of affairs: I can talk about a withering oak which I have never seen, but which I have heard is standing in the backyard, I can see a detailed drawing of the oak; or I can perceive the oak myself. Similarly, I can talk about how terrible it must be for homeless people to sleep on the streets, I can see a television program about it; or I can experience it myself. For Husserl these different ways of intending are not unrelated. On the contrary, the modes can be ranked according to their ability to give us the object as directly, originally, and optimally as possible. The object can be experienced more or less directly, that is, it can be more or less present. The lowest and most empty way in which the object can be intended is in the signitive act. These (linguistic) acts certainly have a reference, but apart from that, the object is not given in any fleshed out manner. The imaginative (pictorial) acts have a certain intuitive content, but like the signitive acts, they intend the object indirectly. Whereas signitive acts intend the object via a contingent representation (a linguistic sign), pictorial acts intend the object via a representation (picture) which bears a certain resemblance to the object as seen from a certain perspective. It is only the actual perception, however, which gives us the object directly. This is the only type of intention which presents us with the object itself in its bodily presence. Thus, on the Husserlian account, perception does not confront us with pictures or images of objects – except, of course, in so far as we are perceiving paintings or photographs – but with the objects themselves. When we say that something appears perceptually, this should consequently not be understood in the sense that the perceptually given is a picture or sign of something else. For Husserl, intentionality is not an ordinary relation to an extraordinary object, but a special kind
of relation to the intended object; a special ‘relation’ that can hold, even if the object does not exist; and that can persist even if the object ceases to exist. When it comes to intentions that are directed toward ‘unreal’ objects, they are in his view just as much characterized by their directedness as are ordinary perceptions. In contrast to normal perceptions, however, the referent – that which would satisfy or fulfill the intention – does not exist, neither intramentally, nor extramentally. In the case of a hallucination, the pink elephant exists neither inside nor outside of consciousness, but the hallucination is still about a pink elephant. In short, although one of the peculiar features of the mind is its ability to think about objects that do not exist, we should not accept the reality of nonexistent objects. The intentional object is not a special kind of object, but rather the answer to the question of what a certain intentional state is about. If the answer refers to some nonexisting object, the intentional object does not exist. If the answer refers to some existent thing then the intentional object is that real thing. So if I look at my fountain pen, then it is this real pen which is my intentional object, and not some mental picture, copy, or representation of the pen. Throughout this kind of analysis, phenomenology contends that consciousness is characterized by an intrinsic intentionality, and resists the attempt to provide a reductive account of intentionality, for example, by trying to explain it by appeal to nonintentional factors such as causality. But how exactly does intentionality work? How do we intend objects? This is where the notion of meaning becomes central. For the phenomenologists, intentionality is a question of meaning. We intend an object by meaning something about it. Does this emphasis on meaning commit the phenomenological account of intentionality to some form of internalism? In this context, internalism refers to the view that a subject’s beliefs and experiences are wholly constituted by what goes on inside the mind of that subject, so that matters in the subject’s natural and cultural environment have no bearing on their content. By contrast, externalism argues that mental states are externally individuated. What we think about, what we refer to, depends upon what actually exists in the (physical, social, and cultural) environment; what we experience depends upon factors that
Phenomenology of Consciousness
are external to the subject possessing the experience in question. There has been a widespread tendency to argue that whereas Husserl was a classical Cartesian internalist, existential phenomenologists like Heidegger and Merleau-Ponty favored a form of externalism since they were fully committed to the view that the mind is essentially determined by its intentional relationship to the world. Although it is quite true that later phenomenologists to a somewhat larger extent than Husserl emphasized the importance of practical and bodily forms of intentionality this standard interpretation nevertheless remains too simplistic. It ignores plenty of evidence suggesting that Husserlian phenomenology differs from traditional internalism. At the same time, however, it is by no means obvious that Heidegger or Merleau-Ponty can be classified as straightforward externalists. A frequent claim found in the phenomenological literature is that the relation between mind and world is an internal relation, a relation constitutive of its relata, and not an external one of causality. Mind and world are not distinct entities; rather they are bound constitutively together. Considering the way in which phenomenologists conceive of the mind–world relationship, considering how they would reject the dualism between a selfcontained mind and a mindless world, it is questionable whether it really makes much sense to classify their views as being committed to either internalism or externalism. Ultimately, we should appreciate that the phenomenological investigation of the condition of possibility for manifestation is more fundamental than any division between inner and outer – a division that the alternative between internalism and externalism remains bound to. Phenomenology might teach us that the forced choice between internalism and externalism is misguided, since there are other options available.
Embodiment The phenomenological investigation of the body is not the analysis of one object among others. That is, it is not as if phenomenology in its investigation of a number of different ontological
181
regions (the domain of logic, mathematical entities, utensils, works of art, etc.) also stumbles upon the body and then subjects it to a close scrutiny. On the contrary, the body is considered a constitutive or transcendental principle, precisely because it is involved in the very possibility of experience. It is deeply implicated in our relation to the world, in our relation to others, and in our self-relation. Phenomenologists object to the metaphysical division between res extensa and res cogitans. If one accepted such a division the only place for the body would seem to be on the side of the res extensa. But phenomenologists deny that the body is a mere object in the world. The body is not merely an object of experience that we see, touch, smell, etc. Rather the body is also a principle of experience, it is that which permits us to see, touch, and smell, etc. Obviously, the body can also explore itself. It can take itself (or the body of another) as its object of exploration. This is what typically happens in physiology or neurology, etc. But such an investigation of the body as an object is not exhaustive. The phenomenological emphasis on the body entails a rejection of Cartesian mind–body dualism. But this does not entail an endorsement of some kind of Cartesian materialism. It is not as if the phenomenological way to ‘overcome’ dualism is by retaining the distinction between mind and body, and then simply getting rid of the mind. Rather the notion of embodiment, the notion of an embodied mind or a minded body, is meant to replace the ordinary notions of mind and body, both of which are considered derivations and abstractions. Merleau-Ponty famously spoke of the ambiguous nature of the body, and argued that bodily existence is a third category beyond the merely physiological and the merely psychological. The first and most basic phenomenological distinction to be made is between the objective body and the lived body (Husserl’s distinction between Ko¨rper and Leib, respectively). This is, a phenomenological distinction rather than an ontological one. It is not meant to imply that each of us has two bodies: one objective and one lived. Rather it is meant to explicate two different ways that we can experience and understand the body. Whereas the former notion focuses on the body as seen from an observer’s point of view, where the observer may be a scientist, a physician, or even the embodied
182
Phenomenology of Consciousness
subject herself, the latter notion captures the body understood as an embodied first-person perspective. A description of the lived body is a description of the body from the phenomenological perspective. On the one hand, it is the way the body appears in immediate experience. On the other hand, it is more than that – it is the way the body structures our experience. The body is not a screen between me and the world; rather, it shapes our primary way of being-in-the-world. This is also why we cannot first explore the body by itself and then subsequently examine it in its relation to the world. On the contrary, the body is already in-the-world, and the world is given to us as bodily revealed. Indeed as Sartre points out, the body is operative in every perception and in every action. It constitutes our point of view and our point of departure. As perceivers and agents we are embedded and embodied agents. Already Husserl stressed that perception rather than being a mere passive intake of information, involves activity, more specifically, bodily movement. We see with mobile eyes set in a head that can turn and is attached to a body that can move from place to place; a stationary point of view is only the limiting case of a mobile point of view. What we see and hear and touch and taste and smell is shaped by what we do, and what we are capable of doing. In ordinary experience perception and movement are always united. I touch something by moving the arm. I see something by moving the head and eyes. That which is perceived is perceived as nearby or further way, as something that can be approached and explored. Perceptual intentionality presupposes a moving and therefore embodied subject. To understand perception is to understand the intentionality of our own body. When I perceive an object, say, a bookcase, the object is never given in its totality but always incompletely, in a certain restricted profile or adumbration. It is never the entire bookcase, including its front, backside, underside, and inside which is given intuitively, not even in the most perfect perception. Despite this, it seems right to say that the object of my perception is the bookcase, and not simply the perspectivally given surface of the front and side of the bookcase. We so to speak see more than we sense. How is this possible? Phenomenologists have suggested that the absent
profiles of the object are given to us as something that is perceptually accessible. Whereas the actual given front of the bookcase is correlated with a particular bodily position, the horizon of the cointended but momentarily absent profiles of the bookcase (its backside, bottom, etc.) is correlated with my sensorimotor capacities. The absent profiles are linked to an intentional if–then connection. If I move in this or that way, then this or that profile will become visually or tactually present. All perception and action involves a component of bodily self-experience. I am sitting in a restaurant. I wish to begin to eat, and so I need to pick up my fork. But how can I do that? In order to pick up the fork, I need to know its position in relation to myself. That is, my perception of the fork must contain some information about me, otherwise I would not be able to act on it. On the dinner table, the perceived fork is to the left of me, the perceived knife is to the right of me, and the perceived plate and wineglass in front of me. Every perspectival appearance implies that the embodied perceiver is herself the experiential zero point, the indexical ‘here’ in relation to which every appearing object is oriented. As an experiencing, embodied subject I am the point of reference in relation to which all of my perceptual objects are uniquely related. I am the center around which and in relation to which (egocentric) space unfolds itself, or as Merleau-Ponty would put it, when I perceive the world, the body is simultaneously revealed as the unperceived term in the center of the world toward which all objects turn their face. Whereas I can approach or move away from any object in the world, the body itself is always here as my very perspective on the world. That is, rather than being simply another perspectivally given object, the body itself is precisely that which allows me to perceive objects perspectivally. In a primary sense, I am not conscious of my body as an intentional object. I do not perceive it; I am it. The body tends to efface itself on its way to its intentional goal. We do not normally monitor our movements in an explicitly conscious manner. Fortunately, for had we been aware of our bodily movements in the same way in which we are attentively aware of objects, our body would have made so high demands on our consciousness that it would have interfered with our daily life. When
Phenomenology of Consciousness
I play tennis, my movements are not given as intentional objects. My limbs do not compete with the ball for my attention. Had that been the case, I would have been so inhibited that I would have been unable to play efficiently. However, this might change if something goes wrong. Imagine that you are playing tennis. Your attention is directed at the ball, which is heading toward you with high speed, as well as on the position of your opponent. Your body tightens in order to return the ball in a masterful smash, but suddenly you feel a sharp and intense pain in your arm. Your smashing opportunity is lost, and the pain is now demanding all your attention. It attracts your attention whether you want it to or not. There is nothing that reminds us of our embodiment (our vulnerability and mortality) as much as pain. Moreover, the painful body can occasionally be experienced as alien, as something that is beyond our control. As is frequently the case in life, it is the privation that teaches us to appreciate what we take for granted. Nothing in this conception of embodiment should lead us to conceive of the body as something static, as if it has a fixed set of skills and abilities. Not only can the body expand its sensorimotor repertoire by acquiring new skills and habits, it can even extend its capacities by incorporating artificial organs and parts of its environment. The classical example is the blind man’s cane. To put it differently and perhaps even more strikingly, the lived body extends beyond the limits of the biological body. It does not stop at the skin. To take embodiment seriously is to contest a Cartesian view of the mind in more than one way. Embodiment entails birth and death. To be born is not to be one’s own foundation, but to be situated in both nature and culture. It is to possess a physiology that one did not choose. It is to find oneself in a historical and sociological context that one did not establish. Ultimately, the issues of birth and death enlarge the scope of the investigation. They call attention to the role of historicity, generativity, and sexuality. Indeed, rather than being simply a biological given, embodiment is also a category of sociocultural analysis. To gain a more comprehensive understanding of the embodied mind, one cannot just focus narrowly on perception and action, one also has to consider sociality.
183
Intersubjectivity The phenomenological tradition contains rich but also quite diverse and even occasionally competing accounts of intersubjectivity. They all share, however, a rather critical appraisal of the argument from analogy. The classical argument runs as follows: In my own case, I can observe that I have experiences when my body is causally influenced, and that these experiences frequently bring about certain actions. In the case of others bodies, I can observe that they are influenced and act in similar manners, and I therefore infer by analogy that the behavior of foreign bodies is associated with experiences similar to those I have myself. Thus, the argument from analogy can be interpreted as an inference to best explanation. An inference bringing us from observed public behavior to a hidden mental cause. Although this inference does not provide me with indubitable knowledge about others and although it does not allow me to actually experience other minds, at least it gives me more reason to believe in their existence, than in denying it. As phenomenologists have pointed out, however, the argument presupposes that which it is meant to explain. If I am to see a similarity between, say, my laughing or screaming and the laughing or screaming of others, I need to understand their bodily gestures and behavior as expressive phenomena, as manifestations of joy or pain, and not simply as physical movements. If such an understanding is required for the argument of analogy to proceed, however, the argument presupposes what it is supposed to establish. In other words, we employ analogical lines of reasoning only when we are already convinced that we are observing minded creatures but are simply unsure precisely how we are to interpret the expressive phenomena in question. The argument also operates with some questionable assumptions. First, it assumes that my point of departure is my own consciousness. This is what is at first given to me in a quite direct and unmediated fashion, and it is this purely mental self-experience that is then taken to make possible the recognition of others. One is at home in oneself and one then has to project into the other, who one does not know, what one already finds in oneself.
184
Phenomenology of Consciousness
Second, the argument also assumes that we never have direct access to another person’s mind. We can never experience her thoughts or feelings. We can only infer that they must exist based on what we perceive, namely her bodily and behavioral appearances. Although both of these assumptions might seem perfectly obvious, many phenomenologists have rejected both. They have argued that we should not fail to acknowledge the embodied and embedded nature of self-experience and that we should not ignore what can be directly perceived about others. In other words, they have denied that our initial self-acquaintance is of a purely mental nature and that it takes place in isolation from others, and they have also denied that our basic acquaintance with others is inferential in nature. The phenomenological account differs from the accounts proposed by the dominant positions within the theory of mind debate, that is, the theory-theory of mind and the simulation theory of mind. Both of the latter positions deny that it is possible to experience other minded creatures. It is precisely because of the absence of an experiential access to others that we need to rely on and employ either theoretical inferences or internal simulations. Both accounts share the view that the minds of others are hidden, and they consider one of the main challenges facing a theory of social cognition to be the question of how and why we start ascribing such hidden mental entities or processes to certain publicly observable bodies. But it is no coincidence that we use psychological terms to describe behavior; indeed we would be hard pressed to describe it in terms of bare movements. Affective and emotional states are not simply qualities of subjective experience, rather they are given in expressive phenomena, that is, they are expressed in bodily gestures and actions, and they thereby become visible to others. There is, in short, something highly problematic about claiming that intersubjective understanding is a two-stage process of which the first stage is the perception of meaningless behavior and the second an intellectually based attribution of psychological meaning. In the majority of cases, it is quite difficult (and artificial) to divide a phenomenon neatly into a psychological aspect and a behavioral aspect – think merely of a groan of pain, a handshake, an embrace. In the face-to-face encounter we are confronted neither
with a mere body, nor with a hidden psyche, but with a unified whole. Scheler spoke of an ‘expressive unity.’ It is only subsequently, through a process of abstraction, that this unity can be divided and our interest then proceed ‘inward’ or ‘outward.’ Phenomenologists have tended to take an embodied perceptual approach to the questions of understanding others and the problem of intersubjectivity. We begin from the recognition that our perception of the other’s bodily presence is unlike our perception of physical things. The other is given in its bodily presence as a lived body, a body that is actively engaged in the world. As Sartre pointed out, it is a decisive mistake to think that my ordinary encounter with the body of another is an encounter with the kind of body described by physiology. The body of another is always given to me in a situation or meaningful context, which is codetermined by the action and expression of that very body. Some phenomenologists suggest that our understanding of others involves a distinctive mode of consciousness which they call empathy. Empathy is defined as a form of intentionality in which one is directed toward the other’s lived experiences. The phenomenological conception of empathy stands opposed to any theory that claims that our primary mode of understanding others is by perceiving their bodily behavior and then inferring or hypothesizing that their behavior is caused by inner experiences similar to those that apparently cause that kind of behavior in us. Rather, in empathy, we experience the other directly as a person, as an intentional being whose bodily gestures and actions are expressive of his or her experiences or states of mind. Most phenomenologists have argued that it makes no sense to speak of an other unless the other is in some way given and accessible. That I have an actual experience of the other, and do not have to do with a mere inference or imaginative simulation, does not imply, however, that I can experience the other in the same way as she herself does, nor that the other’s consciousness is accessible to me in the same way as my own is. The second-(and third-) person access to psychological states differs from the first-person access, but this difference is not an imperfection or a shortcoming. Rather, the difference is constitutional. It is what makes the experience in question, an experience of
Phenomenology of Consciousness
an other, rather than a self-experience. In short, we should not make the mistake of restricting and equating experiential access with first-person access. It is possible to experience minds in more than one way. When I experience the facial expressions or meaningful actions of others, I am having an experiential access to the life of the mind of others. To get a fuller picture of the phenomenological take on intersubjectivity, it must be emphasized, however, that many phenomenologists have denied that sociality is primarily a question of a thematic encounter between individuals, where one is trying to grasp the experiential states of the other. On the contrary, the very attempt thematically to grasp the experiences of others is the exception rather than the rule. Under normal circumstances, we understand each other well enough through our shared engagement in the common world. We encounter others in worldly situations, and our way of being together and understanding each other is codetermined in its meaning by the situation at hand. In fact, in contrast to mentalistic theory-of-mind approaches that define the problem of other mind as amounting to the question of how to gain access to the other person’s (hidden) mind, phenomenological approaches suggest that a more productive focus is on the other person’s world. In effect, to understand other persons I do not primarily have to get into their minds; rather I have to pay attention to the world that I already share with them. On the phenomenological account, self, world, and others belong together; they reciprocally illuminate one another, and can only be understood in their interconnection. See also: Concepts and Definitions of Consciousness; Intentionality and Consciousness; Mental Representation and Consciousness; Meta-Awareness; The MindBody Problem; Perception: The Binding Problem and
185
the Coherence of Perception; Philosophical Accounts of Self-Awareness and Introspection; Psychopathology and Consciousness; Self: Body Awareness and Self-Awareness; Self: Personal Identity; Self: The Unity of Self, Self-Consistency; Sensory and Immediate Memory; Subjectivity; First- and Third-Person Methodologies; Visual Experience and Immediate Memory.
Suggested Readings Drummond JJ (1990) Husserlian Intentionality and Non-Foundational Realism. Dordrecht: Kluwer Academic Publishers. Gallagher S (2005) How the Body Shapes the Mind. Oxford: Oxford University Press. Gallagher S and Zahavi D (2008) The Phenomenological Mind. London: Routledge. Gurwitsch G (1966) Studies in Phenomenology and Psychology. Evanston: Northwestern University Press. Heidegger M (1996) Being and Time. Stambaugh J (trans.). Albany: SUNY Press. Henry M (1975) Philosophy and Phenomenology of the Body. Etzkorn G (trans.). The Hague: Martinus Nijhoff. Husserl E (1982) Ideas Pertaining to a Pure Phenomenology and to a Phenomenological Philosophy. First Book. Kersten F (trans.). The Hague: Martinus Nijhoff. Leder D (1990) The Absent Body. Chicago: Chicago University Press. Merleau-Ponty M (1962) Phenomenology of Perception. Smith C (trans.). London: Routledge and Kegan Paul. Moran D (2000) Introduction to Phenomenology. London: Routledge. Sartre J-P (1956) Being and Nothingness. Barnes HE (trans.). New York: Philosophical Library. Scheler M (1954) The Nature of Sympathy. Heath P (trans.). London: Routledge and Kegan Paul. Sokolowski R (2000) Introduction to Phenomenology. Cambridge: Cambridge University Press. Thompson E (2007) Mind in Life: Biology, Phenomenology, and the Sciences of Mind. Cambridge, MA: Harvard University Press. Waldenfels B (2001) Das leibliche Selbst. Vorlesungen zur Pha¨nomenologie des Leibes. Frankfurt am Main: Suhrkamp. Zahavi D (2003) Husserl’s Phenomenology. Stanford: Stanford University Press. Zahavi D (2005) Subjectivity and Selfhood. Cambridge, MA: MIT Press.
186
Phenomenology of Consciousness
Biographical Sketch
Dan Zahavi (1967) is a professor of philosophy and the director of the Center for Subjectivity Research at the University of Copenhagen. His research areas include phenomenology, philosophy of mind, and cognitive science. He is a past president of the Nordic Society of Phenomenology and a current coeditor in chief of the journal Phenomenology and the Cognitive Sciences. He has written and edited more than 15 books including Self-Awareness and Alterity (1999), Exploring the Self (2000), Husserl und Transcendental Intersubjectivity (2001), One Hundred Years of Phenomenology (2002), Husserl’s Phenomenology (2003), The Structure and Development of Self-Consciousness (2004), Subjectivity and Selfhood (2005), and The Phenomenological Mind (with Shaun Gallagher) (2007).
Philosophical Accounts of Self-Awareness and Introspection W Seager, University of Toronto Scarborough, Toronto, ON, Canada ã 2009 Elsevier Inc. All rights reserved.
Introduction: Self-Awareness This article provides an overview of philosophical approaches to the topic of self-awareness and introspection. It does not attempt to survey the voluminous scientific work in this area (see ‘Suggested Readings’ for references). Some remarkable experimental work has been done, which by and large seems to support the hypothesis that we are quite poor at self-understanding. It is worth noting that the experimental work provides some support for what are called ‘evidence theories’ below which assert that self-knowledge involves self-interpretation (and hence is subject to a host of distortions and biases, often engendered by possession of background ‘theory’ of mind and self ). However, the scientific work generally investigates relatively sophisticated and high-level self-knowledge rather than the core of basic introspectible mental states about which our accuracy remains unchallenged. That is, while it may well be that our ability to introspectively know such things about ourselves as that we are not jealous of a colleague is highly suspect, or if what we are feeling is really love, this does not call into question our introspective knowledge that we are, say, now in pain or seem to be seeing an apple. We seek here to investigate the philosophical roots of self-awareness and introspection. It is important to distinguish the two ideas in the title of this article. Self-awareness is a broad concept that encompasses a much wider range of phenomena than does the concept of introspection. The remarks immediately following are intended primarily to highlight the domain and nature of introspection; for more on nonintrospective self-awareness see the entry for Self: Body Awareness and Consciousness. Some remarks here about the broader phenomenon are nonetheless in order. It is legitimate to include within self-awareness our ongoing awareness of our own bodies, especially proprioceptive monitoring of such things as
the positions of our limbs, muscle tone, joint extension, and so on. In the sudden absence of such information, our sense of ourselves and our place in the world would be radically altered. But ‘body image’ extends far beyond these basic physical parameters, to include culturally specified bodily norms, both of appearance and deportment. One might also include various forms of what might be called ‘situatedness’ within the purview of self-awareness. At least two forms are notable. One is physical or spatial situatedness – our sense of orientation in local space, which includes implicit action patterns, sensitive to the spatial distribution of objects with the potential either to interfere with or aid our actions (an example might be the way we smoothly avoid obstacles as we move around a room). This sort of awareness is seldom the focus of attention; indeed and famously, focusing attention on the smooth interaction we normally enjoy with the environment interferes with, rather than enhances it. Nonetheless, it is a foundational part of our awareness of ourselves in the world. It seems appropriate to include our sense of spatial situatedness within the shadowy (and still insufficiently explored) realm of what William James called the ‘fringe’ of consciousness. Spatial situatedness is just one of the many low-level or background features of consciousness that underlie our sense of ‘fitting into’ our environment. It is interesting and highly suggestive that our sense of, or at least certain aspects of, spatial situatedness can remain in the absence of other forms of conscious awareness of self and environment. An illustration of this is provided by certain forms of visual agnosia such as the well known case of DF, who lacks the ability to recognize objects (even at the level of a basic geometric form, such as the difference between a square and a triangle), but who nonetheless retains the ability to grasp objects appropriately and is able to walk through complex environments without
187
188
Philosophical Accounts of Self-Awareness and Introspection
difficulty. Although DF’s abilities are amazing in themselves, what is interesting from our perspective is the oft noted ‘confidence’ DF exhibits when interacting with objects via visual information. Thus it seems that her sense of (local) orientation in the world has been retained despite the complete absence of consciousness of the objects as such, around her. Another important sort of situatedness is ‘social situatedness.’ Social situatedness is our sense of orientation toward others as they are reacting to their own perception and understanding of us. This encompasses both sophisticated culturally mediated aspects of the social environment but also much more basic ongoing ‘assessments’ of the emotional reactions of others to one’s own behavior. There has recently been much debate about how we acquire knowledge of the mental states of others, especially focused on the issue of whether such knowledge is acquired via simulation – the empathic ability to get ourselves into a mental state that is hopefully similar to our target’s mental state – or via an internal theory of the mind. Probably both are necessary to generate our sense of ‘belonging’ in our social environment. This certainly involves an ongoing ‘comparison’ of our own mental states with those of the people around us, which gauges the appropriateness of these states in relation to each other. Social situatedness and spatial situatedness come together in our sense of ‘personal space,’ the jealously guarded, but culturally relative need to preserve a certain distance from others. As in the case of spatial situatedness, there can be breakdowns in social situatedness, many of which are very familiar (e.g., the paranoia exhibited in a number of syndromes). The very striking, rare, and bizarre Capgras syndrome illustrates the complexity and deep importance of social situatedness. Victims of this affliction form delusional beliefs that those closest to them (typically parents or spouses) have been replaced by ‘duplicates’ (twins or even – in modern times – robots). The delusion is of such irrepressible strength that tragic consequences sometimes result. The cause of Capgras syndrome is unknown, but the most favored hypothesis is that it results from a breakdown in the emotional response system, so that a kind of dissonance is created between the perceptual familiarity of someone and a lack of any appropriate
feelings toward that person. The dissonance is resolved by the ‘hypothesis’ that a loved one has been replaced by a stranger toward whom no emotional attachment exists. These forms of self-awareness are vitally important for normal functioning, and while they are in all probability a part of the necessary background enabling conditions for introspection, they do not themselves constitute introspection.
The Characteristics of Introspection Introspection is a special kind of knowledge we have of our own mental states. One way it is special is that it is does not appear to be mediated by inference from other knowledge we may possess. For example, it is conceivable that one might come to know that one was in a certain mental state by inferring this from someone else’s opinion (e.g., one’s psychoanalyst). This might be a legitimate form of self-knowledge, but it would not be introspective knowledge. Thus one important question about introspection is the ‘mechanical’ one of how it works to provide us with knowledge of our own mental states in this ‘special’ way. A naturally connected further question is whether the specialness of introspection means that introspective knowledge is somehow more secure or indubitable compared to other forms or sources of knowledge. However, before investigating these issues a more basic characteristic of introspection must be noted. Introspection produces propositional knowledge about our own mental states. Thus it is crucial to emphasize that since all such knowledge is conceptual in nature, introspective knowledge must involve the application of concepts – mental concepts – to ourselves. This is in contrast to the fundamentally nonconceptual aspects of the sorts of self-awareness discussed above and also in contrast to the ‘first-order’ awareness provided by consciousness itself. When we introspect we find any number of mental states of a quite staggering variety: sensory, bodily, cognitive, emotional, or more broadly affective, and so on. Introspection is a kind of knowledge, and in order to ‘know’ that we are in any of these mental states, we must possess the
Philosophical Accounts of Self-Awareness and Introspection
concepts necessary to categorize them correctly. Luckily, we are all in possession of a pretty full range of concepts well tailored for this task. The totality of them and their ‘dynamics’ are frequently referred to as folk psychology, and the invention of this system of concepts should perhaps be ranked as the greatest and most indispensable innovation humankind has achieved. The core idea of folk psychology is based on two foundational mental states: belief and desire. Belief represents the world as we take it be, desire represents the world as it should be (at least from our own point of view). Action results from the rational evaluation of how our beliefs provide a transformation, taking the world as it is to the world as we wish it to be. Thus a thirsty man heads toward rather than away from the river. Who knows how this scheme was first articulated, although obviously it long postdates the existence of creatures who ‘have’ beliefs and desires, but know nothing of them as such. More recently, folk psychology has been explicitly characterized as a kind of theory, of which the biological basis, development, and significance in both normal and pathological cases is the subject of ongoing current debate. And, of course, its complexity, measured both in the kinds of mental states recognized and their interactions, has grown immensely. Whether folk psychology ought to be considered a theory or not, it provides us with a hugely powerful way to understand, explain, and predict the thoughts and actions of our fellow human beings (not to mention helping us to successfully deal with animals). The subtlety and power of advanced folk psychology is most evident in literature. Here, drawn pretty much at random from Jane Austen’s Sense and Sensibility is an illustrative passage: . . .Marianne awoke the next morning to the same consciousness of misery in which she had closed her eyes. Elinor encouraged her as much as possible to talk of what she felt; and before breakfast was ready, they had gone through the subject again and again; with the same steady conviction and affectionate counsel on Elinor’s side, the same impetuous feelings and varying opinions on Marianne’s, as before. Sometimes she could believe Willoughby to be as unfortunate and as innocent as herself, and at others, lost every
189
consolation in the impossibility of acquitting him. At one moment she was absolutely indifferent to the observation of all the world, at another she would seclude herself from it for ever, and at a third could resist it with energy. In one thing, however, she was uniform, when it came to the point, in avoiding, where it was possible, the presence of Mrs. Jennings, and in a determined silence when obliged to endure it. Her heart was hardened against the belief of Mrs. Jennings’s entering into her sorrows with any compassion.
Every sentence of this passage deploys mentalistic notions, exhibiting a supreme knowledge on Austen’s part of how mental states interact with each other, to the end of giving us an excellent sense of the characters of the persons described. Of course, Austen could not have written this without a deep introspective awareness of her own mental states, no less than her evident acuity in understanding others. Introspection is then the application of these mentalistic concepts to ourselves, the very same concepts we use to describe the inner lives of others. While the range of mental ascription is the same for self-as well as other descriptions, introspection appears to possess some epistemic properties that decisively distinguish it from our knowledge of the mental states of others. We can mark out this distinctiveness in terms of a few (putative) core properties of introspective knowledge: privileged access, immediacy, infallibility, and transparency. Privileged access is the idea that introspective knowledge is in some way more secure and has a basis distinct from that of others’ knowledge of our mental states. While it is possible, and generally quite easy, for other people to know what we are thinking or feeling, our own access to these states is of a different nature. One of the most famous defenders of introspection as a special and epistemically privileged mode of mental access was Rene´ Descartes. Descartes’s dualism held that the mind was an independent substance, causally connected to the body by a divine act, which linked a particular mind with a particular body in a system which Descartes’s called the ‘substantial union.’ All conscious experience is a purely mental phenomenon, hence any knowledge of the body is at best indirect. Of course, this general thesis led Descartes to skeptical problems about the nature and existence of the
190
Philosophical Accounts of Self-Awareness and Introspection
external world. But skepticism does not extend into the realm of the mind. It is impossible to be wrong about one’s own existence because any mental act guarantees the existence of the thinker. But how do we know that we are conscious? It would seem to be impossible to be wrong about this, even if we grant the possibility of error about the content of one’s mental states. For try to imagine or conceive that you are not at this moment conscious. The act of conceiving this is by itself a conscious act. Evidently we have some kind of very special access to at least some of the features of consciousness. Introspective access also seems to be unmediated by any other knowledge. When, for example, you are suffering from the pain of toothache there does not seem to be anything ‘between’ the pain and your introspective awareness of the pain. It is important to emphasize that these nonetheless are entirely different states. The pain is a sensation with its own features: sensory qualities of feeling and location plus a distinctive sense of aversion (we know these are distinct because they can become dissociated in certain conditions, notably opiate analgesia). It does not involve the concept of pain, as evidenced by the fact that animals and very young children or infants can feel pain even though they seem to completely lack any mentalistic concepts. The introspective awareness of pain however requires both the existence of the conscious sensation of pain and also possession of the concepts needed to encode knowledge of it. This may suggest a worry about the thesis of immediacy: what guarantees the accuracy of the act of conceptualization? This worry is misplaced at this point and is really directed at the claim of infallibility. Introspective knowledge is infallible if it is impossible for the deliverance of introspection to be in error, or, more precisely, if it must be the case that any subject who believes on the basis of introspection that they are in mental state S is in S. An example where infallibility seems highly plausible is one’s introspective awareness of intense pain. Imagine your reaction to a dentist who replied to your anguished complaint of excruciating toothache that you were mistaken – that you were not in pain at all. The seeming inconceivability of the dentist being right about this illustrates the infallibility of introspection.
Contrast infallibility with transparency, which is the thesis that nothing in the mind is hidden from introspection, or, if a subject is in mental state S, then S can always become introspectively aware of S. A stronger form of transparency would replace ‘can become’ with ‘is.’ Once again, intense pain presents a plausible illustration of transparency. Imagine that you are experiencing an excruciating toothache. Now try to imagine that you are not or even cannot become introspectively aware of the pain. This scenario seems almost as absurd as the idea that one could be mistaken about being in intense pain. Despite the superficial plausibility of infallibility and transparency in these cases, and despite a historical tendency to embrace them (e.g., as in Descartes), upon inspection, neither infallibility or transparency seem likely to be true, or at least not true of all mental states in general. The idea that there could be unconscious mental states, a thesis anathema to Descartes, but embraced by the seventeenth century philosopher Leibniz and now utterly commonplace, puts pressure on the transparency thesis. However, given the weak formulation above, transparency fails only if there are mental states of which introspective access cannot be attained. Freudian-type examples thus seem to fail to undermine transparency since they are such as can – under the appropriate circumstances of psychoanalysis – become introspectively known. But a host of ‘subliminal’ perceptual states which have been shown to affect behavior in ways consistent with their mental status show quite clearly that transparency is not true in general. For example, in semantic priming it can be shown that the ‘meaning’ of a word which is presented visually for too short a time for subjects to be aware of it affects processing of consciously apprehended words. It also seems entirely possible that a conscious experience could be so overwhelming that their subjects would not be able to ‘step back’ to engage in introspection (perhaps the first time one attempts sky-diving would provide an example of such an experience). It is also possible that the character of certain experiences would be necessarily altered by any attempt to introspect them. It is also worth mentioning the pervasive phenomenon of self-deception, which strongly suggests that there must be vast areas of the mind hidden from
Philosophical Accounts of Self-Awareness and Introspection
introspection. At least, it is hard to see how the self-deceived could persist in their delusion if they were introspectively aware of the underlying cognitive mechanisms (e.g., denial, self-interest, etc.) Infallibility is not much more plausible than transparency when regarded as a general thesis. Humans are notoriously adept at self-deception and will with apparent sincerity avow mental states (beliefs, desires, etc.), which it is very hard to believe they actually possess. A typical example might be a man who steadfastly maintains that he believes his wife is faithful, but who persistently and carefully avoids all opportunities that might present evidence suggesting otherwise. The best account here might well be that he actually believes his wife is unfaithful but ‘cannot admit it to himself ’ as we say. Closely akin to – perhaps a form of – self-deception is the phenomenon of cognitive dissonance in which belief and desire formation is distorted by various – usually emotionally charged – factors (the fable of the fox and the grapes provides a homely illustration: the fox realigns his beliefs about the grapes on the basis of their unavailability). It appears clear in all such cases that there is a large gap between how we think our minds are working and their actual operation, which is at odds with infallibility. Despite the difficulties with these putative features of introspection, it remains clear that introspective knowledge of our own mental states is a special form of knowledge, possessing a clear form of privileged access, and at least a high degree of transparency and infallibility.
Theories of Introspection Modern views of introspection broadly divide into four quite distinct forms, which I will label here ‘perception’ theories, ‘evidence’ theories, ‘transcendental’ theories, and ‘displaced perception’ theories. All of them, and certainly the first two and the fourth approaches, take one of their primary tasks to be the integration of introspection into a naturalistic view of the world. The main force of the term ‘natural’ is ‘anti-Cartesian,’ specifically anti-Cartesian dualism. These theories are also opposed to Descartes’s endorsement of infallibility and transparency and aim to provide
191
a theory of introspection that will smoothly integrate with the scientific world view and current psychological evidence. Perceptual Theories The most obvious approach to introspection is simply to equate it with perception. According to naturalism, mental states are states of the brain (more or less) and so introspection is perception of the brain. Of course, if a surgeon holds a mirror up so that I can see my brain as she operates on it or if I am watching the functional magnetic resonance imaging (fMRI) readout of my own brain operations, I am not, in virtue of this unusual perception of my own brain, also introspecting. Just as in the case of perception, introspection requires its own sense organ, which must itself be a part of the brain with its own specialized ‘inner receptors.’ Thus is born the inner-scanner theory of introspection, which postulates a functional brain system – let’s call it the ‘I-scanner’ – designed, by evolution presumably, to actively seek out information about those brain systems which realize mental states. David Armstrong explains the action of the I-scanner thus: In perception the brain scans the environment. In awareness of the perception another process in the brain scans that scanning. . . .
The I-scanner is by its nature rather more intimately connected to the objects of its perceptual capabilities than are our various sensory ‘scanners,’ and this is supposed to explain both the accuracy of introspection and the long-standing philosophical illusion of perfect introspective infallibility and transparency. According to I-scanner theorists, introspection is, as a matter of fact, ‘practically’ infallible, but no more so, and only in the way that perception of nearby objects directly before us in good light, and so on, is characteristically infallible. However, following the analogy of sense perception one can envisage at least the philosophical possibility of radical errors in introspection. For example, an adept science fictional brain surgeon could, in principle, activate a variety of I-scanner states that would yield introspective belief in all sorts of mental states, none of which
192
Philosophical Accounts of Self-Awareness and Introspection
were actually present. This is the natural result of downgrading the privileged access supposedly enjoyed by introspection to a mere generally reliable causal link. One could raise several objections against the I-scanner theory, but the most salient difficulty stems from a curiosity of the phenomenology of introspection. If introspection were properly understood as a kind of perception, one would expect there to be a distinctive phenomenology of introspection; mental states ought to have ‘introspectible qualities’ just as the objects of perception have perceptible qualities such as color and form. Unfortunately, it is a plain fact that there is no distinctive introspective phenomenology whatsoever. That you have current introspective access to your perceptual states of consciousness no one will deny, but all the phenomenology you can find is exhausted by the perceptible qualities of the objects of your current perception. Attending to your perceptions may well heighten or otherwise alter your perceptual state, but it does not introduce a new realm of introspectible qualities. It would be desperate folly to claim that the introspectible qualities of mental states are an exact ‘copy’ of the perceptible qualities of objects, and even worse to claim that we are never aware of anything but introspectible qualities of the mental. Lacking the nerve to make this latter assertion, we might ask the I-scanner theorist why, when we introspect our perceptual states of consciousness, the evident ‘outerness’ of perception is not replaced by an ‘innerness’ appropriate for introspective awareness of the brain (or, even, the mind)? The introspection of intentional mental states is generally of greater significance to us than the introspection of perceptual states. The I-scanner theory seems to fare even less well in this domain. It is very difficult to convince oneself that beliefs and desires are known by a process analogous to perception. Intentional states do not have a set of quasi-perceptual qualities by which they are internally recognized. This is evident in common wisdom. If Mary wonders whether she really loves Tom, she ‘thinks’ about Tom and their relationship. The advice to forget about thinking and just look inside oneself gets one nowhere. Of course, there are various feelings Mary might have when she thinks about Tom and she can be conscious of
these as feelings, but the question is, are they ‘signs’ of love. And there, the problem is that there just does not seem to be anything more to ‘look at’; our intentional mental states do not form an inner world of objects with their various introspectible properties. A defender of a perceptual account of introspection might try the following reply. Couldn’t Mary imagine, for example, what it would be like to live with Tom? She might imagine various domestic scenes or possible situations and ‘play out’ how Tom might act in these situations. Of course she could and this is no doubt part of what thinking about our own intentional states involves, but the central point is that this is not introspection. Consciously imagining something is ‘not’ in and of itself introspection. I can vividly imagine the Canadian flag, with its bright red stripes and maple leaf, but I am not necessarily introspecting when I do this. Of course, it is not a great leap from imagining to introspection. Knowing that and what I am imagining, as well as how I am imagining it, is introspective knowledge, but I do not get this knowledge by somehow perceiving my imagining of the flag. My imagining does not look like anything, though my image of the flag does (at least there is a distinctive and obviously perceptual phenomenology of imagining). But ‘looking’ at my image is just the imagining itself, not an act of introspection, although of course I can, even in imagination, attend to particular aspects of the imagined sensory qualities. Evidence Theories of Introspection What I shall call ‘evidence theories’ provide an entirely different approach to introspection. The basic claim of evidence theories is that our knowledge of our own mental states is formed in precisely the same way as our knowledge of others’ mental states; there is no asymmetry between the ‘first-person’ and ‘third-person’ knowledge of the mental. Gilbert Ryle was the most forceful exponent of such theories. Ryle put forth his view succinctly as follows: ‘‘Our knowledge of other people and ourselves depends upon our noticing how they and we behave.’’ There are some advantages to the evidence theory. It dissolves the surprisingly vexing problem of
Philosophical Accounts of Self-Awareness and Introspection
other minds, or at least transforms it into no less a problem about other minds than about our own. However, this advantage depends upon the assumption that the problem of other minds does not stem from a genuine and fundamental asymmetry in the relation we bear to our own mind and the minds of others. Thus, an obvious objection to the evidence theories turns the advantage on its head. A direct refutation of the evidence theory follows from its denial that we stand in relation to our own states of mind, quite distinct from the relation we bear to the mental states of others. The evidence theory also exhibits a nice theoretical economy. Since it is given that we do have methods for attributing mental states to others it is rather elegant to enlist these methods for self-attribution. This eliminates any theoretical need for a special and, at least in its traditional guise, epistemically suspicious introspective faculty or sense, or even any special mode of selfinterpretation. It also retains an acceptable, albeit rather weak, form of ‘privileged’ access to our own mental states in its allowance that we all generally have access to more, if not a different kind of, evidence about ourselves than others, while possessing abundant means to account for the evident failures of self-knowledge. Now, it is hardly surprising that in some cases people will attribute mental states to themselves on the same basis that they attribute them to others, for people are capable and fond of thinking about themselves and their motivations. As noted above, the complex system of concepts of mental states and their dynamic interconnection of folk psychology provides a very rich and hugely successful system for understanding ourselves and others. We quite naturally apply this knowledge to ourselves no less than to others. But this is a frail basis upon which to build an evidence theory of introspection for it simply ignores the vast range of self-knowledge to which the model applies very badly if at all. It is, as philosophers have long remarked, extremely easy to gain introspective knowledge of one’s own perceptual states. In fact, this ease of access provides the core intuitive support for the special features of introspective knowledge such as infallibility and transparency. But it is evident that this access does not require one to observe,
193
imagine, or remember one’s own behavior or utterances. I don’t need to say to myself ‘‘I am seeing red’’ or hear myself saying this to someone else, to know that I am in a state of ‘seeing red,’ nor do I need to imagine red (in fact, when one is actually seeing red, it is next to impossible to imagine red at the same time, and obviously such imaginings would gain nothing over the perception). Of course, ‘seeing red’ is not by itself an introspective act, so perception is not itself an objection to the evidence theories, but introspective knowledge of perceptual states is freely available simply by virtue of the fact that these are conscious states. We simply do not have to watch what we do or say to know that or what we are seeing or hearing, whereas the only way to know what perceptual experiences others are having is by observing this sort of behavior. The evidence theory no more plausibly explains introspective knowledge of the vast majority of intentional states such as beliefs and desires than it does our knowledge of perceptual states. Knowing that I believe that, for example, 2 + 2 = 4 is (a rather trivial) case of introspective knowledge. It is ludicrous to suppose that my knowledge of this belief depends upon or stems from an observation of my own behavior or from some act of imagination (what would it be: imagining myself writing down ‘2 + 2 = ’ and then noting that my imaginary self completes the equation with ‘4’?). It is no different in the case of desire. I know that I like strawberry pie, but not because I find myself ordering it in a restaurant, or because I can remember that I have ordered it often in the past, or because it is easy to imagine myself ordering it in the future, or because I remember hearing myself say frequently enough ‘‘I like strawberry pie,’’ or because I can easily imagine hearing myself say this, or yet because I frequently say to myself ‘‘I like strawberry pie’’ (which in fact I do not). Nonetheless, the evidence theory is not entirely hopeless. Although it cannot be regarded as a serious candidate for a theory of introspection per se, not all my self-knowledge is introspective. The evidence theory is plausible only for cases of sophisticated or ‘difficult’ attribution of complex intentional states and only a tiny fraction of my self-knowledge, albeit a highly significant fraction, involves these. This zone of accuracy for the
194
Philosophical Accounts of Self-Awareness and Introspection
evidence theory is important. It points to the very real intellectual and practical problems we face in knowing ourselves as complicated intentional beings within the complex social environment we have constructed for ourselves. A good part of our self-knowledge stems from self-interpretation as the evidence theory claims. In fact, it is arguable that the most important elements of our selfknowledge are based on self-interpretation of the kind the evidence theory appeals to – introspective knowledge is generally more easily obtained and the majority of it is relatively trivial. But despite the insights the evidence theory embodies, it seems to be entirely inadequate as a general account of introspection. If we insist on regarding it in this light the evidence theory is seriously misleading about the fundamental source and nature of introspective knowledge. Transcendental Theories of Introspection In philosophy, it is always at least thought to be possible that the appearance of a problem is mere appearance, which can be resolved by a more careful analysis of the concepts involved. And so it is with the problem of introspection. The ‘transcendental’ theories of introspection are really more of a denial that there is a problem of introspective knowledge than a substantive theory of introspection. I adopt the term ‘transcendental’ from Kant’s usage in which a transcendental ‘proof ’ of some subject matter depends on laying out the conditions of its possibility. For example, Kant gives a transcendental proof that our experience conforms to the dictates of causality (or reveals a world to which the category of causation applies) by trying to show that one of the conditions for the possibility of experience is that it conforms to causality. Thus, given that we do have experience, we can infer that it meets the condition of causality. How is this supposed to work as applied to the case of introspection? Of course, any transcendental account of introspection will admit the existence of extensive self-knowledge and will go so far as to allow that this sort of knowledge has a very special nature and perhaps has a more secure status than most other kinds of knowledge. But the story of introspection as told by transcendentalists
is quite different from those we have heard so far. The basic idea is that a subject S having accurate self-knowledge is a condition for the possibility of all attributions of mental states to S. Philosophers, notably Donald Davidson, have argued that a general failure of self-knowledge would render the interpretation of a subject as believing and desiring incoherent. Roughly speaking the argument goes like this. Suppose that in general some person acted as if they believed, say, that elephants have tusks, even going so far as affirming that elephants have tusks (i.e., uttering ‘elephants have tusks’), but denied that they believed this or even claimed that they could not tell whether or not they believed that elephants have tusks. Such ‘metabehavior’ would render the initial interpretation of the subject’s belief unsustainable and if such paradoxes were the norm rather than very rare exceptions, it would throw into doubt the idea that the person had beliefs and desires at all. Note that this sort of a priori defense of self-knowledge leaves room for errors of introspection, so long as these errors remain relatively infrequent and do not become, so to speak, flagrant. A very elegant and compressed argument for a transcendental theory of introspective knowledge has been provided by Akeel Bilgrami who ingeniously links self-knowledge with the possibility of interpreting people as moral agents. Roughly speaking, the proposal is that only those who are aware of, or know, what they are doing can be held morally responsible for the consequences of their actions. Then, Bilgrami argues, since it is a condition for the possibility of taking others to be persons that they be thought of as morally responsible for their actions (at least most of the time), it is a condition for the possibility of taking others to be persons that they (generally) know what they are doing, and one cannot know what one is doing without some insight into the intentional states, which actually explain one’s actions. Selfknowledge of one’s own intentional states thus emerges as a condition of personhood. An immediate problem with such accounts is that they threaten to implausibly reduce the kinds of creatures which can possess minds. Intuitively, many creatures enjoy consciousness and a fairly rich mental life, while lacking the conceptual equipment necessary for introspective knowledge.
Philosophical Accounts of Self-Awareness and Introspection
I think a very large number of animals fall into this category (and it may well be that no animals save for human beings are capable of introspection). Probably, human infants and even young children perhaps up to age three or so also have not yet grasped the mentalistic concepts needed for introspection (there is a large literature on the way that children acquire mental concepts and in general it seems to support the idea that full acquaintance with folk psychology arrives rather late and may have a genetic component). Thus there is a serious problem lurking here with respect to the minds of animals and young children. If having a mind itself somehow entails that the subject possesses generally accurate self-knowledge then they will end up mindless. I take it that this is not a welcome result and thus requires some kind of hierarchy of minds, only some of which will be capable of introspection. It is not altogether clear that those of a transcendentalist bent can accommodate a scale of more or less primitive animal minds. Leaving that worry aside, there is no doubt that there is something correct in transcendental analyses like these, but it is very far from clear that they either solve or dissolve the problem of introspective knowledge. This is because of two related failings that reveal the inability of the transcendental theories to evade the need to provide a positive account of introspection. First, although the transcendentalist may well be right that generally accurate self-knowledge is a condition of the possibility of ascribing mental states to others and also ourselves, one must nonetheless ask how creatures for which this possibility arises are structured so as to realize the possibility. The original transcendentalist, Kant, consigned this structure to an unknowable noumenal realm into which human thought dare not venture on pain of unintelligibility. This is hardly a strategy modern, naturalistically inclined transcendentalists about self-knowledge should wish to emulate. Nor is it defensible to refuse the request for an explanation of how self-knowledge is attained simply by appealing to the transcendental proof that, after all, it ‘must’ be attained. Here, the transcendental methodology is closely akin to the use of the anthropic principle in cosmology, which well illustrates the general limitations of a transcendental or anthropic account of some
195
fact. For example, we observe that the Earth is neither too near nor too far from the Sun to support life. This is no surprise, for the Earth being at such a distance is a condition of the possibility of observers. The anthropic, ‘transcendental’ account of the Earth–Sun distance is, however, no substitute for an explanatory tale of the genesis of the solar system in which each planet attains its appointed place via natural forces, with no help from the fact that the Earth would someday spawn astronomers who can measure the Earth–Sun distance. Every realized possibility must be realized via some mechanism; so too our introspective capacities must have some actual source which, when articulated, will amount to a positive theory of introspection. Second, it is evident that self-knowledge is a cognitive achievement that takes some effort. Selfknowledge, like any other knowledge, does require evidential warrant. When we have such knowledge, we generally also know why we know what we know about ourselves. This point is intended to be uncontroversial and not metaphysically or epistemologically deep. Suppose I report that I am in pain and someone asks me how I know this. The obvious, and correct, answer is ‘‘I feel it’’ (given that, of course, I know what pain is). Self-knowledge cannot be freely generated simply by asserting something about oneself. Self-knowledge must answer to the same epistemic canons to which any potential domain of knowledge must answer, such as those of evidence, warrant, and the possibility of error. Further, my self-knowledge must be integrated with my knowledge of others’ mental states and my knowledge of the world. But then the range and type of evidence required for selfknowledge is problematic and demands an account. Displaced Perception Theories of Introspection Recently, a novel theory of introspection has been advanced as an adjunct to an account of consciousness that identifies consciousness with certain forms of mental representation (the theory is usually known as the representational or intentional theory of consciousness and has been developed in most detail by Fred Dretske and Michael Tye).
196
Philosophical Accounts of Self-Awareness and Introspection
It denies that there are peculiarly mental features of all experience (the so-called qualia), and instead asserts that consciousness is essentially a matter of presentation to the mind of the content of mental representations. For example, if one imagines the Canadian flag the experience is of a vivid red maple leaf shape on a white background flanked by two similarly bright red bars. The mechanism of consciousness is that the content of a representation of the flag is the content of the conscious experience. This has certain obvious advantages. It avoids postulating bizarre entities with little prospect for assimilation in the scientific picture of the world (e.g., the imaginary Canadian flag which cannot be found anywhere in the physical universe). The prospect of integration with a naturalistic outlook is also present, or at least reduces to the problem of naturalizing mental representation. It is worth mentioning here that the main rival to this account of consciousness is the socalled higher order thought (HOT) theory (see Seager article on consciousness). The mechanisms of introspection of the two theories are quite similar insofar as both require conceptualization of the mental state, as mental as a condition for conscious introspection. (For HOT theories, unconscious introspection exists and in fact is what accounts for the target of the introspective state being a conscious state.) The theory of introspection that naturally accords with this view of consciousness asserts that introspection depends on an interplay of consciousness and concepts via a mechanism called displaced perception (which extends beyond the realm of introspection). An example of Dretske’s is, learning that the postman has arrived by perception of the dog’s barking. To get such knowledge one must hear the dog (consciousness) and one must also know what the dog’s barking signifies (concepts). Introspective knowledge of our own perceptual states similarly requires that we consciously perceive, but also that we know what perceiving is. As we have noted, knowledge is conceptual and so requires an appropriate field of concepts for its formulation. Introspective knowledge requires the field of concepts that together form our notion of the mind, or folk psychology, a set of concepts ready to hand for all reasonably mature human beings.
This theory does not require that there be any perception of our mental states as such, but instead simply requires that we possess certain mentalistic concepts that we can successfully apply to ourselves as we perceive the world (including our own bodies). Thus, when I am consciously perceiving red, I have introspective knowledge of my experience because I can and do apply the concept of perceiving red to this instance of my conscious perceptual experience. I don’t need to perceive my perceiving (as the I-scanner theory asserts at the bottom) to make this application any more than I need to perceive my perceiving of a barking dog to apply the concept of ‘barking dog’ to that object. Of course, I do need to be perceiving red to make the introspective application of the concept ‘perceiving red,’ but that is simply a matter of being perceptually conscious. The displaced perception model of introspection shares with the evidence theories the idea that introspection requires an input of information from which mentalistic conceptualization will follow, but it denies that introspection depends on exactly the same kind of evidence we use in attributing mental states to others. Instead, it exploits the fact that we are conscious beings and we can develop or acquire the ability to describe our consciousness at, so to speak, one step removed, as a mental feature of ourselves. The ability to comprehend the epistemic distance between the world and the experience of the world is not some kind of benighted proto-Cartesianism; it is a vital step toward self-consciousness and an awareness of one’s own identity. One difficulty with the displaced perception model is that it appears to transform introspective knowledge into inferential knowledge, thus disparaging even the acceptable level of infallibility and transparency which introspection evidently possesses. That is, it does not account for what is truly special about introspective knowledge. It does indeed seem odd to say that we infer that we are experiencing, say, the taste of an apple based on our consciousness of the apple’s taste. Although many proponents of the displaced perception model of introspection liken introspective knowledge to inferential knowledge, the idea can be resisted. Instead, it may be possible to liken introspective knowledge to nothing more than
Philosophical Accounts of Self-Awareness and Introspection
conceptual categorization. It does not seem quite right to say that one ‘infers’ that a dog is in the yard (a dog which is perfectly visible and in fact is seen) from some underlying purely sensory material, even if such material plays a role in conceptualization. Similarly, we don’t need to infer that we are ‘seeing’ a dog from our visual experience. We need only be in possession of the concept of seeing and have the ability to apply that concept in the appropriate circumstances. An interesting question about this model of introspection is how it is to be extended beyond the realm of introspective knowledge of conscious perceptual states, to include intentional and emotional mental states. For example, consider the problem of how we attain introspective knowledge of our own beliefs or desires. In all probability you, the reader, believe that there are no giraffes on Mars and you have, at least you have now, introspective knowledge that you possess this belief. But it does not seem that one can discover any special internal mental feature by which one recognizes, first, which belief is in question (e.g., giraffes live on Mars vs. grass is green) and, second, whether that content is actually one which is believed (as opposed, say, to being entertained, wondered about, doubted, etc.). Instead, it seems that one knows that one believes that no giraffes live on Mars simply in virtue of figuring out (or recalling, if for some reason you had worked this out in the past) that no giraffes live on Mars. It is as if one says to oneself: ‘‘giraffes on Mars? No way (what would they breathe, how did they get there, etc.),’’ and from this, one both verifies that giraffes do not live on Mars and one believes that giraffes do not live on Mars. The most straightforward way to assimilate this insight about believing is to posit some feature of conscious experience, which we could label ‘taking for true.’ There is, on this view, a basic experiential difference, which differentiates the thought ‘‘I am six feet tall’’ when thought by me and when thought by someone who knows their height to be less than six feet. I take this for true, it seems correct to me, I happily absorb or maintain it within my cognitive economy; the other person rejects it, finds it – in the literal sense of the term – incredible, and will not take it up into his belief system.
197
If the postulate that there is an act of the mind of ‘taking for true,’ then the displaced perception model of introspection can be easily extended to the case of belief. One knows that one believes that giraffes do not live on Mars because one ‘sees’ the truth of this proposition and knows that this licenses application of the concept of belief to it. The model can also be extended to other intentional states. Besides belief, the second and equally fundamental intentional state is desire. To accommodate desire one must posit something like an experiential feature of ‘taking to be good.’ Perhaps this is intuitively even more acceptable than is ‘taking for true.’ We have many expressions that seem almost directly to refer to such a feature of experience, for example, when we say of something that it ‘looks good to eat.’ With an account of introspective knowledge of belief and desire in place it is not difficult to further extend the theory to other intentional states and the highly significant subclass of states – involving both cognitive and sensory elements – that comprise emotional consciousness. Apart from the worry that the displaced perception account makes introspection implausibly similar to inferential knowledge, another serious worry is that it mistakenly equates introspection with a kind of self-generation of our mental states rather than a kind of access to preexisting mental states. Roughly, but hopefully vividly, the problem is that this account cannot distinguish between introspective knowledge of a belief and the (presumably also introspective) knowledge that one has just acquired a belief. In fact, there is a danger that displaced perception could yield a false introspective claim if it is taken to generate knowledge about what one’s beliefs were, prior to setting the displaced perception mechanism into action. It is not entirely clear that this is a bug rather than a feature. After all, if a child comes to disbelieve in Santa Claus simply by reflecting on the implausibilities inherent in the usual stories, then it seems quite correct for introspection to generate the claim that the child does not believe in Santa Claus. But there is an interesting issue here about whether introspection should be regarded more as a kind of ‘measuring instrument’ revealing the current state of one’s mind or more as an engaged
198
Philosophical Accounts of Self-Awareness and Introspection
faculty, which is at least in part actively forming the mind according to principles of rationality as well as, perhaps, other normative principles. The notion that introspection plays a role in selfconstruction is attractive (and has obvious affinities with the transcendental approach), but does not seem to be at odds with the positive accounts of introspective mechanisms discussed above.
Conclusion The topic of introspection, self-awareness, and self-knowledge is complex and far from settled. The nature and status of introspective knowledge remains hotly debated and the issues raised in discussions of introspection are closely linked to the wider philosophical topics of mind, naturalization, and consciousness. In addition, the recent and accelerating growth in our understanding of the neural processes underlying introspection is likely to have a huge impact on the philosophical accounts and general attitudes toward our knowledge of ourselves. See also: Phenomenology of Consciousness; Self: Body Awareness and Self-Awareness; Self: Personal Identity; Self: The Unity of Self, Self-Consistency.
Suggested Readings Abilgrami A (1995) Belief and Meaning. Oxford: Blackwell. Aydede M (2003) Is introspection inferential? In: Gertler B (ed.) Privileged Access: Philosophical Accounts of SelfKnowledge. Burlington: Ashgate. Byrne A (forthcoming) Introspection. Philosophical Topics. Carruthers P (1996) Simulation and self-knowledge: A defense of theory. In: Carruthers P and Smith P (eds.) Theories of Theories of Mind. Cambridge: Cambridge University Press.
Davidson D (1987) Knowing one’s own mind. Proceedings and Addresses of the American Philosophical Association 60: 441–458. Dretske F (1995) Naturalizing the Mind. Cambridge, MA: MIT Press. Dunning D (2005) Self-Insight: Roadblocks and Detours on the Path to Knowing Thyself. New York: Psychology Press. Gertler B (ed.) Privileged Access: Philosophical Accounts of Self-Knowledge. Burlington: Ashgate. Gertler B (2003) Self-knowledge. In: Zalta EN (ed.) The Stanford Encyclopedia of Philosophy. Stanford, CA: Metaphysics Research Lab. http://plato.stanford.edu/ archives/spr2003/entries/self-knowledge/. Gertler B (2008) Do we look outward to determine what we believe? In: Hatzimoysis A (ed.) Self-Knowledge. Oxford: Oxford University Press. Goldman A (2006) Simulating Minds: The Philosophy, Psychology, and Neuroscience of Mindreading. Oxford: Oxford University Press. Gopnik A (2001) Theory of mind. In: Wilson R and Keil F (eds.) The MIT Encyclopedia of the Cognitive Sciences. Cambridge, MA: MIT Press. Kind A (2006) Introspection. The Internet Encyclopedia of Philosophy. http://www.iep.utm.edu/i/introspe.htm. Lycan W (1997) Consciousness as internal monitoring. In: Block N, Flanagan O, and Gu¨zeldere G (eds.) The Nature of Consciousness. Philosophical Debates, pp. 755–771. Cambridge, MA: MIT Press. Lyons W (1986) The Disappearance of Introspection. Cambridge, MA: MIT Press. McLaughlin B and Rorty A (1988) Perspectives on Self-Deception. Berkeley: University of California Press. Moran R (2001) Authority and Estrangement: An Essay on Self-Knowledge. Princeton, NJ: Princeton University Press. Nichols S and Stich S (2003) Mindreading: An Integrated Account of Pretence, Self-Awareness, and Understanding of Other Minds. Oxford: Oxford University Press. Ryle G (1949) The Concept of Mind. London: Hutchinson & Co. Schwitzgebel E (2008) The unreliability of naı¨ve introspection. Philosophical Review 117: 245–273. Seager W (2002) Emotional introspection. Consciousness and Cognition 11(4): 666–687. Shoemaker S (1996) The First-Person Perspective and Other Essays. Cambridge: Cambridge University Press. Tye M (1995) Ten Problems of Consciousness. Cambridge, MA: MIT Press. Wegner D (2002) The Illusion of Conscious Will. Cambridge, MA: MIT Press. Wilson T (2004) Strangers to Ourselves: Discovering the Adaptive Unconscious. Cambridge, MA: Belknap.
Philosophical Accounts of Self-Awareness and Introspection
199
Biographical Sketch
William Seager is a professor of philosophy at the University of Toronto. He was born and raised in Edmonton, Alberta. He obtained his BA and MA from the University of Alberta and then moved to Toronto for his PhD work, where he has remained ever since. His main research interests are in the philosophy of mind, especially the problem of consciousness. He has published Theories of Consciousness (Routledge, 1999) and many articles on mind and consciousness, the most recent being ‘The intrinsic nature argument for panpsychism’ (in A. Freeman (ed.) Consciousness and Its Place in Nature, Imprint 2007). William Seager would especially like to thank Matt Habermehl and Adrienne Prettyman, as well as a number of assiduous anonymous referees, for very helpful comments on this article.
The Placebo Response L Y Atlas and T D Wager, Columbia University, New York, NY, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Analgesia – Significant lessening or complete termination of pain sensation. Nociception – Pain sensation. Placebo – An administered treatment that on its own has no known beneficial effect on a given condition. Placebo effects – Observed improvements in outcome measures attributed to the influence of the placebo treatment. Placebo response – The set of endogenous processes whereby conscious expectancies and conditioning recruit brain and body mechanisms to elicit beneficial effects on symptomatology and processing.
Introduction The history of placebo effects over the last century has been a turbulent one. They have been variously dismissed as artifacts and embraced as compelling evidence of the healing power of the mind. In either case, the rigorous scientific study of placebo provides a unique window into the relationship between the mind, brain, and body. Such research is beginning to reveal the potential power of, and the limits on, how the brain affects the physiological and psychological manifestations of health and disease. In some cases, positive outcomes have been demonstrated to result from conscious expectations of therapeutic benefit. But some of the most potent placebo responses have been elicited by learning that may be outside of conscious awareness and at least partially impenetrable to volition and conscious thought. These studies, and more broadly the kinds of evidence that have and have not demonstrated the power of placebos, are the subject of this article.
A placebo treatment is one that is expected to have no direct physical or pharmacological benefit – for example, a starch capsule given for anxiety or pain, or a surgery where the critical surgical procedure is not performed. For this reason, placebos are routinely used as comparison conditions in clinical studies, against which to evaluate the effects of investigational treatments. However, placebo treatments have also frequently been used to actually treat a variety of ailments; they have had a place in the healer’s repertoire for thousands of years, and are used as clinical treatments by physicians in industrialized countries today with surprising frequency. To the degree that placebo treatments are healing agents, their power lies in the psychobiological context surrounding treatment, resulting in an active response in the brain and body of the patient. This endogenous response is referred to as a placebo response, or meaning response. In many cases, the psychological meaning of the treatment induces shifts in cognition, emotion, and corresponding brain and nervous system activity that produce a palliative effect. In other cases, placebo responses result from associations learned in the brain through a process known as conditioning. In this process, associations between the elements of the treatment context (the shape of a hypodermic needle, or the color of a pill) and helpful neurobiological responses are formed in the brain. These associations may be learned and activated outside of the patient’s consciousness. Thus, although a placebo treatment is itself inert, the placebo response on the part of a patient may have real healing benefits. What is typically measured in a study, however, are placebo effects – observed improvements in signs or symptoms attributed to the influence of the placebo treatment. Although the two terms are often used interchangeably, for the purpose of this article we distinguish between the placebo effect, an observed effect on specific outcome measures (e.g., the difference between a placebo-treated group and an untreated group in a study), and the placebo
201
202
The Placebo Response
response, which is a set of endogenous processes (brain and body mechanisms) recruited by placebo-induced expectations and conditioning. By this definition, placebo effects are directly observed in studies, and placebo responses are psychophysiological processes whose nature must be inferred by observing placebo effects. A placebo effect may thus provide evidence for an underlying placebo response, though the strength of that evidence depends on the nature of the comparison performed in the study and range of alternative explanations. (We note that many scholars use the term ‘placebo response’ to refer to any improvement on a placebo treatment, which is not the sense in which we use the term here.) The potential clinical significance of the placebo response has led to the standard use of placebo groups in clinical trials for therapeutic drugs and procedures. Clinical trials are experimental studies performed to test the therapeutic efficacy of medicines, surgical procedures, and other interventional procedures. The use of placebos has proven invaluable in providing a baseline against which to assess interventions; however, clinical trials are not typically designed to test the therapeutic effects of placebo treatments themselves. Thus, for reasons described below, they provide little evidence either for or against the existence of placebo responses as defined above. In an influential paper published in 1955 entitled ‘The powerful placebo,’ Henry K. Beecher analyzed a series of placebo-controlled clinical trials in order to examine the efficacy of the placebo itself. Stemming from this first effort, another class of experimental studies is designed to specifically assess the therapeutic effects of placebo treatment and patients’ expectancies, and these studies provide the most direct evidence for active psychobiological placebo responses. These studies offer scientists a unique opportunity to investigate interactions between the brain and body. Behavioral, pharmacological, physiological, and most recently neuroimaging methodologies have allowed us to begin to understand the mechanisms by which placebos exert their effects. This, in turn, offers scientists a window into understanding how the brain exerts control over behavior, emotional experience, and physiological processes in the body. This overview is devoted to a critical review of these studies.
False Placebo Effects Placebo effects have been the center of heated debates among researchers that have persisted in different forms throughout the past century. The central debate revolves around the issue of whether active, psychobiological placebo responses exist, and whether they affect health and disease processes in meaningful ways. Two alternative explanations for observed placebo effects have been offered: First, placebo effects in some studies may be statistical artifacts; and second, placebo effects in many studies may reflect changes in subjective reporting processes only, and not in meaningful disease processes. The argument that the so-called ‘placebo effects’ are statistical artifacts applies primarily to clinical trials that are not designed to test the efficacy of placebo treatments, and are thus not carefully designed to avoid such artifacts. The argument that observed placebo effects are subjective reporting biases applies to clinical trials and experimental studies of placebo alike. In this section, we outline each argument, and in the following section, we evaluate the experimental evidence on reporting biases in various disease processes. Placebo Effects as Statistical Artifacts In a typical clinical trial, patients are randomly assigned to treatment either with a therapeutic intervention (e.g., a study drug) or a placebo, and outcomes are assessed under ‘double-blind’ conditions: neither the patient nor the assessor knows which treatment a person is taking. There are many variants on this procedure, but the vast majority of trials share this element in common. Comparisons between the treatment and placebo groups are performed in order to estimate the active effects attributable to the treatment (treatment efficacy). Two assumptions are made in this comparison: First, that all nonspecific effects of being in the study (natural outcome improvement or worsening over time, effects of patient expectation and motivation, health care setting effects, etc.) are common to both treatment and placebo groups, and second, that nonspecific effects and drug effects combine additively, so that a simple subtraction between the two will yield the active
The Placebo Response
treatment effects. The critical question for placebo research is what those ‘nonspecific’ effects are, and how much of them can be attributed to causal effects of the placebo treatment itself (i.e., the placebo response). Put in concrete terms, placebo groups in studies of depression typically improve about 7–8 points on the Hamilton depression inventory (a clinical measure with a range from 0 to 54 points, with 24 or above indicating severe depression); but it would be an error to attribute all of this effect to the placebo response, because other factors also contribute to ‘nonspecific’ improvement in placebo groups. Some of these are (1) spontaneous remission, (2) natural symptom fluctuation, (3) regression to the mean, and (4) participant sampling bias. Natural history: spontaneous remission and natural symptom fluctuation Every disease has a natural history, a time-course that the disease state would follow without any intervention. Many disease states are time-limited, and patients experience spontaneous remission – most depressed patients, for example, eventually recover, as do those with anxiety disorders, sleep disorders, and many other conditions. Improvement in a placebo group in a clinical trial could therefore be due to spontaneous remission. Other diseases – such as hypertension, pain syndromes, Parkinson’s disease, irritable bowel syndrome, and many others – do not have high spontaneous remission rates, but the signs and symptoms fluctuate over time. A study may end while a patient is relatively symptom-free, which would look like healing attributable to placebo, but may in fact be a transient improvement after which symptoms may reappear. Regression to the mean Regression to the mean refers to the observation that when measurements are repeated, subsequent measurements are likely to be closer to the mean than the initial ones. This phenomenon is especially applicable to clinical improvements in placebo groups. If a patient seeks treatment and is enrolled in the study when his or her disease is worse than average, the patient’s state is likely to improve by the time the disease is next measured due to the natural course of the condition, rather
203
than anything having to do with placebo administration. It is tempting in many contexts to look at subgroups of individuals on the basis of their initial disease state – that is, the most depressed patients – and follow them over time. But this is problematic due to regression to the mean: If one were to follow the most severe patients in the placebo group in almost any disease state, they would be seen to improve over time. These first three factors demonstrate the importance of including a natural history (nontreatment) control group in clinical trials for assessing placebo effects. If a no-treatment group is included, then a comparison between the placebo-treated and the no-treatment group can be used to assess the effects of the placebo per se. Unfortunately, very few clinical trials include such groups, because the placebo effect itself is not of primary interest in these studies. Participant sampling bias Participant sampling bias comes about as a result of the fact that participants who experience beneficial results from a treatment are more likely to continue in the study and adhere to treatment regimens than those who experience either no effect or adverse effects. Thus, the participants who complete clinical trials in placebo groups (as well as active treatment groups) are more likely to be those who improve over the course of the study. The result is that participants may appear to improve over the course of the study, but this apparent improvement actually reflects changes in the sample over time.
Placebo Effects as Reporting Biases Another important source of potential error has to do largely with the important subjective aspects of many conditions in which the placebo effect is studied. In pain, depression, and nearly every illness, an essential component of the illness is the subjective experience of the condition. In some domains, there are clear objective measures of current state, such as motor performance or outwardly observable physical symptoms, but in many important clinical states objective measures are not available. For example, because pain is a subjective experience, clinical and experimental measures of pain are based on patients’ reports.
204
The Placebo Response
Physiological measures based on skin conductance, heart function, and pupillary response can be measured, but they are indirectly related to pain; thus, while they may be more precisely measured in some cases, they cannot be assumed to accurately reflect the pain experience. The problem with self-report-based measures is that they might be influenced by a number of factors that are not central to the disease process being studied. Thus, demonstrating placebo effects in subjective outcomes may say little about the power of placebo treatments to effect meaningful changes in disease progression. For example, a treatment that improves subjective well-being in cancer patients may be beneficial for this reason alone, but its viability as a specific treatment for cancer must certainly rest on its effects on the growth of the cancer. A few of the factors that create biases in subjective outcome assessment are described below. Ruling out reporting biases as causes of observed placebo effects is difficult in many cases; we address relevant evidence in the following sections. Demand characteristics and Hawthorne effects Demand characteristics refer to changes in patients’ reports and behavior on the basis of their perception of how they are expected to behave. Often, these expectations are communicated by inadvertent cues from the investigator: Even a very subtle nod or widening of the eyes from a physician may communicate agreement or surprise on the physician’s part. Such influences can be avoided by keeping experimenters and physicians blind to condition, when possible. In one interesting study, for example, physicians’ expectations about a drug’s effectiveness were shown to affect how effective patients thought it was in relieving pain. Other types of cues about expectations might result from the study procedures themselves or the way questions are worded. The question ‘‘How much did your pain decrease?’’ implies that it should have decreased to some degree. Being observed often causes changes in behavior. Such changes are classically referred to as Hawthorne effects, after a landmark study that showed workers’ performance improved dramatically once they were under observation in the
study. Hawthorne effects generally result from social influences that affect participants’ reports and behavior. Male experimental participants, for example, will tolerate higher levels of painful stimulation when an experimenter is watching them. Demand characteristics and reporting biases cover a range of different psychological effects, including Hawthorne effects in some cases, social compliance effects (in which patients say what they feel should be said), self-presentation biases (individuals often say what makes them look better in the eyes of others), and self-consistency biases (consistency with past behavior, to avoid admission that past behavior was in some way incorrect). Response bias Many experiments have demonstrated that decisions about the presence or absence and intensity of a stimulus (such as a disease symptom) are not a function of a stimulus alone, but of prior beliefs and the relative benefits and costs of the decision. This approach is the basis of signal detection theory (SDT). Consider the following simplified scenario. A patient is given a medication and is asked to judge whether the drug relieved pain. In simplified form, the patient truly feels different or she does not, and she may choose to respond ‘‘Yes, the drug helped’’ or ‘‘No, it did not help’’. A ‘yes’ response might be true pain relief or a false affirmative (an error). A ‘no’ response might reflect a true lack of pain relief or a false negative (another kind of error). The answer that the patient reports depends partly on the how well the patient can discriminate true from null effects, and partly on the relative costs and benefits of making the two kinds of error. The SDT framework acknowledges that a cost–benefit analysis influences patients’ decisions on what to report, and provides a way of estimating both the discriminability of true versus null effects and the patient’s response bias towards responding either ‘yes’ or ‘no.’ SDT analysis has been employed since the late 1960s in order to investigate whether placebo administration alters sensitivity to pain (measured by the ability to discriminate between different levels of noxious stimulation), response bias, or both. These studies suggest that the placebo effect is actually a change in response bias alone, without any change in sensitivity. A complicating factor is
The Placebo Response
that a change in ‘response bias’ after placebo treatment amounts to subjects reporting that noxious stimulation simply feels less painful after the treatment, even if they are no less accurate at discriminating different levels of intensity. These studies cannot provide evidence on whether those reports of feeling less pain occur because of changes in pain processing in the brain and spinal cord, or simply because of a cost–benefit analysis involved in the reporting decision. One way of disentangling these alternatives lies in the direct measurement of brain responses to painful stimuli, and we return to this alternative below.
205
studies have found placebo effects in immune function, cortisol levels, and growth hormone levels. In Parkinson’s disease, placebo effects have been reported in disease-relevant neural and neurochemical activity and in clinical signs of disease severity. In major depressive disorder, treatment with placebo induces changes in brain activity that mimic treatment with an active drug. Finally, in asthma, placebo effects have been reported in clinical measures of airway responsiveness. It is possible that placebo effects exist in other physiological processes; more research remains to be done to demonstrate the causal effects of placebo treatment in these and other areas.
Active Placebo Outcomes
Placebo Effects in Pain Physiology
It is clear that placebo treatments may influence subjective reports and observable conditions without affecting any underlying disease process. That is not to say that placebo effects on some outcomes, such as reported relief from pain or improvement in reported quality of life, are not valuable in their own right: In many cases, these outcomes are central to the comfort and happiness of patients, and to their ability to work productively and maintain positive social relationships. Nonetheless, it is valuable to review evidence for placebo effects on measurable biological processes in the brain and body. These are most often likely to be essential elements of a clinically meaningful placebo response in disease progression. In this section, we review some of the evidence for placebo effects on physiological outcomes, focusing on physical pain – one of the best-studied outcomes experimentally – as a model system. In brief, placebo effects have been reported on physiological indices of nociceptive (pain-related) processing, including measures of brain activity and neurochemical responses to painful stimuli, and placebo analgesic treatments have been shown to interact with active pain-relieving drugs. In clinical studies, there is a growing body of literature demonstrating placebo effects on reported clinical pain (a clinically meaningful outcome) in pain disorders such as irritable bowel syndrome. In other domains, particularly when conditioning is involved and sensory or internal cues become associated with active treatment, experimental
While placebo effects have been demonstrated in a wide range of domains, the majority of laboratory studies demonstrating causal effects of placebo treatment have been conducted in the context of pain. Pain is a common and debilitating condition with great biological significance for all organisms. Though it typically has a peripheral, physiological origin and physiological consequences in the body, pain is a subjective experience thought to be determined by an interplay between sensory, affective, and emotional brain systems. Because pain can be manipulated experimentally, and because much is known about the neural and physiological correlates of human and animal nociceptive processing, pain is uniquely suited for experimental investigations of the placebo response at multiple levels, from the involvement of specific endorphins and neurotransmitters, to the contributions of different brain regions, to the motivational significance of placebo effects for an individual. Because of the clinical significance of pain experience, the vast majority of placebo studies to date have demonstrated effects on subjective experiences of analgesia or pain relief. The classic example of placebo analgesia is when a person is (1) experiencing pain, which may be ongoing or caused by noxious (normally painful) stimulation in a laboratory study; (2) the person is given a placebo treatment, often with information indicating that the treatment will diminish pain; and (3) pain under the placebo treatment is compared with pain in an otherwise comparable no-treatment
206
The Placebo Response
condition, either in the same or different individuals, and pain reports decrease reliably with the placebo. However, in spite of the predominant (and clinically appropriate) use of reported pain as a primary outcome, several lines of research demonstrate placebo effects in objective physiological outcomes. Early studies of placebo analgesia, in the late 1960s, suggested that placebo treatment affected pain reports, but not the ability to discriminate stimuli of different temperatures from one another. This research was done in the tradition of thenrecently developed SDT, which provided a method to separately measure an observer’s ability to discriminate stimuli of two or more categories and the so-called ‘response bias’ in reporting on the basis of costs and benefits of making different types of errors. Because placebo effects were found in ‘response bias,’ but not discrimination ability, it was concluded that placebo treatments may amount to nothing more than an increase in the decision criterion for reporting a given stimulus as painful. However, another set of studies published in the late 1970s provided partial refutation of this conclusion by providing evidence that placebo treatments resulted in the release of brain opioids – endogenous (produced by the brain) neuropeptides known to play a key role in clinical pain relief. In these studies, the researchers gave participants a placebo treatment, which produced analgesia; but when they gave the same placebo treatment along with naloxone, a drug that blocks opioid receptors, the placebo analgesic effect was eliminated. These findings suggested that endogenous opioid release was a necessary component of at least one type of placebo analgesia (that elicited by verbal instructions and consequent expectancies of pain relief). Because opiate drugs are among the most widely used and best-known treatments for clinical pain and because opioids are involved in inhibition of pain-related neural signals at the earliest stages of processing in the spinal cord, the broader implication was that placebo treatment must have affected the brain physiology of pain. These studies were replicated and extended in the late 1990s and early 2000s. Although naloxone can itself block pain or enhance pain depending on the dose and perhaps on the psychological conditions of the study, several studies demonstrated that naloxone could reverse placebo analgesia
without otherwise affecting reported pain, and that placebo analgesic effects were specific to the particular body site to which they were applied. This latter finding implied that placebo analgesia in these paradigms was not a general, ‘global’ response to the placebo treatment, but rather was mediated specifically by expectancies that pain would be reduced at a particular body site. Another important set of findings was that placebo effects could be driven by expectations or conditioning with an opiate drug, in which case they tended to be naloxone-reversible and thus implicate endogenous opioid release, or they could be created by conditioning with nonopiate analgesics, in which case they do not appear to be naloxone reversible, implicating nonopioid mechanisms. While the inferences from these naloxone studies suggest that active endogenous opioid processes play an important role in the placebo response, these indirect inferences fail to fully resolve the question of whether pain processing itself is affected under placebo. Opioids could affect pain reports in nonspecific ways or offset the aversiveness of pain without affecting specific nociceptive processing. In order to resolve whether nociceptive processing is affected, changes in the brain processes underlying pain must be examined. Neuroimaging methodologies, including positron emission tomography (PET) and functional magnetic resonance imaging (fMRI), have allowed researchers to examine the brain’s activity in response to pain and to identify a network of regions involved in the pain response, frequently referred to as the ‘pain matrix.’ The first study to use neuroimaging to examine the placebo response used PET to identify a common network between placebo analgesia and opioid analgesia. When compared to a condition in which participants received noxious stimulation without any analgesic administration, both placebo and opioid conditions revealed increased activity in rostral anterior cingulate (rACC) and the brainstem, two regions that are known to play important roles in pain modulation. A second approach to the study of the placebo response assumes that if the placebo response truly decreases pain processing, this ought to be evident through decreased pain-related activity in regions of the brain known to be responsive to pain. Thus,
The Placebo Response
researchers can contrast the response to stimulation under placebo with stimulation in a nonplacebo control condition and look for differences in pain-related neural activity. The first study to use such an approach used fMRI, and did indeed find evidence of placebo-induced decreases in regions of the pain matrix that have been shown to be associated with the affective components of pain, including rACC and insula, as well as thalamus. In addition to these decreases, placebo administration induced increases during pain anticipation in orbitofrontal cortex (OFC), lateral prefrontal cortex (PFC), and, importantly, an area of midbrain surrounding the periacqueductal gray (PAG), a structure rich in endogenous opioids that is known to play an important role in pain modulation. Decreases in thalamus and insula, however, occurred late in the pain response, near the time when participants were asked to make ratings of the pain they had experienced; this therefore did not fully resolve whether the pain process itself had changed or whether the observed changes had been mediated by evaluation processes. Researchers followed this initial investigation with studies employing methodologies with improved temporal resolution, including electroencephalography (EEG) and magnetoencephalography (MEG). Just as research using PET and fMRI has identified specific regions involved in the pain experience, EEG research has identified specific temporal components that are known to be associated with pain. Thus, a similar approach can be employed in EEG and MEG research to examine whether the placebo response changes pain processing; researchers can look for placebo-induced decreases in the amplitude of evoked nociceptive event-related potential components. Importantly, the temporal resolution of these methodologies allows researchers to specifically examine painrelated components that follow nociceptive stimulation quickly enough to be unaffected by subsequent modulatory cognitive processes (such as poststimulus decision making). This approach is particularly suited to examine a long-standing hypothesis known as the gate control theory, which hypothesizes that descending mechanisms inhibit nociceptive input at the level of the spinal cord, thereby preventing nociceptive information from ascending to the cortex for processing.
207
Several studies have employed this approach, and all have offered support for at least some decrease of pain processing as part of the placebo response. One study examined the effect of cues on perceived pain, and found that invalid cues (those inducing expectations for lower or higher pain than applied stimulation) modulated activity in components associated with activity in secondary somatosensory cortex (SII) that occurred just 160 ms after pain stimulation. However, later components reflecting activity in posterior ACC varied with stimulus intensity, but were not affected by induced expectations and changes in perceived pain. A study that examined two of the major components of laser-evoked pain potentials, the P2 and N2, found that the amplitude of the P2 response was decreased with placebo treatment. Importantly, the P2 differences habituated over the course of the experiment, although placebo differences in reported pain persisted, suggesting that effects on early nociceptive processing and reported pain are dissociable. Furthermore, the extent of P2 decreases was less than that which would be expected if the gate control theory entirely explained changes in pain processing under placebo. These results suggest that while the placebo response does directly affect nociceptive processing, cognitive components are also important in the placebo response, as indicated by the large and persistent placebo effects on reported pain, relative to the smaller effects on P2 amplitude, which habituated over time. Since these initial findings, recent neuroimaging studies have added to our understanding of how the placebo response affects neural processing. Similar to the rationale behind the high temporal resolution studies reviewed above, one study sought to resolve whether placebo-induced decreases occurred early in the pain period. Rather than using a methodology with high temporal resolution, the authors used fMRI but acquired images only during nociceptive stimulation, and stopped the scanner during the pain rating periods. Decreases were observed in the right mid-insula, medial thalamus, and ACC during painful stimulation, again suggesting that the placebo response can indeed decrease nociceptive processing in the brain. Together, these studies provide evidence that the placebo response modulates pain processing,
208
The Placebo Response
and begin to provide evidence of how this modulation may occur. Consideration of placebo responses in other domains allows us to consider candidate central and proximal mechanisms for these observed effects. Physiological placebo effects in other domains A controversial meta-analysis of clinical trials that compared placebo groups with no-treatment control groups found evidence of beneficial placebo effects only in subjective outcomes, particularly in studies of pain. The authors asserted that this was presumably a result of reporting biases and artifacts. The series of neuroimaging studies of placebo analgesia reviewed above clearly demonstrate that the placebo response does indeed influence physiological processing of nociceptive information. However, the subjective component of pain may play a critical factor in the etiology of the placebo response. In order to determine the kind of effect the placebo response may have on outcomes without any subjective component, a meta-analysis was conducted that examined placebo effects in clinical trials of peripheral disease processes that used outcome measures assessing the state of peripheral organs, tissues, and body fluids was conducted. There was a significant positive effect of placebo administration on physical processes, such as blood pressure and forced expiratory volume (a clinical sign of asthma severity), but no consistent placebo effect was demonstrated in biochemical process measures, such as cortisol and cholesterol levels. Thus, placebo effects appear to be strongest in physiological processes that can be directly controlled by the brain, and their effects on biochemical processes in the body are likely to be substantially weaker. However, these findings do not conclusively show that there are no effects on biochemical processes. First, some of the individual clinical trials included in the meta-analysis may indeed have shown support for placebo effects on biochemical processes, but there may have been significant heterogeneity among studies, causing the overall results of the analysis to be null. Second, a limitation of analyses of clinical trials is that they are not designed to study placebo effects, and expectancy effects in these studies may in some
cases be very weak. Laboratory studies use instructions and conditioning to elicit positive expectations (i.e., ‘This cream is known to be effective in relieving pain’ for the placebo, ‘This is a control cream with no known effect’ for the control), whereas in double-blind clinical trials, participants are told that they may receive either an active agent or an inert substance. Consistent with this notion, another meta-analysis has shown that laboratory studies of placebo mechanisms elicit stronger placebo effects than clinical studies in which placebo groups are used as a control. Experimental investigations have demonstrated that placebo treatments can alter peripheral physiology across a variety of conditions. In asthma, placebo bronchodilator administration increases forced expiratory volume, a clinical measure of lung function. Preconditioning with active agents has been shown to elicit powerful placebo effects in the domains of growth hormone and cortisol secretion, as well as immunosuppression. Placebo analgesia not only affects pain-related physiology in the brain but has also been shown to reduce b-adrenergic activity in the heart. Experiments designed to investigate nocebo effects – whereby expectations for increased pain or negative outcomes lead to hypersensitivity or increased symptomatology – reveal that nocebo responses in pain involve hyperactivity of the hypothalamic–pituitary–adrenal axis, as evidenced by increased cortisol release and adrenocorticotropic hormone.
Mechanisms of Placebo Effects The evidence reviewed above powerfully supports the existence of a placebo response, albeit one that varies in magnitude across outcome measures and disease processes, and with the strength of expectancy manipulations and use of conditioning procedures. However, a mechanistic understanding of the placebo response is critical to understanding how, in what outcomes, and under what conditions placebo responses occur. In summary, the placebo response can be divided to two mechanistic levels: central mechanisms and proximal mechanisms. Central mechanisms are those that may operate across domains (i.e., in pain, Parkinson’s disease, and depression), whereas proximal mechanisms
The Placebo Response
tend to be domain-specific. We will again employ pain as a model system, since much is known about the central and peripheral pathways of pain processing.
Central Mechanisms Central mechanisms are those involved in translating verbal instructions into positive expectancies and maintaining the activity corresponding to those expectancies in the brain. They are the mechanisms of generating and maintaining positive beliefs, which are likely to operate in similar ways across disorders. Conditioned placebo effects may involve additional, separate mechanisms. Studying placebo effects can thus provide a window into how these basic processes work, and how they shape mind–body interactions across a range of conditions. Expectancy versus conditioning: two routes An important theoretical debate about the central mechanisms subserving the placebo effect seeks to understand the psychological processes that give rise to the placebo response. Do conscious expectations mediate the placebo effect, or are changes in placebo due to classical conditioning? We will review the arguments of each perspective, noting that they are not mutually exclusive; both general expectancies and specific learning (conditioning) are likely to play important roles in placebo effects. Conditioning and the placebo effect
Some researchers have argued that placebo effects are a result of conditioning. In the most common understanding of classical conditioning, an unconditioned stimulus (US) that normally elicits a certain response (the unconditioned response, or UCR) is paired with a neutral stimulus that elicits no response on its own. Over repeated pairings of the two stimuli, the previously neutral stimulus comes to elicit the same response as the US; the previously neutral stimulus is then referred to as a conditioned stimulus (CS), and the evoked response is a conditioned response (CR). Conditioning can occur in aversive contexts (in fear conditioning, a light may be paired with a shock,
209
to elicit freezing in response to the light) or appetitive contexts (as in Pavlov’s classic experiments, food can be paired with a tone and animals eventually salivate in response to the tone). In the case of the placebo effect, there are several routes by which conditioning may result in placebo effects. For example, a pharmacological agent (US) might be administered in the form of a pill (CS) to elicit the specific effects of the drug (UCR); subsequent administration of the pill without the active pharmacological ingredients might still elicit the same effects. Researchers have even suggested that the medical context itself can serve as the conditioned stimulus; thus, people may associate intrinsically neutral items such as syringes or even doctors with the changes associated with treatment, and placebo effects may result from conditioning to these contextual stimuli. Proponents of the classical conditioning model of placebo have argued that a lifetime of medical treatments serve as conditioning trials to pair the medical context (CS) with therapeutic effects (CR). Support for the classical conditioning model of placebo comes from research demonstrating that placebo effects are stronger after exposure to active drugs, that placebo analgesics are more effective when labeled with well-known brands, and from research that shows that some side effects of drugs that are unlikely to be consciously perceived or mimicked – such as respiratory depression following opiate treatment – may be reproduced by placebo treatments after conditioning. In addition, some placebo effects, such as increased cortisol and growth hormone after conditioning with an active drug, are not reversible by telling subjects that the medication is a placebo, suggesting that some learning has occurred that is not modifiable by conscious expectancies. Brain mechanisms of conditioning
While conditioning has been studied for more than half a century, and we know much about the neural circuitry involved in fear conditioning, mechanistic research that would directly support a conditioning model of placebo is quite scarce. Primary support for a model in which conditioning recruits endogenous mechanisms comes from studies of conditioned immunosuppression. In a series of studies in rats, saccharin was paired with an
210
The Placebo Response
immunosuppressive agent, and the rats that had been preconditioned exhibited decreased antibodies when given saccharin in a later test phase. A similar approach has been used to show evidence of conditioned immunosuppression in humans, measured by immune factor expression in mRNA and lymphocytes (white blood cells). These studies offer support for conditioning-based placebo effects on immune responses. However, while this offers some insight into the role of conditioning in endogenous processes, little is known about brain mechanisms specific to conditioning-based placebo responses in humans, and mechanisms supporting conditioned immunosuppression are unlikely to generalize to other domains, such as pain. In one PET study mentioned earlier, researchers compared brain responses to painful stimulation under opioid-based analgesia with responses to placebo analgesia. Opioid administration always preceded the placebo analgesia condition, which may have induced a conditioning-based placebo effect. Brain responses to each were compared with a pain control condition, and both were associated with increased activity in rACC and increased rACC-brainstem connectivity. While promising, this and other studies have not directly compared conditioning processes with nonconditioning expectancy manipulations (verbal instructions only), and the nature of the conditioning-specific placebo response remains yet to be elucidated. Expectancy and the placebo effect An alternative view proposes that conscious expectancies mediate the changes associated with placebo effects. In this view, the internal beliefs and expectations associated with the inert treatment are responsible for the endogenous regulation of processes in order to produce the requisite changes associated with placebo response; put simply, one experiences changes associated with placebo administration because one expects to. Expectancies involve appraisals of the significance of a stimulus or event in the context of its anticipated outcome. For the most part, expectancies are conscious at the time when decisions are made, or, if they are not conscious (as may happen during rapid decision making), expectancies can be brought into consciousness when attention is drawn to them. This is an important distinction
from conditioning theories, which assume that organisms need not have conscious awareness of contingencies between stimuli in order for conditioning to occur. Thus, one way to define the distinction between expectancy effects and conditioned responses is that expectancy effects depend on the participant’s state of mind, whereas conditioned responses do not. By this definition, expectancybased effects can be altered by instructions to subjects, whereas conditioned effects cannot. The power of expectancies has been illustrated in many areas of research, from basic perception to complex physiological processes. One way that researchers have learned about the influence of expectations on physiological processes is by comparing the effects of hidden and open drug administration. Positive expectancies (expectations for relief) are active when patients are aware that they are receiving a given drug, as is the case when drugs are administered in full view of the patient. When this is contrasted with conditions in which drugs are administered surreptitiously (e.g., when a drug is administered intravenously under the guise of saline administration), researchers generally find that the open administration has a greater beneficial effect. When the contextual cues surrounding treatment are the same in both open and hidden cases, a conditioning-based explanation for the effect is unlikely. A practical application of this research is evident in hospitals, where patients are allowed to self-administer analgesic agents such as morphine; it takes far less morphine to produce the same painrelieving effect when patients control drug delivery and expect relief than when doctors administer the drug without patient expectations. The same positive expectancies may therefore be the driving force behind the power of placebo. The expectancy model of placebo is supported by research demonstrating that placebo effects can occur with verbal instruction alone (i.e., without prior experience with a drug or active treatment). In addition, in some studies, placebo effects on pain that have been induced through a conditioning procedure have been reversed completely by revealing to the participants that the placebo treatment was a sham. Thus, the placebo effects in these studies do not meet the criteria for conditioned responses (involving specific learning not modifiable by beliefs).
The Placebo Response
It is also worth noting that for conditioning to occur, the brain must be capable of learning an association (i.e., forming a pathway) between a CS and either the UCS or the UCR, and reactivation of that pathway must be able to elicit the UCR. Placebo responses do not always fit these criteria. For example, during the extinction phase of conditioning, a conditioned stimulus that is not reinforced will cease to elicit a conditioned response; however, placebo effects can remain far longer than extinction would permit. In other cases, an association formed over repeated experiences can be reversed immediately by a change in instructions, which is not consistent with models of conditioning. Brain mechanisms of expectancy
Placebo instructions change the cognitive context in which pain stimulation is perceived, and these altered appraisals of the situation give rise to changes in expectations about pain, harm, and pain relief. The brain mechanisms of such expectations are likely to be similar to those involved in executive functions – basic cognitive processes coordinating the maintenance and manipulation of information. Information about context is known to require dorsolateral prefrontal cortex (DLPFC), which interacts with working memory – systems for maintaining information in an active state in the brain – in order to maintain expectancies induced by placebo manipulations. Expectations about the value of upcoming stimuli are also critical to the pain process and are potentially highly involved in the placebo response across domains. Effective placebo administration induces expectations for reduced symptomatology or diminished pain, which affects how the brain processes the condition or stimulus. The processes most likely to be altered are those that assign value and meaning (for the self, or survival) to the stimulus. Orbitofrontal cortices and rACC have been shown to be highly involved in the process of valuation. Brain representations of active contextual information and stimulus value may directly or indirectly influence more basic perceptual, behavioral, and somatic (peripheral) processes through connections with other parts of the brain that represent percepts, motor actions, and somatic states.
211
By drawing on knowledge from brain mechanisms of executive function, researchers can define reasonable hypotheses about brain mechanisms supporting an expectancy-based placebo response. These can be tested by contrasting anticipatory activity in a placebo condition to anticipatory activity during a control condition, so that one can identify processes related to pain expectancy that are shaped by placebo treatment. This approach was used in an fMRI study of placebo analgesia, which revealed increases in DLPFC, OFC, and rACC activity during anticipation of pain with placebo. These anticipatory increases correlated with placebo effects on reported pain, and anticipatory increases in DLPFC and OFC correlated with subsequent placebo-induced reductions in brain activity during thermal stimulation. Other studies have replicated and extended this result, showing that placebo treatments for negative emotion activate the same brain regions, and that endogenous opioids – neurochemicals linked to relaxation, euphoria, and pain relief – are released in these regions following placebo treatment. The expectancy versus conditioning debate It is difficult to resolve the relative contributions of expectancy and conditioning to placebo effects, because the two are not always mutually exclusive; in some cases, conditioning procedures are likely to shape both learning and expectations. There are two ways to distinguish between learning and expectancy mechanisms: One relies on behavioral observations and the other on measurement of the brain. Earlier, we suggested that conditioning results in learning that persists over time, in spite of expectancies; when a CS is presented without the UCS, extinction of the CR is relatively slow. Thus, effects that can be reversed in a single trial or affected by verbal instructions are not likely to be the result of conditioning, but rather expectancies. In a classic placebo study, expectancies of pain relief were manipulated by surreptitiously turning down the stimulus intensity during testing of a topical placebo solution. Placebo effects on pain developed over the course of this manipulation. However, some participants were informed that the intensity was being reduced during the placebo administration, while others were told that it was
212
The Placebo Response
not; this latter group presumably attributed the reduction in pain to the placebo. Although the physical stimuli were identical for the two groups, including the putative CS (placebo application) and UCS (reduction in pain), placebo effects were about 7 times as large when the verbal instructions led participants to expect large reductions in pain attributable to the placebo. Other studies have demonstrated that ‘conditioned’ placebo effects in pain and even basic fear conditioning in humans are reversible by changing the instructions to subjects, suggesting that expectations are mediating the effects rather than automatic, learned associations. However, some kinds of placebo effects are not reversible by changing the instructions. In another classic study, researchers repeatedly injected sumatriptan, a drug that induces cortisol and growth hormone release, thereby forming an association between the injection and the drug response. After this conditioning procedure, injecting saline alone elicited cortisol and hormonal increases, and these increases were not blocked by changing the verbal instructions. A second way to discriminate between conditioning and expectancy is by measuring brain activity itself. The patterns of placebo-induced activity increases in OFC and rACC, and increases in DLPFC, suggest that general mechanisms of appraisal and expectancy are at work. Such effects have been found in pain and, though less well studied, depression; in pain, these same expectancy- and appraisal-related regions have been shown to exhibit placebo-induced opioid release. A difficulty, however, is that there is no way to ensure by looking at the brain that these responses are not the result of some conditioned association being activated. Another difficulty is that it is currently difficult or impossible to measure learned associations directly in the human brain; whereas synapse strength, gene expression, and other molecular markers of learning can be investigated in animal models, the techniques for probing them are invasive and cannot be used in humans – and, in addition, the area in the human brain where cellular learning underlying placebo effects may be taking place is still unknown. Other central mechanisms Opioid release in anticipation of pain under placebo correlates with dopamine release in the
nucleus accumbens – a region known to be highly involved in reward processing. Similarly, the placebo response in Parkinson’s disease has been linked to increased dopamine release in this region. A promising new theory of placebo has thus suggested that placebo administration may actually be rewarding, and that it is this positive affective shift that results in observed placebo effects on pain and motor performance in Parkinson’s disease. Work demonstrating correlations between dopamine release under placebo with subsequent activity in a reward task shows that individual differences in reward response correlate with individual differences in placebo. Researchers have postulated that two activation systems that are mutually exclusive exist – a positive, approach system (behavioral activation system), and a negative, withdrawal system (behavioral inhibition system). Placebos may induce a shift from a withdrawal system to an approach system, and this may induce concomitant effects on pain, depression, Parkinson’s disease, and other conditions, which would be observed as placebo responses.
Proximal Mechanisms Proximal mechanisms refer to the pathways whereby central mechanisms interact with the actual processing of a stimulus or condition in order to elicit changes in domain-specific activity. Proximal mechanisms are therefore likely to be different for different disorders. We will specifically review proximal mechanisms for pain, depression, and Parkinson’s disease, as these placebo responses are the most well understood. Proximal placebo mechanisms in pain A central question in the study of placebo analgesia has concerned the level (or levels) of the pain pathway at which the placebo response has its effect. The naloxone studies reviewed earlier illustrate the important role of endogenous opioids in the expectancy-based and opioid-conditioned placebo responses. Endogenous opioids facilitate modulation of both descending pain-control circuits and central processing of nociceptive information, illustrating their significance as one of the key mediators of the placebo response in pain.
The Placebo Response
The gate control theory and its successors postulate that descending modulatory signals under placebo can cause inhibition of nociceptive processes in the spinal cord, before signals reach the brain. In animal work, a specific type of opioid receptor, the m-opioid receptor (MOR), has been shown to inhibit nociceptive transmission at the level of the spinal cord’s dorsal horn. The PAG, a region that has been repeatedly demonstrated to be involved in placebo analgesia, is rich in MORs and plays an important role in this descending modulation in animals. As mentioned earlier, an fMRI study that directly examined the placebo response found placebo increases in an area of the midbrain surrounding the PAG during anticipation of pain. This activity could be consistent with the gate control hypothesis, in that the placebo context would increase opioid release by the PAG, and descending opioids would inhibit subsequent pain at the level of the spinal cord’s dorsal horn. Further support for the role of the PAG in the placebo response comes from PET studies showing that placebo treatment causes increased opioid release in PAG during pain. Spinal inhibitory mechanisms in humans are also supported by a study showing that placebo manipulations can reduce the area of skin that develops hyperalgesia with repeated thermal stimulation. Animal studies have shown that this hyperalgesia is due to spinal neuron sensitization; hence, placebo effects that decrease the area of hyperalgesia are considered as evidence of descending modulation at the level of the spinal cord. While these studies and others suggest that early modulation of ascending nociceptive signals may play a key role in placebo analgesia, it is important to recognize that this may not be the only factor responsible for observed placebo effects in pain. The EEG results reviewed earlier suggest that while early inhibition does indeed occur, it cannot entirely account for the magnitude of decreases in reported pain. Recent studies have taken a mechanistic approach to investigating the role of endogenous opioids in the placebo response, elaborating on the knowledge available from naloxone studies. These studies have shown where in the brain placebo changes opioid release, strengthening the argument for placebo changes in central nervous system processing. In PET studies of placebo analgesia that
213
measured MOR activity directly, PAG opioid activity both decreased in anticipation of pain under placebo and increased during painful heat, suggesting that placebo might diminish the threat associated with upcoming nociceptive stimulation and enhance opioid release induced by pain. These studies also demonstrated that placebo treatment increases m-opioid system activation in the OFC, perigenual ACC, rACC, right anterior insula, left dorsolateral PFC, thalamus, amygdala, and the nucleus accumbens, suggesting that placebo treatments affect opioid responses in many regions critical for affective valuation. The rACC and insula regions corresponded to those that had been shown to decrease during pain in the aforementioned fMRI experiments, suggesting that these changes in pain responses may have been opioid-mediated. In one of these studies, placebo increased connectivity between PAG and rACC, as well as other distinct subsystems of correlated regions, suggesting that placebo treatment involves central opioid release that increases functional integration across regions. These effects could influence the central processing of pain above and beyond effects mediated by descending spinal control systems. Proximal placebo mechanisms in major depressive disorder In depression, placebo effects on the brain have been examined by using PET imaging to measure baseline metabolic activity before, during, and after treatment with either placebo or an active medication. Many changes that were observed as part of successful treatment with the active drug were also observed in placebo responders, including metabolic decreases in subgenual anterior cingulate. This region has been shown to be consistently affected in depression and is a target of deep-brain stimulation in patients who do not otherwise respond to treatment. Other common sites of activity included increases in prefrontal, parietal, and posterior cingulate cortex. Importantly, these common results differ from patterns of brain activation over the course other types of treatment, such as cognitive behavioral therapy. This suggests that both active drug and placebo treatments work in part by changing central systems involved in affective valuation and motivation. Much more work remains to be done
214
The Placebo Response
to unpack the brain mechanisms involved in both verum and placebo treatment for depression. Proximal placebo mechanisms in Parkinson’s disease In Parkinson’s disease, active placebo responses are likely to involve dopaminergic pathways. PET studies of dopamine D2 receptor activity have provided evidence that placebo treatments lead to dopamine release in the striatum. Complimentary evidence has been obtained from neurosurgical studies, in which placebo (sham) stimulation of the subthalamic nucleus – a stimulation site used in the treatment of Parkinson’s – has been shown to affect both subthalamic nucleus activity (decreased bursting and neuronal frequency discharge) and muscle rigidity, which is a clinical sign of the disease. Recent evidence indicates that expectations of effective treatment affect dopamine levels and muscle activity in a dose-dependent fashion. Whether dopaminergic mechanisms play a role in placebo effects in pain, depression, and other domains remains to be tested more thoroughly. It has been proposed that dopamine release underlies positive affective and motivational shifts that lead to improved outcomes across disorders. Whether outcomes can be influenced by changes in affective and motivational states may determine whether placebo treatments are effective; ongoing research is now being conducted to test this hypothesis.
Summary While many factors can potentially lead to observed placebo effects without changing underlying processing, careful experimental manipulations provide evidence of active placebo responses in the domains of pain, Parkinson’s disease, depression, asthma, and conditioned immunosuppression. There is clinical evidence for effects on other conditions, such as hypertension, anxiety, and heart function, but more experimental research is needed to assess whether placebo treatments can truly affect outcomes in these domains. Neuroimaging methodologies allow researchers to identify and examine active central and proximal mechanisms supporting the placebo response. This, in turn, provides a powerful window into mind–body interactions.
See also: Hypnosis and Suggestion; Psychoactive Drugs and Alterations to Consciousness.
Suggested Readings Benedetti F, Colloca L, Torre E, et al. (2004) Placebo-responsive Parkinson patients show decreased activity in single neurons of subthalamic nucleus. Nature Neuroscience 7(6): 587–588. Benedetti F, Pollo A, Lopiano L, Lanotte M, Vighetti S, and Rainero I (2003) Conscious expectation and unconscious conditioning in analgesic, motor, and hormonal placebo/ nocebo responses. Journal of Neuroscience 23(10): 4315–4323. de la Fuente-Fernandez R, Ruth TJ, Sossi V, Schulzer M, Calne DB, and Stoessl AJ (2001) Expectation and dopamine release: mechanism of the placebo effect in Parkinson’s disease. Science 293(5532): 1164–1166. Hrobjartsson A and Gotzsche PC (2001) Is the placebo powerless? An analysis of clinical trials comparing placebo with no treatment. The New England Journal of Medicine 344(21): 1594–1602. Kong J, Gollub RL, Rosman IS, et al. (2006) Brain activity associated with expectancy-enhanced placebo analgesia as measured by functional magnetic resonance imaging. Journal of Neuroscience 26(2): 381–388. Meissner K, Distel H, and Mitzdorf U (2007) Evidence for placebo effects on physical but not on biochemical outcome parameters: a review of clinical trials. BMC Medicine 5: 3. Montgomery GH and Kirsch I (1997) Classical conditioning and the placebo effect. Pain 72(1–2): 107–113. Petrovic P, Dietrich T, Fransson P, Andersson J, Carlsson K, and Ingvar M (2005) Placebo in emotional processing – induced expectations of anxiety relief activate a generalized modulatory network. Neuron 46(6): 957–969. Petrovic P, Kalso E, Petersson KM, and Ingvar M (2002) Placebo and opioid analgesia – imaging a shared neuronal network. Science 295(5560): 1737–1740. Ploghaus A, Becerra L, Borras C, and Borsook D (2003) Neural circuitry underlying pain modulation: Expectation, hypnosis, placebo. Trends in Cognitive Science 7(5): 197–200. Pollo A, Vighetti S, Rainero I, and Benedetti F (2003) Placebo analgesia and the heart. Pain 102(1–2): 125–133. Scott DK, Egnatuck CM, Wang H, Koeppe RA, Stohler CS, and Zubieta JK (2008) Opposite dopamine and opioid responses define placebo and nocebo effects. Archives of General Psychiatry 65: 220–231. Scott DJ, Stohler CS, Egnatuk CM, Wang H, Koeppe RA, and Zubieta JK (2007) Individual differences in reward responding explain placebo-induced expectations and effects. Neuron 55(2): 325–336. Stewart-Williams S and Podd J (2004) The placebo effect: Dissolving the expectancy versus conditioning debate. Psychological Bulletin 130(2): 324–340. Wager TD, Matre D, and Casey KL (2006) Placebo effects in laser-evoked pain potentials. Brain, Behavior, and Immunity 20(3): 219–230.
The Placebo Response Wager TD, Rilling JK, Smith EE, et al. (2004) Placebo-induced changes in fMRI in the anticipation and experience of pain. Science 303(5661): 1162–1167.
215
Wager TD, Scott DJ, and Zubieta JK (2007) Placebo effects on human m-opioid activity during pain. Proceedings of the National Academy of Sciences of the Unite States of America 104(26): 11056–11061.
Biographical Sketch
Lauren Y Atlas began her doctoral work at Columbia University in September 2006. She completed her undergraduate education in 2003 at the University of Chicago, where she worked with John Cacioppo in the Social Neuroscience Laboratory. Her undergraduate thesis research examined cardiovascular differences between lonely and nonlonely participants during active and passive coping. After graduating, she worked as fMRI project coordinator in Stanford’s Mood and Anxiety Disorders Laboratory under the direction of Ian Gotlib, where she was involved in projects investigating the neural bases of cognitive and affective processing in major depressive disorder, social anxiety disorder, and bipolar disorder. Her graduate work with Dr. Tor Wager at Columbia takes a mechanistic approach to the study of how expectancies modulate affective experience. Current projects use fMRI and psychophysiology methodologies to examine brain pathways mediating the relationship between applied nociceptive stimulation and subjective pain.
Dr. Tor D Wager received his PhD from the University of Michigan in cognitive psychology, with a focus in cognitive neuroscience, in 2003. He joined the faculty of Columbia University as an assistant professor of psychology in 2004. His primary research interest is in the neural and psychological bases of cognitive and affective control. His research quantifies behavioral performance and brain activity to investigate the neural mechanisms by which humans have flexible control over their behavior. This approach emphasizes the mutual constraints on interpretation afforded by studying behavior and functional anatomy at the same time. His main research interests along those lines are brain and psychological mechanisms underlying the cognitive control of pain and affect; individual differences in selective attention, inhibition, task switching, and other executive processes; and the relationship between affective regulation and cognitive control. He is also interested in developing image analysis and statistical
216
The Placebo Response
modeling methods that will improve our ability to use fMRI as a research tool in cognitive and affective neuroscience. Current projects along these lines include optimization of experimental design for fMRI experiments, meta-analysis of functional imaging data, nonlinear alternatives to hemodynamic response fitting, robust regression techniques in massively univariate linear models, and application of multivariate techniques.
Psychoactive Drugs and Alterations to Consciousness A Dietrich, American University of Beirut, Lebanon ã 2009 Elsevier Inc. All rights reserved.
Glossary Hallucination – Abnormal perceptual experiences that occur in the absence of a physical stimulus and while the person is awake and conscious. Opiates – Drugs that act on opiate receptors and mediate relief from pain and produce feelings of euphoria. Psychedelics – Drugs that induce hallucinations or have ‘mind manifesting’ or ‘expanding’ properties. Psychoactive drugs – Drugs that act on the central nervous system and alter behavior and mental processes. Sedatives and hypnotics – Drugs that depress or inhibit brain activity and produce drowsiness, sedation, or sleep; relieve anxiety; and lower inhibition. Stimulants – Drugs that produces behavioral arousal via a variety of neural mechanisms. Therapeutic index (TI) – A measure of the safety of a drug that takes into account the difference in dose between an effective dose and a lethal dose. Tolerance – A compensatory process in which the effect of a particular drug diminishes with repeated exposure to the drug. There are several types of tolerance, such as cross, metabolic, physiological, behavioral, and learned tolerance. Transient hypofrontality theory – A theory that proposes a common neural mechanism for all altered states of consciousness, the transient downregulation of brain activity in the prefrontal cortex.
Introduction Drugs have been used and abused to change consciousness since recorded history. Indeed their use
goes back probably much longer, as animals are known to seek out psychoactive substances. Alcohol, perhaps the oldest such drug, was discovered independently by most cultures and the psychological effects of many naturally occurring compounds were known to the Chinese, Indians, Egyptians, and Greeks. The pursuit of intoxicated happiness is so pervasive – even among other animals – that it has been called the fourth drive, after hunger, thirst, and sex. The history of drug use and abuse from antiquity to the present day drug pharmacology – especially their mechanisms of action – and the legal and health implications are all relentlessly fascinating topics, but this article is not the place to pursue these issues. This article focuses instead on the changes to phenomenology, with particular emphasis on the mind-altering effects induced by hallucinogens and psychedelics. Apart from these chemical altered states, this article also examines the problem of defining altered states in general and how they fit, with nondrug states, into an overall framework for altered states of consciousness (ASC).
Problems with Definitions Consciousness is notoriously hard to define. This makes ASC even harder to capture in a single neat definition. Several attempts have been made to crystallize the concept into a usable definition but, to put it bluntly, these efforts have not been widely adopted. In consequence, we have a poor understanding of the subject matter under investigation and a lack of agreement about the type of phenomena or states that should be counted as examples of ASC. From an empirical perspective, the matter gets worse due to the ephemeral nature of these altered states. Altered mind states are difficult to induce reliably in the laboratory and any effect an induction procedure might have on consciousness is even harder to observe, let alone quantify. In short, the independent variable cannot 217
218
Psychoactive Drugs and Alterations to Consciousness
be readily manipulated and the outcome measures rely on introspective verbal report – a set of circumstances that has led to little research being done on the subject. In addition, ASC is a topic that most neuroscientists do not see as a suitable subject for building a scientific career. There are, of course, exceptions – sleep research or experimental hypnosis, most prominently – but by and large research on altered states is not supported by funding agencies and, as a consequence, few psychologists or neuroscientists can afford to make them their primary area of interest. This polite neglect is also due to the bad reputation of ASC. Many view them with suspicion – some sort of abstruse psychopathology at the lunatic fringe frequented mostly by potheads and meditating yogis. This bias is made worse by the excessive use of esoteric language used to describe these mind states. Thus, an undeserved aura of lawlessness and unscientific hogwash sadly surrounds them. However, studying altered states is probably one of the best ways to study consciousness itself because most phenomena, mental or otherwise, are usually studied by manipulating them. As is the case for consciousness as a whole, there has been a resurgence of sound research on altered states in recent years. This is motivated also by evidence pointing at bona fide medicinal benefits for some altered states, such as hypnosis, meditation, or cannabis use. The advantage of this progress, apart from restoring some scientific legitimacy to the topic, is much exciting new data that informs our understanding of the phenomenology and brain mechanisms of ASC. The prospect of understanding altered states in terms of their neural substrates holds the great promise that we might clarify at a deeper level questions regarding the nature of altered states, such as, for instance, how ASC differ from normal consciousness and how they are all related to one another. Trying to fit states of consciousness into some kind of system – a continuum or hierarchy, perhaps – is not novel but the recent advances in cognitive neuroscience have made this prospect more possible than ever. Two earlier models, one by Charles Tart and a more recent, neurobiologically informed attempt by Allan Hobson arrange conscious states by positioning them in a multidimensional space. Hobson’s AIM model maps states
of consciousness in a brain–mind space along three parameters: Activation, an energy dimension that depends on the activity of the reticulo-thalamocortical system; Input–output gating, an information source dimension that estimates the extent to which external and internal information is being processed; and Mode, a chemical modulation dimension that estimates the mix of aminergic (norepinephrine and serotonin) and cholinergic (acetylcholine) influences, which runs from low aminergic to high aminergic. This is particularly useful for drug states as the mechanisms of action of most psychoactive substances that induce profound alterations to consciousness are well known. Tart’s phenomenal space consists of two dimensions: irrationality and ability to hallucination. Tart, who also coined the term and popularized the concept of ASC, sees states of consciousness as discrete entities. Some positions in this multidimensional space are stable and can be occupied for a long time, while others are unstable and cannot be occupied except for very brief moments. The twilight between sleep and waking is an example of such an unstable position. We cannot operate in this zone for long and consciousness quickly gives way to the stable state of either waking or sleep. This led Tart to propose the model of what he called state-specific sciences. The concept is that each state of consciousness (SoC) has its own reality, logic, and physical laws. Any one state of consciousness cannot, therefore, be understood by a person in a different state, as it is observer dependent. It follows that all knowledge, scientific or otherwise, is only relevant to the SoC in which it is produced and can only be understood by someone residing in that same state. This places consciousness in the center of science – a bold move indeed. If you find it challenging to wrap your mind around the idea of state-specific sciences and its implications consider that the reason for your doubts may have to do with you not being in the right states of consciousness! This position also argues against the commonsense notion that there is one normal state and all other states are altered states. There is, then, no qualitative difference among SoC, echoing the thinking of William James who wrote many years earlier: ‘‘Our normal waking consciousness is but one special type of consciousness.’’ This is in tune with others who
Psychoactive Drugs and Alterations to Consciousness
have also emphasized that ASC should not be considered higher, or lower, as the case may be, states of consciousness, because this introduces a value judgment without proper evidence. Another upshot arising from the new evidence on the neural underpinnings of ASC is the old hope that altered states could be defined, in the future, by objective criteria, perhaps through better neuroimaging tools or an entirely new technology not yet invented, rather than by introspective selfreport. In the meantime, ASC can only be defined in subjective terms. Two such definitions are commonly used in this context. The first comes from Tart, who defined ASC as ‘‘a qualitative alteration in the overall pattern of mental functioning, such that the experiencer feels his consciousness is radically different from the way it functions ordinarily.’’ A nearly identical definition is proposed by William Farthing, who identifies ASC as ‘‘a temporary change in the overall pattern of subjective experience, such that the individual believes that his or her mental functioning is distinctly different from certain general norms for his or her normal waking state of consciousness.’’ Although definitions such as these do seem to capture the essence of ASC, several obstacles arise that make it impossible to use them in scientific research. Because subjective definitions rely, by their very nature, exclusively on introspection, we can only determine for ourselves whether or not we have entered an altered state. This makes altered states unverifiable from an objective, thirdperson perspective. Consider a sample of troublesome questions that expose the problem clearly. To start, take two people who have the exact same level of alcohol intoxication (matched for variables such as tolerance, weight, etc.), but one claims to be in an altered state while the other vehemently denies it? Or what about antidepressant drugs? Is a Prozac state an ASC? It is easy to see that this is an unsatisfactory state of affairs. The same problem arises when trying to compare some altered experience to our default mode of consciousness. What is the default here, we might want to reasonably ask? Normal consciousness itself is highly variable and the baseline, if there is one for each person, differs not only among people but also with each person. Consider more examples highlighting this predicament. If someone is madly in love, should
219
we consider this feeling an ASC? In light of the above, subjective definitions, it is certainly not easy, let alone clear, to decide on this. We can draw out the same predicament for other situations, such as PMS? Apart from bodily changes, women often describe their symptoms in terms of alteration to cognition and emotion. Does that count as an ASC? As for drug states, scientists and users typically agree that drug states constitute ASC. However, even here this troublesome mess cannot be entirely avoided. There is an agreement – by consensus, not via objective criteria though – that psychedelics and opiates alter conscious experience, but what about some less spectacular substance, say, amphetamines, Valium, or ecstasy And even for the clear, putative cases – LSD, mescaline, PCP, etc. – we must consider factors such as dosage or tolerance before we can be sure that a person has entered an ASC. One final aspect of defining ASC is that all definitions, those above included, emphasize that altered states are temporary phenomena. Permanent changes to mental status, such as those that occur in neurological and psychiatric disorders, are usually not considered ASC, although it is typically assumed that consciousness is altered in these conditions. Drug abuse can, however, result in permanent brain damage. How to classify these changes to consciousness is not clear either.
Transient Hypofrontality Theory Given the problem with subjective definitions, one solution has been to sidestep defining ASC altogether. This is also typically done in consciousness research in general. By simply considering, without further examination, certain mental states altered states – dreaming, daydreaming, meditation, hypnosis, runner’s high, drug states, most prominently – researchers can make progress on the neural basis of these states in the hope of finding a better angle from which to consider ASC. One such approach has been the transient hypofrontality theory. The transient hypofrontality theory proposes a common neural mechanism for altered states that can be subjected to empirical study. The theory is explicitly based on functional neuroanatomy. It views consciousness as composed of various attributes, such as self-reflection, attention, memory,
220
Psychoactive Drugs and Alterations to Consciousness
perception, and arousal, which are ordered in a functional hierarchy with the frontal lobe necessary for the top attributes. Although this implies a holistic view in which the entire brain contributes to consciousness, it is evident that not all neural structures contribute equally to consciousness. This layering concept localizes the most sophisticated levels of consciousness in the zenithal higher-order structure: the prefrontal cortex. From such consideration, the transient hypofrontality theory of ASC can be formulated, which attempts to unify all altered states into a single theoretical framework. Because the prefrontal cortex is the neural substrate of the topmost layers, any change to conscious experience should affect, first and foremost, this structure followed by a progressive shutdown of brain areas that contribute more basic cognitive functions. Put another way, the highest layers of consciousness are most susceptible to change when brain activity changes. It follows from this ‘onionpeeling’ principle of sorts that higher cognitive processes such as working memory, sustained and directed attention, and temporal integration are compromised first when consciousness changes. Anecdotal evidence, particularly for drug-induced altered states, supports this idea. All altered states share phenomenological characteristics whose proper functions are regulated by the prefrontal cortex, such as time distortions, disinhibition from social norms, or a change in focused attention. The evidence is particularly strong from psychopharmacology. For animal and human subjects, psychoactive drugs have been regularly shown to negatively affect exactly these mental processes – working memory, executive attention, and the ability to estimate the passage of time. This suggests that the neural mechanism common to all altered states, those induced by psychoactive drugs included, is the transient downregulation of modules in the prefrontal cortex. The reduction of specific contents to consciousness is known as phenomenological subtraction. The deeper an altered state becomes, induced by the progressive downregulation of prefrontal regions, the more of those subtractions occur and people experience an ever greater departure from their normal phenomenology. In altered states that are characterized by severe prefrontal hypoactivity – various drug states such as those induced by LSD
Figure 1 The arachnologist Peter Witt wondered what
would happen to the ‘mind’ of spiders on psychoactive drugs. The results? See for yourself. It makes you reconsider coffee consumption. Witt and Rovner, 1982, reproduced with permission from Palgrave Macmillan Limited.
or PCP, for instance – this change results in an extraordinarily bizarre phenomenology such as hallucinations and delusions, most prominently. In altered states that are characterized by less prefrontal hypoactivity, such as long-distance running, the modification to consciousness is much more subtle. However, drug-induced states also differ from other altered states, such as meditation or hypnosis, in one important aspect. They add something to experience. So, on top of phenomenological subtraction, which is a hallmark of all ASC, drug states are also marked by phenomenological addition. The specific addition depends highly on the drug itself and the kind of neurotransmitter and mechanism of action it works on. In any event, the individual simply functions on the highest layer of phenomenological consciousness that remains fully functional. A consequence of the transient hypofrontality theory is that full-fledged consciousness is the result of a fully operational brain. Thus, and despite popular belief to the contrary, default consciousness is the highest possible manifestation of consciousness, and all altered states represent, by virtue of being an alteration to a fully functional
Psychoactive Drugs and Alterations to Consciousness
brain, a reduction in consciousness. ASC that are often presumed to be as higher forms of consciousness, such as, for instance, transcendental meditation or the experiences reported after taking ‘mind-expanding’ drugs are therefore really lower states of consciousness, as they all reduce cognitive processes – attention, working memory, etc. – that are associated with the highest forms of consciousness. This view is also in contrast to the theories of James and Tart, who maintained that normal consciousness is not qualitatively different from any other state of consciousness. It is difficult to imagine how higher consciousness might look like in terms of brain activity or feel like in terms of phenomenology, but should not it entail an enhancement of mental abilities ascribed to the prefrontal cortex rather than, as is the case in the above examples, their subtraction? If all drug states share this common neural mechanism, why, then, does each feel different? Specific drugs cause unique phenomenological changes that can be readily and reliably distinguished by drug users – or rats, for that matter – from those caused by some other drug. How can we reconcile this with the proposal that prefrontal downregulation is the underlying cause for all drug states, and indeed for all altered states? One clue is the specific phenomenological additions that come with each drug. Another might be that different drugs target slightly different sets of prefrontal modules, so removing quite specific computation from the conscious experience. Irrespective of these considerations, explanations of the neural basis of drug-induced altered states have focused, in the past, almost exclusively on neurochemical mechanisms. The transient hypofrontality theory makes clear though that a broader framework from cognitive neuroscience can provide a much better understanding of these phenomena of consciousness.
Drug Classification Psychoactive drugs are compounds that act on the central nervous system and alter behavior and cognition. They are highly lipophilic and thus readily cross the blood–brain barrier that keeps neurons protected from circulating blood. Psychoactive
221
drugs alter synaptic transmission by modulating neurotransmitter amounts and availability or by affecting receptor activity. In addition to their primary effects on mental processes – arousal, perception, mood,cognition,consciousness–thesedrugsproduce a variety of nonbehavioral effects that are often far more dangerous to health, and, in some instances, can be lethal. There are several different classification schemes for psychoactive substances – according to legal, medical, or pharmacological criteria – but the most common organization is based on their effect on behavior. This scheme classifies these drugs into four broad classes. The first are the sedatives and hypnotics. Drugs in this class depress or inhibit brain activity and produce drowsiness, sedation, or sleep; relieve anxiety; and lower inhibition. Although depressants do not share a common neural mechanism, most of them either decrease the metabolic activity in the brain or increase the transmission of the principal inhibitory neurotransmitter, gamma-aminobutyric acid (GABA). All sedative compounds have the potential for addiction and dependency. Common examples include barbiturates, such as Seconal; benzodiazepines, also known as minor tranquilizers, such as Xanax or Valium; nonbarbiturate sedatives, such as methaqualone; nonbenzodiazepines, such as buspirone; antihistamines and anesthetics; and alcohol. In low doses, alcohol can act as a stimulant, but with increased dosage its main effects are almost always depressive. Marijuana, which is derived from the hemp plant Cannabis sativa, is often misleadingly classified in the category of psychedelic. Given that its most pronounced behavioral symptom is sedation and the fact that it rarely produces – and only in very high doses – sensory distortions of hallucinatory quality, it is also best categorized here as a sedative. The second class are stimulants. These drugs produce behavioral and mental arousal. This class of psychoactive compounds also includes a variety of different substances each of which may also have a different neural mechanism. Common examples are amphetamines; cocaine; the methylxanthines, such as caffeine which is the most widely used psychoactive drug in the world, theophylline, which is present in tea, and theobromine, the psychoactive ingredient in chocolate; nicotine or
222
Psychoactive Drugs and Alterations to Consciousness
tobacco; appetite suppressants; and a variety of exotic plants, such as the betel nut, khat, yohimbe, and ephedra. Stimulants vary in strength, legal status, and the manner in which they are taken; however, all stimulants have addictive potential. None of these preparations, however, induces radical changes to mental status, such as, for instance, hallucinations or delusions, except perhaps cocaine and amphetamines when taken in high doses. The third class consists of the opiates. Drugs in this class act on opiate receptors and their two main effects are that they mediate relief from pain and produce feelings of euphoria, which is experienced in a dreamlike state of consciousness. It is also these two effects, of course, that make them drugs of abuse. In pharmacology, the term narcotics, derived from the Greek word for stupor, is used interchangeably with opiates but the legal systems in most countries use the term narcotics to refer to all illicit drugs, regardless of their mechanism of action. Opiates are highly addictive and can either be natural, semisynthetic, or synthetic. Natural opiates such as opium are derived from the opium poppy. The active ingredients of opium are morphine and codeine. The most famous semisynthetic opiate is heroin, which is two hundred times more potent than morphine. If taken intravenously, heroin produces what is known as a rush, an intense feeling of pleasure of euphoria that, according to first-hand reports, is not comparable in strength to any other experience. This makes heroin the most psychologically addictive drug known. Examples of synthetic opiates, also known as designer drugs, include methadone, naloxone, and the prescription pain medication Demerol, which is the drug most abused by physicians. The fourth and last class consists of the hallucinogens and psychedelics. Drugs in this category share the common feature that they induce hallucinations or have other ‘mind manifesting’ or ‘expanding’ properties. But this is also where the commonality ends. Pychedelics can either occur naturally in a plant, such as mescaline, which is derived from the peyote cactus, or be synthesized in the laboratory, such as LSD. In the latter case, such drugs are also known as designer drugs. In terms of consciousness – or behavior, for that matter – no overarching framework seems to be emerging that can organize their wide-ranging,
phantasmagoric effects. Pharmacologically, they are also all over the place. For every classical neurotransmitter system, there is a drug in this class. In consequence, psychedelics are probably best classified according to the neurotransmitter system they primarily affect. Cholinergic psychedelics include, most prominently, physiostigmine, scopolamine, and atropine, which are well known to cause bad trips. Drugs that alter norepinephrine transmission include mescaline and the popular ecstasy (MDMA), which cannot be said to cause bone fide hallucinations. Drugs that alter serotonin transmission include lyseric acid diethylamide or LSD, psilocybin, and morning glory. These drugs are perhaps most commonly associated with the psychedelic experience. There are also drugs with psychedelic properties that alter dopamine (cocaine), glutamate (ketamine), GABA (muscimol), opioid (morphine), or cannabinoid (hashish) transmission; however, these preparations are not predominantly hallucinogenic and they are best classified elsewhere according to their respective main effect. A peculiar subclass of the hallucinogens and psychedelics category are the psychedelic anesthetics, such as phencyclidine or PCP (angel dust), ketamine, and several gases and solvents, such as ether or nitrous oxide. With the exception of the cholinergic psychedelics, all drugs in this class have a high margin of safety, that is, a high TI and are generally nonlethal – even when taken in large quantities. In addition to these four broad categories, there a number of other drugs that affect the mind. These compounds are used to treat a variety psychological and neurological disorders and include antidepressants; antipsychotic medication, such as clozapine, chlorpromanzine, and haloperidol; and drugs for epilepsy, Parkinson’s disease, the dementias, such as Alzheimer’s disease, and spasticity. In most countries, the legal system takes a different approach to classifying psychoactive drugs. In the United States, for instance, the Comprehensive Drug Abuse and Prevention and Control Act of 1970 regulates these substances. According to this scheme, drugs are classified into five schedules according to the perceived risk of developing dependency to that drug. Schedule I drugs such as heroin, marijuana, and most psychedelics have a high risk of dependency and no accepted medical
Psychoactive Drugs and Alterations to Consciousness
use. These drugs are forbidden and cannot be obtained by prescription or, often, even for research, for that matter. Schedule II drugs, such as morphine, codeine, amphetamines, and certain barbiturates have a high risk of dependency but are accepted by the medical community for some treatment, that is, they have a medicinal purpose. Schedule III drugs have a risk of moderate physical dependency or high risk of psychological dependency and include preparations with limited opiates (e.g., morphine) and barbiturates not included in Schedule II. Schedule IV drugs, such as the benzodiazepines, have a slight risk of mild physical or psychological dependency. Schedule V drugs have less risk of mild physical or psychological dependency. Finally, alcohol and tobacco are not classified under this law. They fall under the jurisdiction of the Alcohol, Firearms, and Tobacco agency, which is a division of the US Department of Treasury. This is as good a place as any for a note of caution. Most explanations on the neural basis of altered states rely solely on neurochemical modulation. The reason for this is that drugs are the most common way to induce altered states and their known mechanism of action for changing brain function was for a long time the only hope to find a sound, mechanistic explanation for the mercurial nature of ASC. But there is a danger of leaning too hard on neurochemistry. An increase in serotonin does not cause happiness any more than an excess of dopamine causes paranoid schizophrenia. Neurotransmitters per se do not carry content in their messages. Drugs that work on the same neurotransmitter system, even via the same synaptic mechanism, can have very different effects on consciousness (Table 1). Prozac, for instance, increases serotonergic transmission and is a well-known antidepressant. LSD, on the other hand, is also a serotonergic agonist. LSD, however, sends you on a trip of some kind, but it certainly is not an antidepressant. Similarly, cocaine is a dopaminergic reuptake blocker and produces alertness and euphoria. Ritalin, the treatment of choice for children with attention deficit hyperactivity disorder (ADHD), works on exactly the same neural mechanism, blocking the reuptake of dopamine. Cocaine gets you high, though, to say nothing of prison, while Ritalin has a calming effect and the
223
medical community has approved it to be administered to kids. The moral here is this. Behavioral pharmacologists have long given up on the naive idea that complex psychological phenomena can be attributed to changes in the concentration of a simple little molecule. Every transmitter system can be modulated by drugs that do and do not alter consciousness. As such, we must also consider the function of the neural structure in which the changes in chemical neurotransmission occurs.
Phenomenology An obvious place to examine drug-induced alterations to phenomenology is the psychoactive substances classified under psychedelics and hallucinogens. Drugs in the other categories may produce changes to consciousness and its content as well but these are not anywhere as profound or interesting, at least for the study of consciousness, than drugs that induce perceptual changes, especially hallucinations, and changes to thought patterns, such as delusions. Note however that although not all psychedelics produce hallucinations or delusions, hallucinations have been studied most often by psychologists in this context. In consequence, this part of the article focuses mostly on druginduced sensory distortions, particularly in visual perception – color, shape, depth, etc. These events occur, however, in all sensory systems and can have, in all of them, hallucinatory strength. Delusions are less frequent and occur mostly with the ingestion of a few selective psychedelics. Sensory distortions, particularly bona fide hallucinations, do not typically occur spontaneously and are therefore usually not considered under normal experiences. In the Western world, such experiences are mostly associated with madness, as a symptom of a disease in the brain, such as psychosis, or malfunctioning of a bodily system, such as in sleep deprivation or starvation. When they occur in response to drug use, hallucinations are also typically seen as neurological junk or signs of temporary insanity. Not all cultures, however, view hallucinations in this negative light. Some cultures highly value hallucination, including druginduced hallucinations, and consider them sources of insight and truth. Great spiritual significance is
A sampler of psychoactive drugs detailing some of the more interesting effects on consciousness Drug and origin
Mechanism of action
Phenomenology
Sedatives and hypnotics
Valium
Binds to benzodiazepine site of the GABAa receptor facilitating the effects of GABA
Alcohol
Binds to barbiturate site of the GABAa receptor facilitating the effects of GABA Receptor agonist on the CB1 receptor; also inhibits 5-HT3 receptors
Produces all degrees of behavioral depression from sedation, to relaxation, to coma; reduces anxiety, disinhibition, sleepiness, tiredness, sense of well-being Relief of anxiety, sedation, disinhibition. Stimulant in low doses. Impairments in vision, speech, motor control, and judgment Main effects are sedation and analgesia; others include heightened sensations, uncontrollable laughing, sense of well-being, intensified introspection, mental slowing Alertness, arousal, insomnia, loss of appetite, and combats fatigue; known as ‘speed’ when taken IV or as ‘speedball’ in combo with opioids (to reduce side effects) Same as for amphetamines; more euphoria and paranoia; freebasing and crack result from purification procedures and intensify the experience, especially the euphoria Analgesia, peacefulness, feelings of warmth and well-being, boundless energy, dreamy imagery; when taken IV, heroin induces a ‘rush’ of intense pleasure that is without equal Restlessness, delirium, vivid hallucinations, disconnection with reality, no memory of the experience; OBEs are common; bad trips can easily be filled with ugly monsters Vivid hallucinations consisting of bright colors, geometric designs, and animals; synesthesia; no OBE, insight is retained; more color and less form distortion than LSD Induces feelings of empathy and sympathy; creates desire for intimacy and need for personal contact, heightens self-awareness; no hallucinations or loss of reality Vivid, kaleidoscopic sensory changes (e.g. mosaic patterns on all surfaces), powerful feeling of love, mystical oneness, synesthesia; bad trips include anxiety, panic, violence Pleasant feelings of relaxation, distortions of space, visual hallucinations with more intense color than LSD; effects resemble more mescaline than LSD; fewer bad trips Short-acting and intense psychedelic rush (businessmen’s LSD); loss of all reality, overwhelming visual hallucinations; communication with Gods, spirits, or the deceased Anesthesia with depersonalization, loss of ego boundaries, changes in body image, floating, OBE, NDE, distortion of time and space, full amnesia, euphoria; frequent bad trips
Marijuana (Cannabis sativa) Stimulants
Amphetamine Cocaine (Coca plant)
Stimulates the release of DA and blocks its reuptake; also activates adrenergic receptors DA reuptake blocker
Opiates
Morphine, heroin (opium poppy)
Activates opioid receptors
Hallucinogens and psychedelics
Scopolamine, atropine (Nightshade)
Cholinergic antagonist
Mescaline (Peyote cactus)
In a class of its own; structurally related to NE and alters NE transmission
MDMA (Ecstasy)
Pharmacologically promiscuous but primarily stimulates the release and blocks the reuptake of 5-HT2 Activates 5-HT2a receptors
LSD Psilocybin, psilocin (Mushrooms)
Structure resembles 5-HT and activates 5HT receptors
DMT (Virola tree)
5-HT agonist
PCP (Angel dust), ketamine
Primarily NMDA receptor agonists but also alter opioid and monoamine transmission
Dietrich A, Introduction to Consciousness, 2007, Palgrave Macmillan, reproduced with permission of Palgrave Macmillan.
Psychoactive Drugs and Alterations to Consciousness
Category
224
Table 1
Psychoactive Drugs and Alterations to Consciousness
often attached to them and people interpret them as visions that allow them to see into other realms of reality, realms inhabited by Gods and their ancestors. Because of this sacred meaning, hallucinations are sought out in elaborate rituals and only selected members of the group, such as shamans, are entrusted to have them. These rituals, in addition to drugs, often involve repetitive and rhythmic motion – spinning, whirling, dancing, and jumping – which is coordinatedwith some sort of cadenceddrumming, but hallucinations can also be induced through such behavioral techniques alone – without the aid of a psychedelic substance. Drug states are not so much an altered state as they are a whole family of altered states. Depending on the nature of the compound, people can experience many different kinds of hallucinations, delusions, and emotional changes. On the positive end, hallucinations may include intense colors, kaleidoscopic visions, fantastic images of animals and landscapes, mosaic patterns on all surfaces, as well as vibrating, rotating, or exploding designs that retreat into infinity. This is especially the case for the three major psychedelics: LSD, psilocybin, and mescaline. The delusions may include merging with one’s surroundings, such as a mystical feeling of oneness with God or the forces of nature; however, they may also be less sacrosanct experiences, such as becoming one with the ashtray on the table. The emotional changes may include pleasant relaxation, mild amusement, or even sweeping euphoria. On the negative end, changes may range from some tenseness or thoughts of paranoia to nail-baiting fear and terror. Drug users may encounter menacing beasts that chase them mercilessly with such a demented roar that would make the average nightmare look quaint in comparison. Other grotesque delusions may contain distortions to the sense of self, such as macabre changes in body image – the nose grows to double its size, for instance – bizarre depersonalizations where the psychonaut, a term Siegel introduced, floats in a void, and out-of-body or near-death experiences (OBE and NDE). As with most drug effects, whether you go to heaven or hell depends heavily on set and setting, that is, where and when you take the drug and in what state of consciousness or mood you happen to be in. Given the noncholinergic psychedelics’ high margin of
225
safety, the psychological scars that may result from the so-called bad trips may represent the biggest danger to health while visiting these alternate realities. Some psychedelic drugs, even in low doses, cut all ties with reality and take the user into an altogether different realm of experiences. PCP and scopolamine delusions are good examples. The danger of losing all understanding of what is real and what is imagined is that PCP-powered phantasms become ontologically fully real and when psychonauts, in their desperation, run from them and jump out of a real window, they are really dead, in this world. Other psychedelic drugs keep the psychonaut on this side of reality and maintain some type of lucidity as into what is real and what is imagined. Mescaline, of peyote fame, ecstasy, and several other norepinephrine psychedelics are drugs that tend to preserve such a tentative hold on reality. This often allows users to retain the insight and thoughts they had during the trip. In addition, for some drugs the changes to mental status are so subtle that they allow users, within limits and at low doses, to temporarily snap out of it. The anecdotal and esoteric descriptions by drug users, even the above broad overview, often give the mistaken impression that psychopharmacological karma comes in an infinite variety. Different drugs certainly evoke different cognitive and emotional effects, to the extent that they can, as said, be readily distinguished by drug users, even if they happen to be laboratory rats, but all still share common characteristics. The passage of time, for instance, loses all its meaning and people show no evidence of cognitive flexibility, that is, they simply lose any ability to extract themselves from their immediate surroundings – the freeze-frame of the here and now. Psychoactive substances, then, can be said to induce a sort of mental singularity, which, again, strongly questions the idea that these drugs are somehow mind-expanding, which is the original meaning of the term psychedelic, or induce a higher state of consciousness. Another common feature is the loss of ego boundaries, that is, people feel that they dissolve into the Universal Ocean, merge with the Almighty, or become one with Void. Such experiences are clear indications for the disengagement of cognitive functions supported by the prefrontal
226
Psychoactive Drugs and Alterations to Consciousness
cortex, as it is there that we form a sense of the self and delineate it from other selves. But taken together, drug states are also quite different from nondrug states – meditation, hypnosis, or rhythmic movement – because they are, in addition to phenomenological subtractions, also marked by phenomenological additions, depending on the systems the drug stimulates. Siegel has investigated the psychedelic experience in detail, collecting and categorizing the phantasmagoric images of the hallucinating mind in an effort to find patterns in what seems to be a phenomenon that can seemingly take on an infinite number of forms. He trained the participants of his experiments on a standardized hallucination code that classified their perceptions in precise language. They would report, for instance, that they see, say, 550 nm rather than a dirty red. His studies revealed that there is a hidden order in the chaos. The brain, it seems, hallucinates in only four basic geometric forms. All hallucinations have some uniqueness to them, some superimposed fiddly bits that provide some creativity, but all are variations on only four recurrent themes. These form constants, as they are called, were first discovered by Heinrich Klu¨ver, who experimented with mescaline in the 1920s. Klu¨ver called them the spiral, the cobweb, the tunnel or cone, and the lattice or honeycomb. Siegel, however, has found the same form constants in hallucinations induced by nondrug altered states, such as migraines, fever, auras, temporal lobe epilepsy, or sensory deprivation. They can be clearly seen in psychedelicinspired art all around the globe, be they from Huichol Indian paintings, or mandelas, or tie-dye T-shirts. The reason for these form constants can be found in the functional architecture of the visual cortex. Computer models using a columnar organization similar to that of the cortex show that the same form constants also crystallize in computer simulations (Figure 1). When the dosage is increased, the distorted sensory experiences become more involved and might contain people, storylines, and composite scenes. Siegel has also cartographed and indexed this more complex phase in the hallucinatory process and discovered that here, too, are underlying rules that govern how, exactly, consciousness is altered. There are, for instance, the effects of
megalopsy and micropsy, phenomena in which people or objects grow huge or shrink infinitesimally in size. There is also duplication and multiplication. A tunnel always leads to more tunnels and one sunflower becomes an entire field of sunflowers. The way in which motion and metamorphosis occur also has order to it. Birds become bats (never the other way around) and things first pulsate before they revolve. The hallucinating brain is not an unconstrained image generator. The best way perhaps to demonstrate this mental order is in visual perception. Siegel and Jarvik showed that specific drugs produce precise alterations in the color spectrum and color intensity and these changes that can be described in terms of dose–response curves. Figure 2 shows that the three major psychedelics LSD, psilocybin, and mescaline tend to be very colorful and bring out red, orange, and yellow hues, while marijuana is more on the bluish end of the spectrum. When psychonauts closed their eyes and used nothing but their imagination, blacks, whites, and violets dominated. A few psychedelic drugs are particularly strange, even for psychedelics, in the way they affect phenomenal consciousness. One is PCP, also known as angel dust. This substance is not popular anymore but in the early 1970s it was for a brief amount of time the drug of choice. The mechanism of action of PCP is not fully understood. It is a promiscuous compound that binds to different types of postsynaptic receptors and thus alters several different neurotransmitter systems at once. PCP is a so-called dissociate anesthetic because it produces, in addition to the usual delusions and hallucinations, two prominent effects: analgesia, the suppression of pain sensations, and an eerie feeling of detachment or dissociation from the world. The phenomenal experience is often extraordinarily bizarre with sensations of depersonalization, complete loss of ego boundaries and self-identity, OBE or NDE, as well as peculiar distortions in body image and selfconcept. Abuse of this drug frequently led to symptoms of paranoid schizophrenia requiring treatment and sometimes even hospitalization. Another peculiar compound is the designer drug ecstasy. Like PCP, it is pharmacologically promiscuous and, although classified as a norepinephrine psychedelic, it acts mostly in serotonergic synapses. This, however, as so often, does not explain its weird
Psychoactive Drugs and Alterations to Consciousness
227
80 Mescaline (300 mg)
Relative percent
70 60
LSD (100 μg)
50
Psilocybin (20 mg)
40 Δ9-THC (20 mg)
30 20
d-Amphetamine (15 μg)
10 Phenobarbitol (60 mg) BOL (100 μg)
V
BI G Co Y lor rep O or t R s W
Dr
B
ug
s
0
Placebo Baseline
Figure 2 Color in hallucinations is a function of the drug and its dosage. Participants saw warm reds, oranges, and
yellows on LSD, psilocybin, and mescaline; somber blues on marijuana and uninspiring blacks, whites, and shades of gray on amphetamine and sedatives. Reproduced from Siegel RK and Jarvik ME (1975) Drug induced hallucinations in animals and man. In: Siegel RK and West LJ (eds.) Hallucinations: Behavior, Experience, and Theory, pp. 81–161. New York: Wiley, with permission from Wiley (New York).
effects on phenomenal consciousness. Much disagreement surrounds ecstasy among psychologists and neuropharmcologists. For one, a controversy in the late 1990s about the neurotoxic properties of ecstasy, a controversy that included, most unfortunately, forged data, has not helped in our understanding whether or not this drug causes permanent brain damage. As for changes to the content of consciousness, ecstasy does not induce true, frank hallucinations. Rather, it heightens the sense of introspection in general and in particular seems to inspire intense feelings of empathy and sympathy. The user feels an increased need and even longing for personal contact and intimate tactile stimulation. This set of properties has led to the popular label of ecstasy as a love drug. In psychopharmacology, the proper term for this is entactogen. This led to two more controversies surrounding ecstasy. The first is the use of ecstasy as a date rape drug. The second is the proposal by some clinical psychologists to use ecstasy’s feeling enhancing effects in psychotherapy. Because the user maintains a high level of cognitive functioning and retains any insights gained during the state of heightened introspection, ecstasy, goes
the argument, might facilitate the self-discovery process. Needless to say, this proposal has been met by fierce opposition among psychologists and physicians, to say nothing of law enforcement. Marijuana is a different matter still. Marijuana’s mechanism of action was not understood until the mid-1990s when the endocannabinoid system was fully identified. This system is a transmitter system consisting of two receptors, CB1 and CB2, and their endogenous ligands, anandamide and 2-arachidonylglycerol (2-AG), which are the endocannabinoids, that is, endogenous cannabinoids. The fact that marijuana does not alter any of the classical neurotransmitter systems explains perhaps the distinct ASC it engenders. Delta-(9)tetrahydrocannabinol, or THC for short, is the psychoactive constituent of marijuana. It exhibits high affinity to the CB1 receptor and activation of this receptor complex, by endocannabinoids or by the ingestion of exogenous cannabinoids such as THC, induces emotional and cognitive changes such as analgesia, sedation, anxiolysis, and a sense of well-being. Other prominent effects include a shortened attention span, impaired working
228
Psychoactive Drugs and Alterations to Consciousness
memory, distortions of time estimation, all mediated by hypometabolism in prefrontal cortex regions, as well as enhanced sensory perception and a state of silent introspection. As said, unless high doses are taken, marijuana induces neither hallucinations nor delusions.
Summary Empirical progress on understanding the nature of drug-induced altered states has been impeded by the lack of an operational definition for these experiences as well as the ephemeral nature of drug effects. Because no single definition is widely accepted, other approaches to the problem, such as the transient hypofrontality theory, have been suggested. Such frameworks, which are based on cognitive neuroscience, raise the hope that we someday might better understand these alterations to consciousness by, perhaps, linking the phenomenology of drug states, or any other altered states, to specific neural markers. These models have already cleared the ground of common misconceptions such as that these states are mind expanding or that they constitute some higher form of consciousness that allows people to connect to the Great Beyond. We humans are curious about what it would be like to feel different. This is surely one of the main reasons that attract so many people to experiment with altered states, particularly drug states. It is also perhaps the main reason why drugs have featured so prominently in most societies and cultures throughout history. Drug states come in such a great variety and form that they seem to defy classification. There exist several different categories of psychoactive substances and each contains a host of structurally very different compounds that work via separate mechanisms of action involving all known neurotransmitter systems. What’s more, there is the bewilderingly complex phenomenology that appears to be highly idiosyncratic. There is a pattern underneath this diversity, though. The brain hallucinates in only a very limited number of stable configurations – the four form constants – that give
way to more intricate delusions and hallucinations also according to a number of seemingly simple rules. Colors, for instance, vary in an orderly fashion as a function of the type of drug and its dosage. All this suggests an underlying blueprint, a kind of universal hallucination code. Moreover, drug states share common themes, such as the loss of time perception, the lack of cognitive flexibility, and the disintegration of the self concept, which suggests that the one common neural mechanism underlying these altered states is the downregulation of the prefrontal cortex. See also: Altered and Exceptional States of Consciousness; General Anesthesia; The Neurochemistry of Consciousness.
Suggested Readings Bressloff PC, Cowan JD, Golubitsky M, Thomas JP, and Wiener MC (2002) What geometric hallucinations tell us about the visual cortex. Neural Computation 14: 473–491. Diaz J (2007) How Drugs Influence Behavior: A Neurobehavioral Approach. Upper Saddle River, NJ: Prentice Hall. Dietrich A (2003) Functional neuroanatomy of altered states of consciousness. The transient hypofrontality hypothesis. Consciousness and Cognition 12: 231–256. Dietrich A (2007) Introduction to Consciousness. London: Macmillan. Grinspoon L and Bakalar BB (1997) Marijuana, the Forbidden Medicine, 2nd edn. New York: Yale University press. Grinspoon L and Bakalar BB (1997) Psychedelic Drugs Reconsidered, 3rd edn. New York: The Lindesmith Center. Hardman JG and Limbird LE (eds.) (2000) Goodman & Gilman’s the Pharmacological Basis of Therapeutics, 9th edn. Elmsford, NY: McGraw Hill. Hobson JA (2001) The Dream Drugstore. Cambridge: MIT Press. James W (1890) Principles of Psychology. New York: Holt. Siegel RK (1989) Intoxication: Life in the Pursuit of Artificial Paradise. New York: Penguin. Siegel RK and Jarvik ME (1975) Drug induced hallucinations in animals and man. In: Siegel RK and West LJ (eds.) Hallucinations: Behavior, Experience, and Theory, pp. 81–161. New York: Wiley. Tart CT (1972) States of consciousness and state-specific sciences. Science 176: 1203–1210. Tart CT (1975) States of Consciousness. New York: Dutton & Co.
Psychoactive Drugs and Alterations to Consciousness
229
Biographical Sketch
Arne Dietrich is an associate professor of psychology and the chair of the Department of Social and Behavioral Sciences at the American University of Beirut, Lebanon. He holds a PhD in cognitive neuroscience from the University of Georgia. Professor Dietrich has done research on the higher cognitive functions supported by the prefrontal cortex, focusing mostly on the neural mechanisms of (1) altered states of consciousness (ASC), (2) the psychological effects of physical exercise, and (3) creativity. His major publications include a new theory of ASC, the transient hypofrontality theory, published in Consciousness and Cognition in 2003 and 2004, the proposal of two new mechanistic explanations for the effects of exercise on brain function, the endocannabinoid theory and the transient hypofrontality theory, and a new overall framework for the neural basis of creativity, published in Psychonomic Bulletin & Review in 2004. Professor Dietrich is the author of an introductory textbook on consciousness, published by Palgrave Macmillan in 2007, and his research has been featured prominently in the international press.
Psychodynamic Theories of the Unconscious M Macmillan, University of Melbourne, Melbourne, VIC, Australia ã 2009 Elsevier Inc. All rights reserved.
Glossary Coconscious, subliminal conscious, subconscious, and unconscious – Alternative terms used by the pioneers of depth psychology for describing unconscious mental processes. Instinctual drive – Freud’s concept of a drive having a physiological source that is represented in the mind by an idea which has psychological consequences. Libido – For Freud it is the psychological energy of the sexual instinctual drive; for Jung it is a general type of psychological energy that takes different forms. Psychodynamic theory – A theory that casts mental functioning as the result of conflicting mental forces. Repression – For Freud a mental force that prevented entry of unacceptable ideas to consciousness or the ego. Structure – One of the three agencies of id, ego, and superego making up Freud’s final theory, known as the structural theory, and which supplanted the three systems of the topographic theory. System – One of the three components, Cons., Precons., and Uncons., that comprised Freud’s early topographic theory of the mental apparatus.
The hypothesis [of unconscious cerebration] rests entirely upon the testimony of consciousness, and this testimony should be considered very suspicious. (Alfred Binet, Alterations of Personality, 1892/1896, p. 355).
Introduction Although the impression is sometimes given that Sigmund Freud discovered unconscious mental
processes, this view is incorrect. Freud made no such discovery. What he did was to invent a psychodynamic concept of unconscious mental processes and locate them in a particular repository, the unconscious mind. There is now more than one such conceptualization, but each derives from Freud’s. It is now difficult to recall that in the late 1800s, when Freud invented his concept, a whole new kind of psychology called depth psychology had come into being. Depth psychology was not a monolithic psychology, but rather a loose movement, grouped around the central proposition that aspects of mental life relevant to understanding unusual mental states were unconscious. The theoretical and practical contributions to this view made by Jean-Marin Charcot, Pierre Janet, Alfred Binet, and Hippolyte Bernheim in France, Joseph Delboeuf in Belgium, Charles Samuel Myers and Frederic William Henry Myers in England, and William James, Morton Prince, and Boris Sidis in the United States, to name only a few of its pioneers, were regarded as being at least as important as Freud’s. The depth psychologies differed in three fundamental ways. First, they conceptualized nonconscious mentation very differently. For some, these nonconscious mental processes were like their conscious counterparts, or even superior to them, but for others they were inferior. Second, and related to the first difference, there was profound disagreement over what was hidden in the depths. Was it another kind of self, an essentially normal but subconscious personality, or was it material disavowed by normal consciousness organized according to principles very different from those of waking life? Lastly, what caused the mentation to be so lost? Had a passive process caused mental fragmentation or did it result from something active? To these questions we must add that of asking how each theorist explained how we become aware of its workings.
231
232
Psychodynamic Theories of the Unconscious
Consciousness and Unconsciousness Although precise definition may elude us, most of us agree about the everyday use of the terms conscious and unconscious. Consciousness refers to what is currently in our minds or, more correctly, that of which we are presently aware. Our awareness of what occurred 20 min ago is different, but not markedly so, and can be retrieved with little effort. It is the mark of consciousness that it is there, and immediately so. Unconsciousness is quite different. Not being able to recognize it directly in ourselves, we do so only by being told by another about something that occurred when we were asleep or unconscious in some other way. As a pendant to our not being able to know about our unconsciousness, we usually have no difficulty in recognizing that state in others. Lancelot Whyte showed in the early 1950s that there had been an extensive debate about unconscious mental processes in the European philosophical literature from the second quarter of the 1700s to 1880. One of the main pre-1890s philosophical issues was whether any meaning at all was attached to the term unconscious idea. Some argued that because an idea was not immediately present in consciousness it could not be an idea at all. To explain the recall of an idea experienced 20 min ago, it was assumed that its physiological basis had been revived strongly enough for it to cross the threshold of consciousness. What seemed to be an unconscious idea was merely a physiological process with a potential to become an idea – as a potential it was simply not an idea at all.
Sensory-Motor Physiology and Psychology Throughout the second half of the 1800s, the dominant model underlying the workings of the nervous system and the psychological processes they supported was that of the sensory-motor reflex. Based on a generalization from the simple reflex arc, impulses were pictured as entering the nervous system through the sense organs, being transmitted to the brain and emerging from it as commands to move particular muscles in particular
ways. Incoming stimuli and motor responses were associated with one another because they had affective or emotional consequences. In some mid-century writings the important consequence was the gaining of pleasure, but in Alexander Bain’s physiological psychology of the 1850s it was the avoidance of pain or unpleasure. Thinking itself was modeled on associationism. One idea gave rise to another because both had previously been experienced at the same time or because of some similarity between them. What was true for a single association was also true for complex chains: one moved from idea A to idea Z because experience or similarity provided the intermediate pathways A ! B ! C ! D. . .W ! X ! Y ! Z. A pathway like this was usually conscious, but the notion of a threshold allowed for some degree of unconsciousness, such as when links C ! D. . .W ! X failed to be conscious. These links were present physiologically, but at an intensity that could not carry them across the threshold of psychological awareness or consciousness.
Freud and Associationism Of the versions of this kind of thinking that influenced Freud, the most important were those formulated in the period 1860–90 by Theodor Meynert and Eduard von Hartmann. Meynert’s was a physiological thesis in which associations were made by fibers in the white matter that connected the areas in the cortex where sensations were registered. In his famous example, fibers connected the auditory image of the sound of a lamb’s bleating with the visual image of its appearance such that the visual image was revived when the bleat was heard. The causal presence of the real lamb was an inductive or logical inference from its sound. Associations, singly or in trains, were thus identical with logical and causal connections. Once the starting image or idea was revived, nothing could derail the train from reaching the image or idea at its destination. Von Hartmann proposed that trains of thought were ordinarily evoked by a consciously willed purpose or motive. Were the conscious purpose dropped or abandoned, an unconscious one took over and directed the thinking to its goal.
Psychodynamic Theories of the Unconscious
We know that Freud was quite familiar with Meynert’s ideas. Although he came to disagree with his mentor about the localization of brain function, he consciously adopted Meynert’s view that trains of association were physiologically and internally determined, and even used the example of the sound and sight of the lamb. Von Hartmann’s books on unconscious mentation circulated very widely among the German reading intelligentsia. Freud had read them as a student although he claimed it was not until about 1914 that his attention was drawn to von Hartmann’s thesis about the unconscious guidance of thinking.
Unconscious Ideas and Symptoms None of the debate Whyte documented was about the unconscious production of symptoms or about their removal. Some fifteen years later, from about the mid-1960s, Henri Ellenberger located clinical concepts of unconscious mental processes in a late 1700s European development in what became known as psychiatry. The trends identified by Whyte and Ellenberger came to fruition in the work of many people by 1890, Freud among them. The notion that some symptoms, which we would now class as psychiatric conditions resulting from mental processes, is an old one, going back in Europe to at least 300 BCE, to those Greek temples devoted to the healing cult of Aesculapius and to at least 600 CE in healing practices like those of the Islamic Sufis. Central to practices like these was a well-defined sleep ritual from which the patient woke cured. Where there are reasonably detailed descriptions of the symptoms, or realistic representations of them in the votive offerings of grateful patients, it is almost certain that most of the disorders so treated were produced by unconscious mental processes and were hysterical. Here we need to consider very carefully what is meant by the term hysteria, especially as the concept is now often regarded as ‘incorrect’ in some politically irrelevant sense. Briefly, hysterical symptoms are a class of organic-like symptoms in which there is a loss of function, but which lack the organic basis that explains them. They include paralyses, contractures of the limbs, anesthesia,
233
inability to speak or understand language, loss of sight or hearing, and sometimes convulsions. Hysteria in this sense has nothing to do with the extravagant displays of emotion of popular language and is not related to gender. It is equally important to recognize that its symptoms are not feigned and that patients are not malingering. Charcot It was through the experimental work of JeanMartin Charcot, the great nineteeth century neurologist and student of hysteria, and the astute analyses of hysterical symptoms by Pierre Janet, his psychologist colleague, that ideas were positively identified as determining the symptoms of hysteria. Charcot observed that the symptoms of hysteria corresponded in every detail with those that could be produced by direct verbal suggestion under hypnosis. The characteristics of muscular paralyses and associated loss of feeling generated by hypnotic suggestion, ‘You are unable to move your arm,’ for example, exactly matched those of hysteria. There was a similar correspondence between hysterical symptoms that followed real physical trauma, such as falling on one’s shoulder, and those caused by indirect suggestion when a hypnotized subject was suddenly hit on the shoulder with moderate force. Charcot proposed that both physical trauma and verbal suggestion in hypnosis caused the ego to lose control over the realization of the ideas called up by the sensations. Here he applied and extended the concept of unconscious mental processing first proposed as part of sensory-motor physiology and psychology by Laycock in the 1840s and by Bain and Carpenter in the 1850s. Sensory-motor physiology had it that a mover could make a movement only if the idea of it was in the mind beforehand. Realization was the largely unconscious process that enabled its execution. Charcot incorporated the idea of a loss of movement and a loss of sensation into this schema. The patient’s real fall and the blow to the hypnotized subject called up sensations of paralysis and anesthesia in exactly the same way as a blow causing a ‘corked’ or ‘Charley horse’ arm. The ego would ordinarily have prevented that idea being realized, but after trauma and in hypnosis its control was
234
Psychodynamic Theories of the Unconscious
weak or absent so that the unconscious idea of loss was realized as a symptom. The basis of the symptom was thus an idea, or as it was sometimes called, ‘the imagination.’ Several times Charcot made the point that this did not mean the patient was malingering: although paralysis might result from an idea or the imagination, the paralysis was not imaginary. Janet Janet’s contribution was to notice that in hysteria and hypnosis it was the idea of the function or organ that was lost rather than those functions dependent on the innervation of particular muscles and receptors. Hysterical paralysis or anesthesia of the arm affected the everyday concept of the arm, that is, everything from the shoulder girdle down to the wrist. Janet explained this observation – the validity of which is quite independent of any of Charcot’s observations – by assuming that the idea of the arm had become dissociated from normal consciousness. The sensations making up the idea of the organ or function had not been synthesized into a unitary perception or assimilated into the subject’s idea of his or her self. Only with this act of what Janet called personal perception could the subject say ‘I feel.’ Dissociation followed a restriction of the field of consciousness during a traumatic experience, or during hypnosis, or when there had been prolonged inattention to the sensations. Breuer and Freud Freud adopted Janet’s concept of the role of everyday ideas in determining the details of symptoms (initially giving him full credit for it and later claiming it as his own). He and Breuer based their explanation of symptom formation on the principle that the nervous system acted to reduce excitation impinging upon it by directing it into movement, including the movements of speech. An idea was always associated with a charge or quota of excitatory energy, later to be called cathexis. When this quantity was especially large and ordinary channels of discharge unavailable, it was converted into muscular or sensory symptoms. Normal discharge could not take place in hypnotic-like states (Breuer) or might not do so if intentionally
prevented (Freud). The idea or function was then lost because its excessive affect prevented it from forming associations with other ideas. Freud By 1900 Freud had abandoned the theories of Breuer and the French and set out a different theory of how mental systems and the mental forces housed in them caused symptoms. The systems comprised Conscious (Cons.), Preconscious (Precons.), and Unconscious (Uncons.) minds. A mental force located (vaguely) in Cons. or Precons. opposed unacceptable ideas from entering Cons. and banished them to and kept them in Uncons. The ideas were motivated by drives and once in Uncons. were governed by a primary process: drives brooked no delay in their gratification, drive energy could readily be displaced from one idea to another or combined with it, and there was no sense of time or order and therefore no logic. On the other hand the systems Cons. and Precons. operated according to a rational and logical secondary process, oriented toward reality. Freud derived the notion of a force in Cons. and Precons. from the failure of his patients to follow the trains of their associations to the memories of the events he believed had caused their symptoms. The failure occurred at precisely the point where Freud thought the associations were closest to the causal memory, and it was there that patients displayed the most resistance to further recollection. It was also the point where Freud had to exert the most pressure or force to overcome the patients’ failure to recall. Repression was, Freud concluded, a force located within the patient’s self or ego that corresponded to the resistance. From the three systems and their mode of operation, Freud derived explanations for the formation of symptoms, dreams, and faulty actions (parapraxes), like slips of the tongue and pen, and forgetting proper names. All these mental products were compromises between the efforts of repressive forces seeking to keep unacceptable ideas from Cons. and forces in Uncons. just as determined to push their way in. Freud’s systems and the forces within them make up what is usually referred to as the topographic theory. Many of the functions carried out by Cons. and Precons. are similar to those in the theories of his
Psychodynamic Theories of the Unconscious
rivals. What makes Freud’s distinctly different is his picture of mental life as resulting from conflict between mental forces. It is that which constitutes his theory as the first psychodynamic theory.
235
‘‘the context calls for them inexorably.’’ Others again said they knew and did not know at the same time. These different reactions make the nature of the unconsciousness of Freud’s psychodynamic theory difficult to understand.
Clinical Unconsciousness All five, Charcot, Janet, Delboeuf, Breuer, and Freud, agreed that an unconscious memory was responsible for symptom formation, but they obviously differed about the mechanism. Charcot and Janet both hypothesized that lost memory was in a dissociated or subconscious part of the mind, and Delboeuf would probably have located it in a sleep-like state that alternated with normal consciousness. Breuer suggested that symptoms formed in a hypnotic-like or hypnoid state, an explanation he formulated only after Janet’s and Delboeuf ’s cases were published. It is almost certain that his thinking was influenced by Freud’s knowledge of the role Charcot and Janet attributed to ideas and to their notion of subconscious processes. Freud himself initially agreed with Charcot, Janet, and Breuer about the roles of hypnoid states and subconscious ideas. Gradually he came to think that it was always a psychological force that banished ideas, which conflicted with the individual’s standards, into the unconscious mind. Nevertheless, exactly like his contemporaries, Freud also had to postulate a constitutional tendency to form a secondary personality as a necessary basis for symptom formation. It is also a postulate common to the concepts of co-consciousness or subliminal consciousness, formulated by Morton Prince and F.W.H. Myers, which are not considered here. All share the same logical difficulties. For a set of ideas to be cut off from, but able to intrude into normal consciousness, disconnection and connection had to be present simultaneously. More, as Binet pointed out, if the secondary state was a rudimentary secondary personality, its characteristics could not be explained by the associationist psychology on which all drew. There was a particular problem with Freud’s view of the unconscious as a repository of unacceptable ideas. Some patients appeared to recall and experience as real his reconstruction of what he said had happened to them; others had no memory at all, agreeing, he said, only because
Unconscious Ideas and Therapy From about the mid-1850s unconscious mental processes were sometimes seen as being responsible for the removal of symptoms. Ellenberger described a whole class of what he called magnetic diseases in which this occurred. These very rare conditions were characterized by marked alterations of consciousness and included lethargy, catalepsy, somnambulism, and multiple personality. In those affected, mental processes were very different from normal, and resembled those produced by hypnotism, still then mainly known as animal magnetism. Although rare, magnetic diseases were known well enough through the popularization of authors like Euge`ne Scribe, whose play was utilized by Bellini in his opera La sonnambula, and E.T.A. Hoffmann in his stories – although not in those utilized by Offenbach in his opera. They were frequently cured by the patient leading the person conducting the treatment to a particular form of therapy in which the patient set the conditions for cure, often on a date they specified. Direct suggestion under hypnosis was often added to the therapeutic mix. Janet’s and Delboeuf’s Therapies Once the role of ideas in determining hysterical symptoms was clarified further, more rational therapies based on modifying the ideas underlying the symptoms were worked out. Charcot made no direct contribution to this development because he was decidedly skeptical about the therapeutic value of hypnosis, using it only to demonstrate experimentally how ideas could produce and remove symptoms. The first of the more rational therapies appears to have been that used by Janet in late 1888 or early 1889. He treated the 19-year-old Marie for recurrent hysterical crises accompanied by deliria, hallucinations, and violent bodily contortions. The attacks began two days after the onset of each of
236
Psychodynamic Theories of the Unconscious
her menstrual periods and menstruation was suppressed during this time. Janet hypnotized her and found her first menstruation to have been an entirely unexpected event to which she reacted with shame. She made an attempt to stop the menstrual flow by immersing herself in cold water. Menstruation ceased, but she then had a severe attack of shivering followed by several days of delirium. Menstruation did not recur until five years later and, when it did, the symptoms came with it. Marie had minor hallucinatory attacks of terror, which were repetitions of feelings she had experienced after seeing an old woman fall down some stairs and kill herself. She also had a left-sided facial anesthesia and was blind in the left eye. Janet then used what he termed ‘the singular means’ of using hypnosis to revive her memory of ‘the initial conditions’ and convince her that the idea that her menstruation had been arrested was absurd. The symptoms disappeared completely. Similarly her hallucinations were removed once Janet modified the idea she had formed at the age of 16 years that the old lady had died. Her blindness and anesthesia were removed by modifying her memory of a time when she was 5 years old and had been forced to sleep next to a child with a left-sided impetigo. Josef Delboeuf, the Belgian philosopher and psychologist, made a similar application of the same therapeutic principle apparently independently of Janet. In early 1889 he published his successful cure of a woman whose repetitive hallucinatory visions reenacted her inability to come to the aid of her dying son. Delboeuf had his patient revive the memory of the traumatic event and then used direct hypnotic suggestion to ‘combat it’ while it was in ‘its state of rebirth.’ Both Delboeuf and Janet used hypnosis to return their patients to the state in which the trouble had first manifested itself; both claimed to have cured their patients by removing or modifying the causal memories and images. Breuer’s Therapy Breuer’s patient, the 22-year-old Bertha Pappenheim – pseudonymously Anna O. – helped Breuer develop a similar method. When Breuer first
examined Bertha in 1880 for a nervous cough he immediately diagnosed her as mentally ill. She frequently stopped in the middle of a sentence to pause before continuing, seemingly unaware of these absences (as Breuer came to call them). It also soon became apparent that hers was a case of dual personality in which the transition from one self to the other took place in a well-marked late afternoon state of auto-hypnosis. She had a number of specific and variable hysterical symptoms including frightening hallucinations, partial aphasia, visual and auditory disturbances, paralyses, and contractures. In the best tradition of magnetic illness, Bertha gradually guided Breuer into what she termed the talking cure. During the afternoon auto-hypnosis she would tell stories like those of Hans Christian Andersen, but with sad or tragic themes, which nevertheless lightened her mood when she regained normal consciousness. After some months of so talking, Bertha extended her guidance by telling Breuer about the circumstances under which she had acquired her symptoms. Again, she gradually led him to identify what he called the root of her illness in an event that had occurred while nursing her seriously ill father. Dozing, she had had a hallucination of a snake about to strike at him, but her arm had ‘gone to sleep’ and she had been unable to raise it to defend him. Nor could she call a warning in German; she could utter only an English prayer. Her hallucinations, her inability to speak in German while retaining her ability to speak and write in English, and her paralyses and contractures repeated her inability to act and speak at that time, and were elaborated into symptoms during her later absences. Breuer’s published account had it that each of her separate symptoms was removed after she told the stories of the events during which they had formed, and that she was finally relieved of them by reenacting, on a day she nominated as the day of her cure, the root cause of the snake hallucination. That, Breuer said, ended the whole of her hysteria. Nevertheless there were significant limitations to Bertha’s recovery: Some were reported by Carl Jung in 1923, others by Ernest Jones in his 1953 biography of Freud. Fuller details emerged in 1972 after Ellenberger found a copy of Breuer’s notes on Bertha. Albrecht Hirschmu¨ller’s 1976 complete
Psychodynamic Theories of the Unconscious
publication of the notes and Breuer’s related correspondence reveal that Bertha’s talking calmed her mood temporarily, but had little or no effect on her symptoms. In fact, at one stage while regularly using the method, her symptoms became so much worse that she was ‘forcibly’ hospitalized. Unlike the published version, Breuer’s notes state only that she had been greatly relieved by the treatment. Only 5 weeks after the treatment ended Bertha was admitted to a convalescent hospital, partly or mainly for the treatment of a morphine addiction caused by Breuer’s prescription of the drug for her facial pain, but some of the same hysterical symptoms, including the inability to speak and understand German, were still present. She remained there for four and a half months. Over the next 5 years she was hospitalized three more times for a total of 10 months, each time usually showing remnants of her hysterical aphasia and possibly other speech disturbances, absences, hallucinations, and convulsions, as well as evening facial pains, which because of their variability may have been part of the hysteria, even though storytelling now made them worse. She did recover, but what brought that about is a total mystery. Freud’s Therapy Despite the limitations of Breuer’s treatment, Freud began by basing his own therapy on a variant of it that became known as cathartic therapy. Whereas Breuer’s published and unpublished descriptions of Bertha’s treatment say she merely talked about her symptoms in much the same way as she told her stories, Freud’s variant had it that there was an accompanying discharge of the excess affect attached to those ideas. The recovery of the lost idea, and hence the effectiveness of the treatment, was because of that discharge or abreaction. Freud began to use this technique between 1885 and 1895 at a time when, in the light of what he had learned from Charcot and Janet, he seems to have pondered over what Josef Breuer had told him about Bertha. Freud’s first use of a purely cathartic method in the waking state was probably toward the end of 1892 with Lucy R. She was governess to the Viennese family of a widower with two children, and
237
troubled by a recurrent smell of burnt pudding – an olfactory hallucination. She easily recalled a recent event after which the symptom formed. Her thinking about leaving her employment had been interrupted by a game with the children during which a pudding had burned. Lucy R.’s presenting symptom was its smell. On Freud’s further inquiry she fairly easily recalled a much earlier event during which she felt she was falling in love with her employer. Freud found it difficult to accept these as the memories responsible. By this time he conceptualized sets of memories being linked in a pathogenic memory structure in which at least one memory had to have been deliberately repressed. Neither of Lucy R.’s had been, and at Freud’s insistence, she recalled an event in which her employer had criticized her severely for allowing a female guest to kiss the children. Realizing that this meant there was no possibility of marriage, she had therefore intentionally pushed the thought out of consciousness. With a little further prompting, Lucy R. then recalled a fourth event in which a male guest had been criticized for kissing the children. Cigarsmoke had been present at the time and even though she had not banished the memory, a transient minor olfactory symptom of the smell of cigar smoke had troubled her. At the center of the structure, Freud placed Lucy R.’s deliberately repressed memory of being criticized over the female guest. It formed an unconscious core or nucleus to which the memories of the feeling of love for her employer and the criticism of the male guest were drawn, even though neither of them had been repressed. During her thinking about leaving, which was also not repressed, the affect associated with all three memories was pooled, and reached sufficient intensity to be converted into maintaining the simultaneously present olfactory innervation caused by the burned pudding, the way for which had been paved by the cigar-smoke.
Sexuality and Repression Freud had originally formulated repression for explaining hysteria, but soon generalized its scope to what he called psychoneuroses. In obsessional
238
Psychodynamic Theories of the Unconscious
neurosis, for example, the affect was displaced from the unacceptable idea to a neutral one. He made this and similar generalizations in a context partly determined by the sexual theory of nasal neurosis, formulated by his Berlin friend and colleague, Wilhelm Fliess. Freud set out to establish an exclusively sexual etiology for the actual neuroses of neurasthenia and anxiety neurosis. In doing so, he applied Koch’s postulates, used for identifying the causes of bacterial diseases, in such an incomplete and methodologically indefensible way that his conclusion that they had sexual causes was unwarranted. All it did was to generate a false expectation that he would find sexual causation in the psychoneuroses when he later investigated them.
Perverse Sexuality, Determining Quality, and Free Association In Charcot’s patients and experimental subjects, the sensations experienced in the trauma were repeated in the symptoms. Freud called this property determining quality and adopted it as the criterion for identifying the traumatic memories at the core of the pathogenic memory structure. Symptoms containing oral, anal, and genital sensations, therefore, had to be based on memories of frankly sexual experiences in which those zones of the body were stimulated. Given that Freud wanted to establish sexual causes for the psychoneuroses, it is not surprising to find him placing enormous pressure on his patients to recall events possessing those kinds of sexual content. The main outcome of Freud’s pressure was the seduction theory, or more correctly, the seduction hypothesis. He came to believe that brutal perverse sexual assaults involving those zones had been perpetrated on his patients during childhood, usually by an adult, and sometimes by their fathers. He therefore expected his patients to have memories of them. Some did so, but for the most part the causal ‘memories’ were reconstructions he made from the very fragmentary recollections of his resisting patients. For those patients unable to recall sexual traumas of any kind, only the inexorable demands of the context forced them to agree with Freud. In arriving at the seduction hypothesis Freud discounted his influencing the patients’ recollections.
To understand this we need to consider the assumptions underlying the method by which he gathered his data, that of free association. When patients thought in a nonpurposive way about their symptoms Freud required them to report with absolute honesty everything that came to mind, including thoughts that intruded, however irrelevant. Freud’s application of what he termed, ‘the fundamental rule of psychoanalysis’ depended on trains of thought being psychologically continuous – gaps in them being filled by psychological and not physiological processes – and of them being guided by unconscious ideas. During free association, all the ideas that were produced, including those that intruded, were under the guidance of unconscious ideas related to the symptom. In one of his earliest implicit references to the rule, Freud said it was demonstrably untrue that trains of thought were ever purposeless; when conscious purposive ideas were abandoned, unconscious ones simply took their place. This was the thesis proposed explicitly by Meynert and von Hartmann and it gave no place to unconscious suggestion. Not long after leaving the Salpeˆtrie`re, Freud had vigorously defended Charcot against the charge that many of the phenomena of hysteria and hypnosis produced there were products of unconscious suggestions transmitted to the patients and subjects. In believing that his patients’ trains of thought were similarly uninfluenced, Freud was wrong. External factors like unconscious suggestion could not be disregarded: the determinants of what was produced were not completely inside the patient. In addition to many of the seduction ‘memories’ being reconstructions, it is sometimes clear from Freud’s reports that he imposed them on the patients. This is already evident in the pre-seduction hypothesis case of Elizabeth von R., on whom he practically forced the idea that her painful leg pain was the result of her repressing the idea that she loved her brother-in-law. The later post-seduction hypothesis patient, known as the Wolf Man, resisted until the end of his life, Freud’s reconstruction of the root cause of his problems as his witnessing his parents’ coitus a tergo when he was only 18 months old. What actually happened in these and other cases is difficult to establish. In the case of the Rat Man, for example, on whom some of Freud’s case notes have survived, the discrepancies between the
Psychodynamic Theories of the Unconscious
239
notes and what Freud published make it impossible to know anything reliable about his reconstruction of the supposed threat made by the Rat Man’s father.
drive: repression became a two-stage process, and the unconscious became more than a repository of unwanted thoughts.
A Sexual Instinctual Drive
Two-Stage Repression
The seduction hypothesis required that the traumatic childhood experiences at the nucleus of the neuroses be of perverse sexuality. At the beginning of 1898, when Freud repudiated the reality of these ‘memories’ because he no longer believed in them, he was left without a source for the sexual sensations in the symptoms. His solution was to propose a childhood sexual instinctual drive with component drives that generated the perverse sensations in the oral, anal, and genital zones of the patient’s own body, that is, autoerotically, and which gave rise to fantasies of seduction. The only direct observational evidence Freud had for any component was for an oral drive and it was an unwarranted sexual interpretation of the ‘sensual sucking,’ described by the Hungarian pediatrician Samuel Lindner. In his study of mainly thumb and finger sucking in 500 children, Lindner described four ‘exultant’ suckers who went into a ‘rapture.’ Freud claimed that Lindner recognized pleasure-sucking in general and the rapture specifically as sexual, even though Lindner made no such claim and his descriptions do not warrant that interpretation. Once proposed, Freud’s hypothetical childhood sexual drive helped explain what he saw as the singularity of repression in being directed only at the memory of sexual events. He supposed that in remote prehistory the stimulation of all three zones produced sexual pleasure. During evolution, when humans adopted the upright stance, mental forces of disgust, shame, and morality emerged and reversed the affect. Stimulation of the zones now generated repellent sensations and the affect that was repression. Because Freud believed that individual development recapitulated that of the species, the same mental forces automatically caused repression when the zones were stimulated during childhood.
In Freud’s new conception, primary repression was the first of the two stages and took place when the component drives lying dormant in the phylogenetically abandoned but still-sensitive erotogenic zones were stimulated and released unpleasure. The other was secondary repression and took place much later when and if direct or indirect stimulation of those zones again caused unpleasure. As well as explaining why repression acted only on sexual memories, yesteryear’s revived phylogenetic memories provided a nucleus around which today’s repressed ontogenetic experiences could be grouped. Of each kind of repression the question may be asked: What is it? Obviously Freud tells us what each does, but he leaves us ignorant of how each produces its effects. Both repressions are therefore uncharacterized theoretical terms, lacking properties of their own. The defect is already found in Freud’s earlier and simpler concept: How is an idea stripped of its affect? Converted into muscular innervation? Displaced on to another idea? Similarly, what brings about a reversal of affect? Another and seemingly fatal objection to both the old and the new concepts is that an idea is lost to consciousness when stripped of its affect but finding it and abreacting it requires the affect still to be present. How can that be? Similarly, the adult perverse activity Freud allowed for, only took place if the sexual instinctual drive was strong enough to override the unpleasure, especially when the ‘average uncultivated woman’ was led to it. What conditions allow overriding? How is the strength of the overriding force measured independently of its effect?
Sexuality and the Unconscious There were two main theoretical consequences of Freud’s introducing a childhood sexual instinctual
Drives and Mental Structures Postulating a sexual drive in childhood required Freud to propose a second drive, one strong enough to effect the repression of sexuality at a time when the ego and its standards were necessarily weak. Freud had always based his explanations on the conflict between two mental forces as,
240
Psychodynamic Theories of the Unconscious
for example, in the opposition of a conscious system to an unconscious one, and it is not surprising that he now chose an ego-instinctual drive to give the ego the energy to control the sexual drive. The ego-instinctual drive also accounted for the development of a sense of reality, one that found the real objects that satisfied the child’s real needs rather than its hallucinated images of them. By 1923, Freud had drastically modified even this duality and replaced it with another. The ego and sexual drives were fused into a life-instinctual drive, termed Eros, which was challenged by the destructive force of a death-instinctual drive, termed Thanatos. Freud advanced two kinds of arguments for this conceptualization. First, he reinterpreted a number of behaviors as seeking unpleasure rather than avoiding it. Thus, whereas he had originally seen repetitions of unpleasurable traumatic dreams as attempts to master the traumas, he now said they were instances of a repetition compulsion. Similarly, when patients transferred on to the analyst their childhood feelings toward their parents, they were repeating painful experiences that had never brought satisfaction. Transference was now under the death instinct and very far from an attempt to avoid unpleasure. To these examples Freud added the repetition of what he took to be an unpleasant experience in his grandson’s play. His ‘maturer reflection’ now allowed him to conclude that all these behaviors were motivated by the death instinct. Freud also advanced a complex speculative argument that he believed supported the existence of a biologically based instinctual drive that unconsciously sought the individual’s death. Structures and Unconsciousness These theoretical reconsiderations had consequences for the topographic theory. Without Uncons. where were the new instincts housed? Freud again reinterpreted an early observation, this time on resistance, and drew on an apparently new one. The reinterpretation was that the resistance to recovering associations was unconscious. Paradoxically, although Cons. and Precons. were the organs of consciousness and repression, they were now required to have unconscious components. Freud’s apparently new observation was of his patients getting worse precisely after he praised them for
improving. Calling the deterioration a ‘negative therapeutic reaction,’ Freud now attributed it to an unconscious sense of guilt motivated by Thanatos. Even though he had previously said that the sense of guilt was conscious, even ‘overstrongly’ so in conditions such as depression, he now said it was unconscious. Unconsciousness and the Structural Theory Clearly neither Thanatos nor Eros could be easily housed in Cons., Precons., or Uncons. At any level of consciousness their manifestations would have had to be logical and rational, and it would be just as difficult to conceptualize an irrational and illogical system providing organized opposition to them. Freud’s solution was to propose three new mental structures – the ego, superego, and id – and house derivatives of the new instinctual drives in them irrespective of whether the ideas they motivated were conscious or unconscious. The ego drew on a neutral form of energy derived from Eros and had the responsibility for cognitive functions, such as perception, attention, memory, motor control, and for judgments of reality. It was the only structure in which consciousness arose and in which anxiety could be experienced. The superego formed out of the ego and was powered by parts of Thanatos. It contained the individual’s moral standards, a conscience that scrutinized behavior against them, and an agency that punished infringements of them. The id was the original energy store. It contained a fused form of Eros and Thanatos and was governed by the primary process. The assumption of fusion was, Freud said, ‘‘indispensable to our conception;’’ only by combining with Eros could the self-destructive aim of Thanatos be deflected from the individual. Fusion, in turn, required the further assumption of defusion, one that ‘‘forces itself upon us.’’ Only defused drives could provide energy appropriate to the structures. Some degree of unconsciousness attended each of the new structures: the id wholly so, but the ego and superego functions most closely connected with it were also partly unconscious. Freud now decided that theoretical emendation was necessary: the conscious or unconscious status of an
Psychodynamic Theories of the Unconscious
idea did not depend on its belonging to systems like Cons., Precons., or Uncons. Granting he had ‘managed pretty well’ with the ambiguity of both repressed and preconscious ideas being ‘unconscious,’ he now restricted the terms conscious, preconscious, and unconscious to the descriptive. The structural theory, as it became known, thus completely supplanted the topographic theory with its three systems and instinctual driving force.
The Oedipus Complex and the Structural Theory During the formation and resolution, a universally present Oedipus complex changes in instinctual drives gave each of the new structures the energy appropriate to it. One part of the complex was based on the child’s nonsexual primary identification with its parents: the child wanted to be like them. The other part built on the child’s sexual object choice : it wanted to have the parents as sexual objects. This object choice was made only after Eros could be directed from the autoerotic zones on to real objects. The theoretical representation of the complex was, Freud said, ‘‘much more difficult’’ than its description. There is a similar theoretical complexity about what happened next. Primary identification was a change in the child’s ego such that it resembled its parents. In so doing it became an object of Eros, and the change sublimated the energy of Eros into a nonsexual form that then powered the ego’s nonsexual and cognitive functions. At the same time the two parental sexual objects were incorporated into the child’s ego where their capacity to satisfy Eros made them targets of Thanatos. There were thus positive and negative complexes. In the positive, Eros was directed at the mother-object and Thanatos toward the father; in the negative form the development was reversed. Castration Anxiety A universal threat of castration now caused the development of moral standards, a conscience, and a punitive function. Even when not made aggressively, the threat of castration was always experienced against the trace of a phylogenetic memory of a real but historically remote castration.
241
Anxiety generated by the threat caused the forcible withdrawal of the sexual drive from its objects, and all the parental identifications, together with their functions, to be taken into the ego. Thanatos turned inwards as self-hate and powered the superego that formed as a precipitate of the incorporated parental functions. The identifications so taken over were then united ‘‘in some way,’’ said Freud, to confront the rest of the ego with Thanatos. This remnant ego of the structural theory had lost the capacity to repress, which it had had in the topographic theory. In exchange, it gained the ability to experience anxiety. When anxiety signaled an internal threat, it avoided the danger by bringing the superego’s repressive forces into action and escaped the external threat through its cognitive and motor functions. Although, Freud said this ego was now only ‘‘a poor creature’’ menaced by dangers arising from the id and superego within and from the external world, it could still defend against both.
Freud’s and Other Psychodynamic Theories We have seen that Freud first thought of unconscious mental processes in a way not very different from his contemporaries. The systems of Freud’s first psychodynamic theory comprising an unconscious mind driven by sexuality, and operating with an illogical primary process that a secondary conscious process attempted to control, was completely different. The structural theory with its three agencies of id, ego, and superego was even more so. Freud’s topographic theory also differed markedly from the two main psychodynamic theories which followed more or less immediately from his. Alfred Adler’s was the first. He rejected instinctual drives as important determinants of behavior, and with that the concepts of repression, an unconscious mind, and a universal Oedipus complex. He postulated an innate but nonbiologically based drive to mastery that drove development from the individual’s initial sense of inferiority toward a selfperfection that overcame it, the behavior of which the individual was unaware did exist, but it was not in the unconscious mind or caused by or related to repression. If an Oedipus complex developed, it was
242
Psychodynamic Theories of the Unconscious
because of overindulgence, usually by the mother, which prevented normal development. Carl Gustav Jung’s was the second. It was built around an instinctual drive possessing a nonsexual, general form of energy (confusingly called ‘libido’) located in a collective unconscious that also housed the primordial images or archetypes that regulated it. Repressed and merely forgotten material was located in a personal unconscious. Jung’s concept of libido had no autoerotic direction; the moral standards of the ego were innate and drew their energy from part of the nonsexual libido. Jung rejected Freud’s understanding of the Oedipus complex, especially of its universality, and the notion that the superego was descended from it. Adler’s and Jung’s theories have the same conceptual problem as Freud’s. How can theories based on a single motivating force be validated? Can a choice be made among them? Appeal cannot be made to direct observation because all are based on interpretations and reconstructions, which are neither simple, nor direct, nor rule governed. Consider the different interpretations of delusions of grandeur to which each of the three lead: Freud’s and Jung’s in the case of Schreber, and Adler’s in Nijinsky’s. Freud said Schreber’s delusions came about when his libido was withdrawn from its realworld objects and directed on to the ego. He thereupon lost his sense of reality and his ego became his whole world. Jung said that the whole of Schreber’s libidinal drive (as he conceived it) had withdrawn from reality, not just the form affecting his sexuality. Adler proposed Nijinsky’s drive had overcome his great sense of inferiority, and had taken him to a grandeur that denied him in real life. How could a choice be made? Can choice be made among modern psychodynamic theories of the unconscious? To the extent that they are based on similar kinds of evidence, interpretations, reconstructions, and theorizing as Freud’s, it cannot. As an illustration, consider the basic choice between Freud’s topographic and structural theories. Curiously, although Freud clearly believed that the structural theory had supplanted the topographic, many modern psychoanalytic theorists have made the opposite choice, and therefore of concepts of unconsciousness different from Freud’s. Several broad strands of thought, none of which are completely independent, can
be identified among those making this choice. There is a substantial group, among them David Rapaport and George Kline, who reject the structures completely on the grounds that they are too abstract. They revert to something like the topographic theory, including an unconscious as a system, because they say it gives a more meaningful representation of what happens in treatment. On the other hand, Joseph and Ann-Marie Sandler revised Uncons. precisely because it needed to be brought more into line with clinical experience. Some reject the structures completely or modify them significantly (Roy Schafer), or the functions Freud assigned them, as does Susan Millar in requiring the superego to love the ego, or like Heinz Kohut incorporate all the functions into a unitary self. Different concepts of unconscious necessarily follow from the strand that rejects some or all of the instinctual drives (Edward Bibring, Charles Brenner, and Robert Holt), especially of Thanatos (Ernest Jones and David Werman), or the vicissitudes of the instincts (Heinz Hartmann). A related rejection having profound effects on what is to be considered unconscious is of what Freud termed the shibboleth of psychoanalysis – the Oedipus complex – as a phenomenon (Bennett Simon and Jay Greenberg) because it develops quite differently and relatively late (Melanie Klein). How can a variation like this among psychoanalytic theories of unconsciousness be explained? If free association is not influenced by the analyst’s unconscious suggestion, as Freud insisted, each analyst should, broadly speaking, gather the same data. Occasionally one or another modern analyst grants or hints that this is not so ( Jacob Arlow and Charles Brenner, Jeremy Nahun), but none challenges the deterministic assumptions basic to the method. If the data are the same, variation in theory must then be due to their being interpreted differently. Studies of the interpretation of dreams and symptoms by different psychoanalysts clearly show there are great differences, and that is also true of interpretations of more general clinical phenomena such as transference. To the extent that reconstructions are complex interpretations, it is true of them also. Can differences like these be reconciled when there are no rules against which interpretations and reconstructions can be judged?
Psychodynamic Theories of the Unconscious
Conclusion Without rules for guiding interpretations and reconstructions every psychoanalyst can formulate his or her own psychodynamic theory. Following Freud’s example, many do so. Freud based many elements of his concepts of unconscious mental processes on such idiosyncrasies as inferring repression from his feeling of personal effort, arguing for a deathinstinct based repetition compulsion by simply reversing earlier interpretations and giving more mature thought to them, making indispensable assumptions the basis for instinctual fusion and defusion, and abandoning the topographic theory because of the need to do away with ambiguities with which he previously managed well enough. Feelings and interpretations as idiosyncratic as these are found in modern theorists. How then did Freud’s theory win out over its rivals? There is not much hard evidence, but several factors seem likely to have contributed. First, psychoanalytic concepts seem to provide unlimited explanatory scope and are not limited to pathological conditions. Second, they have what Wittgenstein termed ‘charm.’ Despite unconscious mentation allegedly being governed by an alien primary process, every unconscious motive, drive, or process is understandable in the same way as its directly experienced conscious counterpart. Third, explanations can be derived from them as readily as explanations of everyday behavior. Without specific training, everyone can apply them and be their own analyst in generating plausible interpretations of their own and their friends’ behavior. Social influences having little to do with validity, but much to do with social movement psychoanalysis cannot be overlooked. The movement came about largely through the proselytizing of A.A. Brill, Ernest Jones, and Freud’s controlling of a doctrine through a special committee drawn from among his closest disciples – one that Phyllis Grosskurth has called the ‘Secret Ring.’ No other depth psychology had that kind of founding impetus, centralized control, or appeal to everyday explanations to maintain that impetus. Recasting any of the psychodynamic theories of unconscious mental processes into other theoretical frameworks is possible, but can be useful only if
243
there is agreement on the validity of the particular theory being recast. Were those conditions met, a nice illustration of what might be achieved is provided by the radical minimalist connectionist model, Lucynet, implemented by Dan Lloyd, to simulate Freud’s case of Lucy R. Lloyd accepted Freud’s report at face value, a trust that might be misplaced given his other reports. Lucynet generated Lucy R.’s symptoms without a memory archive, an unconscious, or an explicit function for repression. However enthusiastically cognitive psychologists might greet such a recasting, it is difficult to see that many psychoanalysts would. Other attempts to reformulate other psychodynamic theories may be more difficult than Lloyd’s. Almost all their data and therefore the bases of their interpretations remain hidden under the privileged couch. Consequently the evidence to decide which is worthy of recasting derives from the testimonies of two consciousnesses: the introspecting patient’s and the interpreting psychoanalyst’s. May Alfred Binet’s warning of nearly 120 years ago be disregarded or is suspicion still justified? See also: Perception: Unconscious Influences on Perceptual Interpretation; Unconscious Cognition; Unconscious Goals and Motivation.
Suggested Readings Borch-Jacobsen M (1996) Remembering Anna O.: A Century of Mystification. In: Olson K, Callahan X, and Borch-Jacobsen M (trans.). New York: Routledge (Original work published 1995). Breuer J and Freud S (1895) Studies on hysteria. In: Strachey J (ed.) The Standard Edition of the Complete Psychological Works of Sigmund Freud, vol. 2, pp. 19–305. London: Hogarth. Cioffi F (1998) Freud and the Question of Pseudoscience. Chicago and La Salle: Open Court. Ellenberger HF (1970) The Discovery of the Unconscious. New York: Basic Books. Esterson A (2001) The mythologizing of psychoanalytic history: Deception and self deception in Freud’s accounts of the seduction theory episode. History of Psychiatry 12: 329–352. Freud S (1900) The interpretation of dreams. In: Strachey J (ed.) The Standard Edition of the Complete Psychological Works of Sigmund Freud, vols. 4 and 5, pp. 1–621. London: Hogarth. Freud S (1916–1917) Introductory lectures on psychoanalysis. In: Strachey J (ed.) The Standard Edition of the Complete Psychological Works of Sigmund Freud, vol. 15 and 16, pp. 9–496. London: Hogarth.
244
Psychodynamic Theories of the Unconscious
Freud S (1920) Beyond the pleasure principle. In: Strachey J (ed.) The Standard Edition of the Complete Psychological Works of Sigmund Freud, vol. 18, pp. 7–64. London: Hogarth. Freud S (1923) The ego and the id. In: Strachey J (ed.) The Standard Edition of the Complete Psychological Works of Sigmund Freud, vol. 19, pp. 12–59. London: Hogarth. Freud S (1933) New introductory lectures on psycho-analysis. In: Strachey J (ed.) The Standard Edition of the Complete Psychological Works of Sigmund Freud, vol. 22, pp. 5–185. London: Hogarth. Hirschmu¨ller A (1989) The Life and Work of Josef Breuer: Physiology and Psychoanalysis. New York: New York University Press. (Original work published 1978). Klein DB (1977) The Unconscious—Invention or Discovery?: A Historico-Critical Inquiry. Santa Monica, CA: Goodyear.
Lloyd D (1998) The fables of Lucy R.: Association and dissociation in neural networks. In: Stein D and Ludik J (eds.) Neural Networks and Psychopathology: Connectionist Models in Practice and Research, pp. 248–273. Cambridge: Cambridge University Press. Macmillan M (1997) Freud Evaluated: The Completed Arc. Cambridge, MA: MIT Press. Macmillan M (2003) Challenges to psychoanalytic methodology. In: Chung MC and Feltham C (eds.) Psychoanalytic Knowledge and the Nature of Mind, pp. 219–238. London: Palgrave. Macmillan M and Swales PJ (2003) Observations from the refuse-heap: Freud, Michelangelo’s Moses, and psychoanalysis. American Imago 60: 41–104. Timpanaro S (1976) The Freudian Slip: Psychoanalysis and Textual Criticism. In: Soper K (trans.). London: NLB (Original work published 1974). Whyte LL (1960) The Unconscious Before Freud. New York: Basic Books.
Biographical Sketch
Malcolm Macmillan graduated with DSc from Monash University (1992), MSc Melbourne University (1964), and BSc from the University of Western Australia (1950). Since 2005, he has been a fellow of the Academy of the Social Sciences in Australia, and during 2004–05 was the president of the International Society for the History of the Neurosciences. His visiting positions include fellowships at Oxford University and the University of Pittsburgh. Macmillan was a founding member, then fellow (1988), and life member (2005) of the Australian Psychological Society. He was also a founding member of the College of Clinical Psychologists of the APS. Since 1991 he has been a fellow of the Association for Psychological Science. In November 2006, he became one of the three coeditors of the Journal of the History of the Neurosciences. Macmillan is the author of the prize-winning An Odd Kind of Fame: Stories of Phineas Gage (MIT Press), and his Freud Evaluated: The Completed Arc (MIT Press) has had considerable critical success. He is also the author of some 90 papers, book chapters/ encyclopedia entries, and book reviews, and presenter of over 30 conference papers. Currently he is evaluating the work of Alfred Walter Campbell and comparing Campbell’s cytoarchitectonic research with Brodmann’s.
Psychopathology and Consciousness S A Spence, University of Sheffield, Sheffield, UK ã 2009 Elsevier Inc. All rights reserved.
Glossary Cognition – Strictly speaking, this means thinking (as in Descartes’: cogito ergo sum), but the word tends to be used to describe the processing of information by the brain: for example, attention, concentration, and different forms of memory. Compulsion – A behavior linked to an obsession, the submission to which the subject resists, the prelude to which is increasing anxiety and tension, and the performance of which may lead to temporary relief, for example, compulsive hand-washing. Delirium – A global impairment of cognitive function, usually acute in onset, fluctuating in intensity, and potentially treatable. Consciousness is characteristically impaired (‘clouded’). Delusion – An abnormal belief based upon abnormal inference, incorrigible, and out of keeping with the subject’s social, cultural, and religious background. Dementia – A global impairment of cognitive function, usually chronic in onset, progressive, and irreversible. Consciousness is initially preserved (‘clear’). Hallucination – A percept in the absence of a veridical external stimulus, over which the subject has no control. Hysterical conversion – An unconscious mental conflict is hypothesized to have been resolved through its conversion into a physical symptom, for example, paralysis, for which an explanatory physical cause cannot be found. Illusion – A misperception of a veridical external stimulus, more frequent in low stimulus intensity environments (e.g., darkness) or when the subject is distracted.
Imagery – Mental percepts (in the absence of external stimuli), over which the subject exerts control, and which they recognize as imagined (e.g., in the mind’s eye). Neurosis – A mental condition, the symptoms of which are exaggerations of those present in normal mental life, for example, anxiety. Obsession – A recurrent, intrusive thought, usually distressing in content, which the subject recognizes as his own, but which he tries to resist. The subject matter he rejects as being incongruent with his identity (ego-dystonic). Pseudodementia – An apparent global impairment of cognitive function, subacute in onset, during which the subject fails tasks out of an expressed belief that they cannot do them (they tend to ‘give in’ too soon); associated with depression or malingering. Psychosis – A mental condition in which the symptoms present are not held to be features of normal mental life, for example, delusions.
Introduction Psychopathology concerns the study of abnormal states of mind and their correlates in abnormal patterns of behavior. The term has been used in a purely descriptive sense in the context of phenomenological philosophy, as an aid to interpretation in psychotherapeutic approaches to the mind and, more recently, in an experimental context, most notably in the pursuit of the neural correlates and biological underpinnings of abnormal symptoms and syndromes. Psychopathology necessarily invokes the experience of a conscious subject and the attempts of an observer to understand and describe that experience
245
246
Psychopathology and Consciousness
as accurately as possible. Nevertheless, psychopathology is prone to the limitations associated with trying to determine what another person is thinking: the third-person ‘observer’ can never know what the first-person ‘subject’ is experiencing; much depends upon inference. This article describes the main symptoms of mental disorders with particular reference to first-person experience and third-person observation. We offer definitions of key psychiatric terms such as hallucination, delusion, delirium, and dementia. Where the biology of such experiences is understood this is touched upon. Throughout, emphasis is placed upon the accurate understanding of terminology, realistic appraisal of clinical contexts, and the limits upon what may be factually stated regarding another’s mental state. Finally, we consider interactions where communication may be deliberately subverted.
What Is Psychopathology? Psychopathology is the study of the experience and expression of abnormalities of mind. Hence, it inevitably involves a dynamic interaction between the person and their surroundings, the subject (or patient) and their witness (the observer). Psychopathology has been approached in three different ways over the last 150 years: 1. Descriptive psychopathology involves the careful description of mental states of closely observed human experiences, without recourse to explanation (this has been called ‘phenomenology’ and is exemplified by the classic contributions of Karl Jaspers). This form of psychopathology is a prerequisite to accurate diagnosis; it requires time spent listening to subjects; and, by definition, the mental phenomena examined are conscious: they comprise what the subject experiences (and can describe). The formal, structured description of mental symptoms probably began in the 1860s. 2. Interpretative psychopathology also involves the careful description and understanding of human subjective experience, but seeks to explain it mainly through recourse to
hypothesized unconscious mechanisms (this has been called psychodynamic psychopathology and is exemplified by the classic contributions of Sigmund Freud). Hence, much of what might be discussed in psychodynamic psychopathology concerns what the subject herself is unaware of (that which is ‘unconscious’ and initially unknown to her); part of the purpose of therapy is to understand recurring patterns of behavior that might, until then, have eluded recognition. 3. Experimental psychopathology has sought to understand subjective phenomena mainly in terms of brain processes; it currently constitutes the dominant paradigm and might subsume a host of investigators using a variety of physical methodologies to examine the biological processes supporting disordered mental life (e.g., Wilder Penfield’s pioneering use of direct electrical stimulation of the human cerebral cortex to elucidate the focal correlates of subjective experience (voices were heard by subjects when their temporal lobes were stimulated); or the discovery of the huntingtin gene by the Huntington’s Disease Collaborative Research Group in 1993, the latter indicative of a similarly mechanistic approach, albeit applied at a radically different level of explanation: genes rather than anatomically specified cerebral systems). Hence, contemporary psychiatric practice requires the ability to retain an understanding of, and relationship with, individual human subjects, to help them attain autonomy (i.e., to effect change in the real-world) while simultaneously incorporating into their clinical care knowledge gleaned from experiments conducted at many different levels of mechanistic understanding (from genes to communities). This may precipitate tensions and, indeed, within scientific academic psychiatry, there is currently a search for endophenotypes, brain markers that may be objectively diagnostic of specific neuropsychiatric disorders, which could effectively circumvent the need for prolonged, detailed description of individual mental states. These approaches raise ethical issues that have yet to be fully explored.
Psychopathology and Consciousness
What Is a Psychiatric Diagnosis? Most psychiatric conditions comprise syndromes, in other words they are characterized by groups of symptoms that tend to run together. Such syndromes are clearly described in operational terms in documents such as the Diagnostic and Statistical Manual (DSM) of the American Psychiatric Association and the International Classification of Diseases (ICD) of the World Health Organization (WHO). Hence, in schizophrenia (ICD 10, WHO, 1992, category F20) it would not be unusual to encounter a patient who believes that he is being spied upon, that others are reading his mind from a distance, and that there are persecutors who speak to each other about him, whom he can hear (‘‘he is bad, look at the way he holds his fork, he is evil’’). These phenomena (persecutory delusions, thought broadcast, and third-person auditory verbal hallucinations, respectively) would satisfy criteria for a diagnosis of schizophrenia, assuming that they had lasted for more than a month and that they occurred in the absence of any other processes (such as cannabis use, F12, or gross brain disease, F06) that might have provided plausible explanatory alternatives. It follows that most psychiatric diagnoses essentially describe subjective states. In schizophrenia (F20) there is often a theme of persecution, in dementia (F00-F03) the subject is losing awareness of themselves and their surroundings, in depression (F32) they are suffering despair, but in all these states the diagnosis is quintessentially concerned with current consciousness, with what it is like to be that person at that time. For the most part then psychiatric diagnoses differ from many other forms of medical diagnosis, where there are objective, biological pathologies that constitute the evidential basis for diagnosis, and of which the subject need not necessarily be aware: a woman with an ovarian tumor may not know that she has it until it causes pain, a man may have advanced coronary vascular disease long before he is aware of his condition. Hence, these pathologies, these diagnoses, may be ‘silent’ in a way that psychiatric disorders have not been said to be (to date). This is not to say that things might not change, but it does call into question what one might be treating in a psychiatric patient who did not have symptoms.
247
The treatment might be medical, but would it still be psychiatric? Of course, there are exceptions on either side of the dualistic (and outmoded) body/mind divide: several ‘physical’ diagnoses lack a consistent pathophysiology and essentially comprise descriptions of symptoms (e.g., irritable bowel syndrome, fibromyalgia, chronic fatigue syndrome) and it is noticeable that these too constitute descriptions of conscious states that could not be said to occur ‘silently’ (indeed, they also often cross over the hinterland, into psychiatry, when a sufferer is offered a psychotherapeutic intervention). Conversely, there may be ‘psychiatric’ disorders, that, while asymptomatic, might be said to be present (in some sense) before their subject knows of them (e.g., in the undiagnosed carrier of the huntingtin gene (F02.2), or the person developing Alzheimer’s disease pathology (F00) before their memory impairment is apparent). Nevertheless, a manifest psychiatric disorder seems to require a conscious subject for its expression: we do not currently speak of ‘silent schizophrenia.’
Dissecting Phenomenology Traditionally, psychopathological symptoms (experienced by the subject) and signs (witnessed by the observer) have been described, organized, and addressed according to defined phenomenological categories, for example, the perceptual disturbance of the person with schizophrenia (F20) who hears ‘voices’ contrasts with the memory disturbance of the person developing Alzheimer’s disease (F00). While such a distinction is feasible for certain, discrete diagnostic entities (e.g., Korsakoff ’s syndrome, F10.6, can occasionally comprise solely a disturbance of memory, thereby constituting a pure amnestic syndrome), most of the time the disorders patients present with affect more than one domain of mental life, for example, the man with schizophrenia may hear voices (a disturbance of perception), while believing that his thoughts are not his own (a disorder of thinking), and feeling ‘nothing’ (a disturbance of emotion). So, while it might be helpful to categorize disordered phenomena according to discrete phenomenological domains, most neuropsychiatric syndromes impact
248
Psychopathology and Consciousness
more than one domain; hence, a comprehensive account of their pathophysiology would necessarily be required to explain a diverse range of phenomena. Helpful Dichotomies Some traditional distinctions have gone in and out of fashion, but may still be useful when thinking about neuropsychiatric disorders. Indeed, they are still employed clinically. Psychosis versus neurosis : a distinction has been made between psychoses, conditions in which phenomena are experienced that are not present in the normal mental state, where a subject ‘loses contact with reality,’ and neuroses, which may be understood as incorporating exaggerations of ‘normal’ mental experience. A good example of the former is schizophrenia, the latter, generalized anxiety disorder. While a so-called neurotic patient may elicit considerable understanding from an interlocutor, because the experience they describe (e.g., anxiety) is not qualitatively different from normal states of consciousness, the situation for the patient experiencing a psychotic illness can be more troublesome: he may be trying to describe phenomena that are totally alien to his listener; indeed, the situation he finds himself in may well be beyond his powers of description (e.g., the man who feels thoughts entering his head through his skull and can point to their site of entry). The psychotic/ neurotic distinction provides a useful shorthand for describing patients’ situations, although it must be acknowledged that there are limits upon its universal applicability: some neurotic phenomena may be so extreme that few healthy people would really share an insight into what it might be like to experience them (imagine the thoughts of the woman suffering from obsessive–compulsive disorder, F42, who washes her hands throughout the night, with soap and bleach, until her skin splits and her fingers bleed, but she cannot stop because of her fear of contagion); conversely, although hearing voices has long constituted the layperson’s notion of the quintessential symptom of madness, more recent epidemiological research suggests that voices are heard by more people than see psychiatrists (or even tell their doctors about them); such phenomena may be common in the
context of bereavement, sensory deprivation, or significant trauma. They may also arise commonly in the context of falling asleep or waking up (socalled hypnagogic and hypnopompic hallucinations, respectively), when a subject might hear their name being called. Hence, the neurotic may deviate markedly from normality and those who regard themselves as ‘normal’ can experience something of what it might be like to be psychotic. Functional versus organic : Historically, it was always apparent that certain mental disorders were more obviously associated with gross (i.e., obvious) pathology of the central nervous system (CNS) than others. In the eras of Hughlings Jackson, Broca, Wernicke, Charcot, Janet, Freud, Kraepelin, Alzheimer, and many of the other great, early neurologists, neuropathologists, and protopsychiatrists, it was clear that certain mental disorders (such as general paralysis of the insane (GPI), a late complication of syphilitic infection) had visible correlates in the brains and bodies of those afflicted (in life; though confirmation ultimately arrived at postmortem), while other conditions did not seem to exhibit such correlates (e.g., dementia praecox, the diagnosis that eventually became known as schizophrenia). What emerged was an understanding that certain disorders involved gross organic (usually structural) disturbance of the brain, while others were thought to reflect more subtle functional impairments (which might be understood to be psychological or neurochemical in nature). Hence, schizophrenia (F20) and manic-depressive psychosis (which came to be known as bipolar disorder, F31) were regarded as functional psychoses in contrast to GPI or encephalitis lethargica, which could cause organic psychoses (F06). There are now a number of problems with this distinction. One is that, over time, organic correlates of functional brain disorders such as schizophrenia have become well recognized (e.g., subtle enlargements of the fluid-filled ventricular spaces within the brain, and atrophy of the frontal and medial temporal lobes). Another problem is that the functional/organic distinction has been borrowed to refer to an entirely different set of contrasting conditions. In this second scenario, the patient is one who exhibits unusual symptomatology, where there is an element of doubt concerning the veracity of his symptoms. In conversion
Psychopathology and Consciousness
hysteria (F44.4) a man might be unable to lift his arm, yet a biological cause cannot be found. The Freudian, interpretative psychopathological approach to this problem is to look for the causes of his symptom, specifically unconscious psychological causes: Is he unable to lift his arm because to do so would then allow him to strike his wife, whom he now hates on account of her infidelity? In this case, the man’s paralysis is functional (psychogenic: caused by his ‘mind’) not organic (i.e., not a consequence of a stroke or brain tumor, etc.). However, there is a further problem with this use of the word ‘functional’: in itself it does not distinguish the man with hysterical conversion (an ‘unconscious cause’ for his symptom, F44.4) from the man who is malingering (consciously producing his symptom for personal gain, F68.1). Hence, the term functional may slide between several different meanings: from subtle brain disturbance (in schizophrenia, F20), to hypothesized psychogenic (unconscious) mechanism (in hysteria, F44.4), to implied deceit (in malingering, F68.1). We need to be especially careful when applying this and certain other psychopathological terms in clinical practice (see Table 1). Visual versus auditory hallucinations: Similarly, a useful distinction has concerned the modality of hallucinations (perceptions experienced in the absence of a veridical external stimulus). While visual hallucinations are more often encountered in organic brain states, auditory hallucinations (particularly of voices) are more characteristic of the so-called functional disorders such as schizophrenia (F20) and severe (psychotic) depression (F32.3). A good example of an organic state would be that of delirium tremens (F10.4), seen among alcohol-dependent patients withdrawing from alcohol. The street drinker, hallucinating in the doorway, may be pointing and gesticulating at persecutors that only he can see. More frightening still may be the knives and spears he sees penetrating his body. Visual hallucinations should prompt consideration of a treatable, physical disease. The characteristic auditory hallucinations of schizophrenia and depression will be addressed below. However, we should note that though this distinction (between visual and auditory hallucinations) is very helpful clinically, it is still the case that organically ill patients may ‘hear things,’ and that
249
patients with schizophrenia sometimes ‘see things.’ It is the relative preponderance of these phenomena that is clinically informative. Delirium versus dementia: When faced with a patient who is cognitively impaired, for example, disorientated in time and place and unable to account for themselves, it is particularly important to quickly consider two things: the physical health of the patient (they need to be examined by a doctor) and any information that can be gleaned from their family, friends, contacts, or witnesses. When the patient cannot speak for themselves we are crucially reliant upon what witnesses can tell us about them. A global disturbance of cognitive function (meaning: concurrently impaired attention, concentration, memory, perception, language, and spatial awareness) may arise gradually over time or with a rapid, precipitate onset. The speed of onset and the stability of the findings are important for diagnosis, as is the level of consciousness: the alertness of the patient. In delirium, a physical illness (F05), such as a chest infection or withdrawal from alcohol (F10.4) precipitates the patient into an acute state of confusion, and his level of awareness may wax and wane over hours and minutes. He may come and go, get worse overnight, experience moments of apparent lucidity, and then deteriorate again. His sleep-wake cycle is much disturbed. He may be particularly prone to visual illusions (misinterpretations of veridical external stimuli): the coat on the back of the door becomes a demon; delirium is an intensely frightening experience. Hence, it may be defined as an acute or subacute onset of global cognitive impairment, in which the level of consciousness fluctuates, and where the cause is usually potentially treatable. Physical investigation and resuscitation is essential, as is a calming environment, a simple, well-lit room (to reduce the likelihood of visual illusions) and, ideally, the constant attention of the same (i.e., familiar) nursing staff. In contrast, while the dementias also impact global cognitive function, they tend to do so more slowly (over months and years) and, in general, the patient remains in clear consciousness while awake. In addition, while the person with delirium may lack insight into their current difficulties (and hence be at risk of acting on mistaken beliefs; e.g., the man who jumps from
250
Psychopathology and Consciousness
Table 1
Some psychopathological terms that have multiple or disputed meanings
Term
Meaning
Ego-syntonic Ego-dystonic Formal thought disorder
Consistent with the subject’s personality, their long-term values and behaviors. Inconsistent with the subject’s personality, their long-term values and behaviors. Indicative of schizophrenia but used in two ways. The first (more accurate application) is to refer to a disordered flow of thinking as expressed in disordered speech: speech that is incoherent and at its worst comprises ‘schizophasia’ or ‘word salad.’ The second use actually refers to the content of specific delusions seen in schizophrenia, those which relate to thought, such as thought insertion, thought withdrawal, thought broadcast. Has been variously applied to mean: serious disorders that lack a structural brain correlate but which are hypothesized to implicate biochemical or psychological etiologies (e.g., schizophrenia), or a disorder which is purely a product of unconscious mechanisms (as in hysterical conversion) or is malingered (i.e., consciously performed). The degree to which the subjects understand their situation: do they realize that they are ill; do they understand the nature of that illness; do they believe it requires treatment; what kinds of treatment would they accept as necessary? A contentious diagnosis, suggesting that the subject experiences and exhibits more than one personality, usually emerging in the course of interpretative psychotherapy. One ‘personality’ (the ‘host’) may have knowledge of all the others (the ‘alters’). Has been deployed as a forensic defense. A mental disorder where the symptom is an exaggeration of phenomena that might be experienced normally (e.g., anxiety) and where (traditionally) the subjects have not lost contact with reality, that is, they retain insight into their disorder. Usually implies a mental disorder with a demonstrable physical cause, which might be structural (such as a brain tumor), pathophysiological in the brain (as in epilepsy), the body (as in hypothyroidism), or exogenous (as in drug-induced psychosis). In the lay sense, it means feeling picked upon or persecuted, and this meaning is applied in the diagnosis of paranoid personality disorder. However, when applied to schizophrenia, paranoid means deluded, literally out of one’s mind. Hence, one may suffer from paranoid schizophrenia, and be deluded, but one’s delusions need not be persecutory (one might believe that one is Joseph Stalin or Charlie Parker). A confusing term, used in two ways. The first is a way of describing a percept occurring in inner space (i.e., not in the outside world), against the subject’s will, but lacking the reality of a ‘true hallucination’ (this form is characteristically seen in people with personality disorders, an inner voice, which the patient may recognize as their own, urging self-harm: ‘‘Jump in the river, jump in the river’’). A second, less satisfactory application is in the description of a full hallucination, into which the sufferer has gained insight, in other words they are experiencing a hallucination, but they realize it is a symptom of illness and not ‘real.’ To the public, a word often confused with psychotic (out of touch with reality), but really a term describing abnormal personalities. In the legal sense (in the United Kingdom) a psychopath has a severe personality disorder, one justifying involuntary incarceration in hospital; in diagnostic usage, a psychopath describes a particular personality disorder, an extreme variant of the antisocial: one associated with the instrumental manipulation of others, lying, and cheating, premeditated violence, lack of remorse and empathy for victims. According to this usage, a ‘successful psychopath’ might never come to medical/ legal attention. A mental disorder where the symptoms comprise phenomenology that would not be experienced in normal mental life and where the subject is regarded as having lost contact with reality, that is, they currently lack insight. Psychiatrists use this term in three ways: to describe the experience of others being aware of one’s thoughts, at a distance; the experience that one’s thoughts are literally broadcast out aloud and therefore audible to others close by; or the experience that others have access to one’s thoughts via telepathy. All uses are current! Commonly confused with multiple personality disorder (above). Schizophrenia actually refers to a ‘splitting of the mind,’ not between identities (persons), but between functions, so that thoughts and perceptions become disturbed and incoherent.
Functional
Insight Multiple personality disorder Neurosis Organic Paranoid
Pseudohallucination
Psychopath
Psychosis Thought broadcast
Schizophrenia
the roof because he thinks he can fly), the person with dementia may migrate from knowledge of their failing powers (early in the process) toward a lack of recognition of themselves or others
(toward the end); they are at risk of suicide while they know what is happening to them, at risk of exploitation, neglect, or misadventure when they do not.
Psychopathology and Consciousness
Form versus content : Descriptive psychopathology has borrowed this distinction from philosophy and it has considerable bearing upon clinical diagnosis. An experience occurs within a phenomenological domain (e.g., as a thought or percept) and this constitutes its form. The theme or subject matter of that experience constitutes its content. Hence, a man with schizophrenia (F20) may hear voices saying ‘‘he is evil, he is dirty’’; the form is a third-person auditory verbal hallucination, the content derogatory and persecutory. Contrast this case with that of another man, who suffers from obsessive–compulsive disorder (F42): he experiences recurrent, intrusive thoughts, which he knows to be his own, but which prey upon his mind: ‘‘I am evil, I am dirty.’’ Here, the form is an obsessive thought and the content is depressive. While an interpretative psychopathologist might pinpoint the similarity between these cases (i.e., their content is derogatory to the self, the patient is identified as evil), the descriptive psychopathologist will make a diagnostic distinction (the form of the auditory hallucinations is in keeping with schizophrenia, while that of the obsessive thought is consistent with obsessive–compulsive disorder; their treatments are different). Nevertheless, while matters of form are crucial to diagnosis and identifying the appropriate therapy, the question of content remains central to understanding the person, the one who experiences that symptom. In the above examples it would not be surprising if questions of evil or immorality were important to these patients when well, if they had featured in their moral development and earlier life. However, their specific disease processes influence the manifestation of their concerns in the present: as heard voices or intrusive thoughts, respectively. Reactive versus instrumental: The purview of psychopathology extends beyond the domains of mood and perception, and memory and concentration, to include accounts of human moral behavior, particularly those expressions of the personality that may cause suffering to patients or their society (F60–F69). When the cause of that suffering is violence then it may be helpful to consider whether the violent act arose spontaneously, in reaction to the trigger of an acute situation or whether, instead, it comprised the product of premeditation on the part of the perpetrator.
251
So, for instance, a man who is intoxicated with cocaine (F14.0) might feel acutely irritable and hyperaroused, might be overactive and lacking in foresight regarding the consequences of his impending actions; he might then punch a man who blocks his way while leaving a nightclub. Alternatively, a severely psychotic man, suffering from schizophrenia (F20), may come to believe that the devil is stalking him, so that he carries a knife for his protection, and when he is stopped by a stranger who asks him for directions he stabs the man, believing him to be the devil in disguise. Finally, a sadistic psychopathic pedophile (F65.4) may spend many months grooming a victim over the Internet, through an alias adopted while ‘online’ in so-called ‘chat rooms’; he may have repeatedly masturbated, thinking of what he will do to the victim when he meets her, then he acts out his plan with a precision bordering on the obsessional. This behavior is clearly ‘instrumental.’ Levels of Understanding A feature of the clinical psychopathological discourse is the tendency to arrange diagnostic entities hierarchically. By this we mean that because several neuropsychiatric syndromes may share symptoms and signs, items of psychopathology are rarely pathognomonic (solely diagnostic) of one single disorder and clinicians must arrange differential diagnoses according to an order of preference: 1. Partly on the basis of which single diagnosis might most comprehensively (and parsimoniously) account for all of the phenomena described (i.e., applying Ockham’s razor); 2. Partly on the basis of relative frequencies and likelihoods of one condition over another (Alzheimer’s disease, F00, is common, Huntington’s disease, F02.2, uncommon); but also 3. On the basis of a hierarchy of causation (Table 2). Hence, while a symptom such as anxiety might be entirely normal prior to exams or stage performances, it might constitute a disorder if it is extreme (persistent, disabling, interfering with vocational function); but although it may comprise both a symptom and a diagnosis (as in ‘generalized
252
Psychopathology and Consciousness
Table 2 The phenomenological state of anxiety may have many precursors or causes Level of explanation
Exemplars
Normal
Exam stress Public speaking or performance Generalized anxiety disorder Agoraphobia Social phobia Posttraumatic stress disorder Obsessive–compulsive disorder Agitated, major depression Paranoid schizophrenia Caffeine Alcohol (withdrawal from alcohol) Amphetamines Paroxetine (‘Seroxat’) Brain-related: Aura preceding the ictal phase of a temporal lobe seizure Systemic: Thyrotoxicosis Hypoglycemia Anxious, avoidant personality Dependent personality disorder
Neurosis
Psychosis Organic, exogenous
Organic, endogenous
Personality
anxiety disorder,’ F41.1), it can also be a comorbid feature of other diagnoses, to which it is contributory if not central. For instance, anxiety may accompany many of the neuroses, for example, agoraphobia, F40.0, and social phobia, F40.1, may arise as part of a depressive syndrome, F41.2, can accompany a psychotic illness such as schizophrenia, F20, but might also be a symptom of a bodily disorder such as thyrotoxicosis (hyperthyroidism) or pheochromocytoma (a tumor of the adrenal gland), F06.4 (Table 2). In all these cases the patient may experience and exhibit anxiety and, while there may be subtle differences in periodicity and duration (short bouts of panic associated with going out in agoraphobia; persistent dread in the severely depressed), the symptom is nevertheless common to all these syndromes. In eliciting a history and examining the patient, the psychiatrist is attempting to answer the following questions: Is this anxiety the totality of the condition (and if so what is its meaning to the patient, what makes it happen?) or is it part of something else? As will be seen in Table 2, as we move from the level of ‘normality’ through levels of increasingly organic pathology, we also move from what we might call the very psychological to the very physical. We move from the social environment
(the stage or examination room implicated in the healthy subject’s anxiety), through the exaggerated impact of social interactions (in social phobia and agoraphobia) toward physical environment (ingested as caffeine or amphetamines, F15), until we encounter effects endogenous to the brain and body of the sufferer (e.g., his endocrine system), and ultimately the genes that impact his personality. Such a series of transitions also influences the therapies that are deemed appropriate under each condition. We might engage with the psychological levels through psychotherapeutic means, while the purely physical requires a different treatment (e.g., excision of a thyroid tumor). Difficulties arise when an inappropriate level of treatment is applied at the level of the patient’s problem.
Disorders of Emotion Emotions are subjective feeling states; yet they also carry physical correlates that may be apparent to observers. Hence, the frightened man experiences a subjective state of panic, an urge to escape, while his body prepares for ‘fight or flight,’ with arousal of his sympathetic nervous system (increased heart rate, blood pressure, blood flow to the muscles, increased sweating, respiration, and dilatation of the airways); his response may be obvious to those around him. Some emotions are so ‘hard-wired’ that they are similarly expressed across all human cultures. These, so-called basic, emotions (fear, anger, distress, disgust, surprise, and joy) are associated with distinct facial expressions that may emerge even in the absence of an observer (i.e., the depressed woman may still look sad even when she is alone). Other, so-called cognitive emotions necessarily implicate others: love, guilt, shame, embarrassment, pride, envy, and jealousy; these invoke relationships, whether proximate or past. Mood and Affect In psychiatry, it is common to describe emotions using the words mood and affect. Mood describes a subjective state, what the person is feeling. Affect is how they appear to others. Hence, mood and affect may be congruent (the woman with depression is crying) or incongruent (the man with
Psychopathology and Consciousness
schizophrenia is laughing while he describes his stepfather raping him). Mood and affect may be abnormally elevated or lowered, may be exaggerated, diminished, or absent. They may alternate between the extreme highs and lows of bipolar disorder (manic depression; F31) or the lesser peaks and troughs experienced in cyclothymia (F34.0; Table 3). Mania is characterized by extreme changes in mood, affect, behavior, and cognition (F30 and F31.1–2). The mood is elevated, as is the affect (indeed, a subject’s humor may be quite contagious).
Table 3
253
Speech and movement are accelerated, there is no need to eat or sleep; sexual activity increases. Thoughts accelerate and may acquire new significance, so much so that the person loses contact with reality: he believes that he is God, that he controls the world, has acquired new insights, talents and energy and should not be constrained by others (i.e., he becomes grandiose). He may describe frankly psychotic perceptions: voices telling him he is a genius, lights and visions abound. Unfortunately, happiness may quickly give way to irritability and anger, and manic patients may harm themselves or
Disorders of mood and affect
Exemplar
Selected features
Mania
Episode of markedly elevated mood, affect, altered behavior, and thinking; speech and movement accelerated; thoughts grandiose, heightened energy and disinhibition. Mania is a psychotic state (delusions of grandeur and hallucinations), hypomania is the nonpsychotic prelude. Episode of marked lowering of mood, affect, reduced self-care, and impoverished thinking; may vary from neurotic to psychotic (delusions of guilt, poverty, nihilism). Risk of suicide (especially early on in treatment, as energy returns). Intercurrent episodes of mania and depression, the latter predominating over the longer term. Risk of suicide when depressed, misadventure when ‘high.’ A longstanding pattern of instability of mood with mild ups and downs, mood and affect are recognized to be changeable over days and weeks. Might be observed to be ‘moody.’ Persistent lowering of the mood, with some features of depression, but insufficient to satisfy the criteria of full depressive syndrome. Affect may often appear ‘miserable’ over a long term. Persistent anxiety markedly interrupts normal relationships and vocation. Prominent symptoms and signs of autonomic arousal. Anxiety precipitated by specific objects (e.g., ‘simple phobias’ of spiders, snakes, heights, plane travel, etc.) or social situations (e.g., crowded places and being away from home, in agoraphobia; intimate situations where behavior may be observed by others, such as speaking or eating in public, in social phobia). Sudden onset of fear and autonomic arousal, coming ‘out of the blue’; fear that one may collapse and die. May complicate agoraphobia. Similar to panic, but in a setting of extreme environmental threat (e.g., postbomb explosion). A response to a change in life circumstance, such as the end of a relationship, which may be characterized by anxiety or depression. Anxiety may be permanent but heightened in response to environmental cues (resembling the initial trauma); recurrent intrusive imagery of the event; bad dreams, phobic avoidance of similar environments (e.g., avoiding tunnels and stair wells after a train crash). Anxiety accompanies ego-dystonic intrusive thoughts and may increase until the tension can be released by performance of a compulsive act (e.g., hand washing, counting rituals, etc.). Anxiety is associated with the fear of illness; worries or over-valued ideas predominate. Similar theme to hypochondriasis, but the patient presents on multiple occasions concerning physical symptoms, which are found to be without pathological cause. May constitute an oversensitive awareness of normal bodily sensations. Lack of feelings, persistent absence of responsivity (may be seen in chronic schizophrenia or end stage dementia). Can be an alarming disconnection between affect and reported mood, as seen in schizophrenia, for example, laughing when they should be crying. Classically associated with cerebrovascular disease in the elderly; the subject may cry easily, may be excessively moved by sentimental topics. Sudden rage or panic in someone with dementia who is failing on tests of cognition (best avoided by gradual, sensitive examination of the cognitive state).
Depression Bipolar affective disorder Cyclothymia Dysthymia Generalized anxiety disorder Phobic anxiety
Panic disorder Acute stress reaction Adjustment reaction Posttraumatic stress disorder Obsessive–compulsive disorder Hypochondriasis Somatoform disorder Blunting of affect Incongruous affect Emotional incontinence, hyperemotionalism Catastrophic reaction
254
Psychopathology and Consciousness
others while they are ‘high.’ They are also at risk of misadventure, behaving uncharacteristically (e.g., through promiscuity or drug taking). Hypomania (F31.0) describes the nonpsychotic prelude to mania; it may be quite a pleasant state, associated with new thoughts and energy. The patient might be reluctant to seek treatment. In depression (F32 and 33) the mood and affect are lowered; the patient feels sadness, which may be accompanied by tearfulness and slowing, so-called retardation. To the observer, the depressed person appears under-active, hunched and crying, or detached and vacant. However, in the elderly patient there may be pronounced agitation, torment. Appetite and sleep fail, weight may be lost, thinking becomes an effort. Thoughts turn to failure, hopelessness, feelings of worthlessness, and guilt; a minor misdemeanor from the past assumes new significance. If there are psychotic features they are mood congruent: voices are pessimistic, visions are of Hell or disaster, the body feels diseased, the bowels have ‘turned to stone’ (Cotard’s delusion or nihilism; F32.3; Table 3). The depressed patient may kill himself to end his suffering, because he thinks he deserves punishment or because he is a burden to others. The woman with postnatal depression (F53) is at particular risk of killing herself and her child (‘to save him from this cruel world’). Anxiety states (F40–48) are states of fear, marked by an awareness of autonomic arousal: racing heart beats, sweating palms, increased respiration, paraesthesia (numbness and tingling) of the fingers and toes. Such a state may be more or less constant (as in generalized anxiety disorder, F41.1) or episodic (precipitated by specific objects in the phobias, e.g., spiders, F40.2, or arising ‘out of the blue’ in panic disorder, F41.0; Table 3). Sometimes anxiety is accompanied by disgust, as in obsessive–compulsive disorder (F42), when the patient feels contaminated, ‘unclean.’ Sometimes, anxiety and depression are combined (F41.2), the patient wringing her hands, miserable, and agitated. The most difficult mixed state to diagnose can be that of mixed mania, F31.6, where, despite all the behavioral features of increased activity, rapid speech, and fleeting thoughts, the mood described (subjectively) is one of sadness and dread.
In the histrionic individual (F60.4) the expression of emotion may be superficially extreme, but it seems to carry no deeper feeling; it appears exaggerated yet ‘shallow,’ the performance is similar to that of a bad actor: one notes the caricature of an emotion, but detects no authentic feeling behind it. The patient can flit from one emotion to another without any apparent awareness of their incongruity (e.g., the man intoning slowly, with eyes closed, as he describes the funeral of a friend, with apparent pathos, smiles suddenly as he mentions that he will buy a car with the money bequeathed him). In the borderline personality (F60.31) the mood can be labile, changing acutely in response to arguments or disappointments; essentially, self-esteem is unstable, the patient decompensates easily; they may self-harm repeatedly. In these cases the feelings appear ‘felt,’ but they may be transient. In a nonpejorative sense, they appear ‘adolescent’: sensitive, quick to appear, flaring up intensely, and then fading. The problem is that the period of intensity may give rise to severe self-harm (e.g., cutting or overdosing). Deficit states: While one may conceptualize mania, depression, and anxiety as exaggerations of normal human emotions, however much they may have become contextually inappropriate, in some conditions, such as schizophrenia (F20), it is clear that the feeling tone has been lost. The affect can appear ‘blunted’ as if the patient is incapable of feeling happiness or sadness. When asked about his mood he may simply say that he feels ‘blank,’ ‘vacant,’ ‘nothing.’ Such a deficit may be associated with a lack of will and initiative; the patient says little and stays in bed much of the day. Similarly, some patients who have undergone head injuries or the effects of frontal brain tumors or strokes may lack emotion, appearing grossly apathetic or else facetious, embarking upon silly pranks. In organic states, hyperemotionalism may present as a lowered threshold for tearfulness; a man may notice that his moods are stronger, more vivid; he cannot prevent himself from crying when he encounters family or friends. He cries easily at the news on the television. In some forms of epilepsy intense joy, rage, laughter, or crying may each arise, as a prelude (aura) to a seizure (ictus), though they are likely to be short-lived and incongruent under the circumstances. Nevertheless,
Psychopathology and Consciousness
they may be frightening to witness. Ictal violence may be completely disinhibited and very severe. Incongruous affect also appears rather disconcerting and is seen in schizophrenia (where it may be associated with speech disturbance): the patient seems to smile while discussing pain or distress.
Disorders of Thought and Speech Classically, textbooks of phenomenology describe disorders of thought as a single category; however, it actually constitutes an amalgam of disordered thinking, thoughts, and speech acts. There are two types of information that we are attempting to gather in a descriptive psychopathological interview: what the person tells us about their thinking, that is, what it is that they experience (subjectively); and how their thoughts are organized and expressed to us (objectively): hence, we are (inevitably) drawing inferences about their thinking and their thoughts from listening to their speech (and sometimes reading what they have written). For this reason it is helpful to divide thought disorders into those of: 1. thinking (the process), 2. thought form, and 3. thought content (or theme). It should be noted that varied vocabularies have been applied, some theoretically derived, some derived from observation; this account will attempt to be as purely descriptive as possible (see Table 4). Regarding the thinking process, we are interested in the emergence and flow of ideas; the spontaneity exhibited, and the coherence of the stream of thought. If we ignore the content for a moment, we are really listening for whether the flow and cadences of this person’s speech sound as they might if we were engaged in a ‘normal,’ healthy conversation. Clearly, the clinical interaction is not ‘normal’ for the patient; they may be nervous, guarded, embarrassed, or resentful about ‘having to see a psychiatrist.’ Nevertheless, we must listen carefully to their speech. In general, in depression and mania, even if the subject matter is distressing it usually remains understandable; and even if we cannot interrupt the manic patient
255
we can probably follow what they are saying. When speech becomes incoherent, muddled, and disjointed, we are more likely to be dealing with schizophrenia or an organic mental state. When we use the specific term ‘formal thought disorder’ we are denoting the incoherence of schizophrenia (F20). This is one of those psychopathological terms that have to be used very precisely (Table 1). The form of a thought and its content (theme) are summarized in Table 5. The most important forms of thought that a psychiatrist detects are those of overvalued ideas, obsessions, and delusions. Within each of these there is a theme: the concern of the anorexic teenager with her fear of fatness (F50.0), the obsessional’s guilt that he thinks of feces (F42), and the delusional system which transforms the reality of the paranoid patient (e.g., F22). One has to listen, but also to probe ideas, one has to ask questions, even challenge what the other believes, in order to gauge whether they really believe it, whether they might occasionally question it themselves and, perhaps most importantly of all, whether they will act upon it. Risk Most psychiatric diagnoses are associated with an increased risk of self-harm. This is most marked in psychoses (especially depression and schizophrenia), and is especially likely among those who are without hope or in pain, who have lost much through their illness and who lack social support and family. One should always enquire about suicidal ideation, its frequency and intensity, any planning that has occurred and what opposing constraints there may be; suicidal risk may be ameliorated somewhat by religious belief, ‘something worth living for’ and the existence of people whom the patient would not wish to leave behind (bereaved by his death). Self-harm may be ‘driven’ by certain phenomena: command hallucinations, telling the patient to kill herself, or motor restlessness (akathisia) induced by certain drugs (below); sometimes self-harm is a misadventure driven by a delusion (‘‘I am invincible, I can fly’’; ‘‘if my hand does not burn, I have been saved’’). Harm toward others is greatly stressed by the British media. Particular attention is devoted to ‘stranger violence’ perpetrated by men with
256 Table 4
Psychopathology and Consciousness Disorders of the process of thinking and its expression (speech)a
Behavior
Possible causes
Definition
Silence
Mutism
Rarely primary (a patient who never acquired the power of speech), usually secondary – ‘involuntary’ (due to depression, mania, schizophrenia, brain lesion, such as left frontal stroke, hysteria) or ‘voluntary’ (elective mutism). ‘Psychiatric stupor’ describes a mute patient whose eyes follow events in the room, cf. ‘neurological stupor’ in which consciousness is obtunded. From the patient’s perspective the problem may be one of a lack of thoughts (depression, schizophrenia, medial frontal stroke) or their multitude incoherence (mania, schizophrenia). Sudden severe traumatic events may render patients speechless; they may appear psychiatrically stuporose (above). Rarely a schizophrenia patient may later describe how their thoughts were suddenly ‘stopped’; they may implicate a persecutor (someone who ‘took them away’). Primary language, not the language of the interview; patients with low intelligence may speak slowly, as may the suspicious, embarrassed or frightened. Reduced volume (amount) of speech in schizophrenia or depression. Seen with psychomotor poverty or retardation, respectively. Left frontal lobe lesions may render the patient mute, or nonfluent (Broca’s aphasia), in which case the patient will appear to be trying to speak, though unable to do so, and will seem to perceive their impairment. In transcortical motor aphasia a frontal lesion renders the patient unable to speak spontaneously though he may be able to repeat what others say. May reflect potentially reversible causes such as intoxication, drug overdose, or transient ischemic attack; if permanent, suggests neurological disorder, for example, classically, cerebellar disease, distributed cerebrovascular lesions (strokes), multiple sclerosis, motor neuron disease, Huntington’s disease. Repetition of the same words over and over, may indicate confusion, schizophrenia, or noncooperation. The patient’s repetition of the last word of their own phrase is said to constitute an extrapyramidal sign. When the repetition is of the interviewer’s words: this may (rarely) be a symptom of schizophrenia or transcortical motor aphasia or else a sign of noncooperation. Seen in organic brain states (classically frontal lobe syndromes) and chronic schizophrenia. The patient returns to prior themes repeatedly, inappropriately, and out of context. Normal amount of speech but the subject says little. May be normal, longwinded, and circumstantial. May have difficulty concentrating; may be easily distracted by environmental cues, as in delirium, schizophrenia, attention deficit hyperactivity disorder. Rarely ‘talking past the point’ in hysterical states. In schizophrenia, the patient may use stock phrases or excessive, pseudotechnical, ‘stilted,’ arcane forms of speech, but still convey little. He may lack a goal and wander away from the point and not return. Sometimes speech is so self-referential that it conveys little new information (the patient brings all enquiries back to himself and his concerns). Language structured according to its surface qualities (meanings and sounds, respectively), in schizophrenia, hypomania, and sometimes hysteria. (‘The present sure is tense,’ with apologies to Captain Beefheart.) Speech fast but coherent, may be difficult to interrupt. Tends to become loud and emphatic. Characteristic of hypomania, then mania. In mania, the patient rushes from topic to topic although he can still be followed. Again, running from thought to thought, but there is an elliptical connection between successive themes. Mention of birds may turn to birds of prey, then to forms of prayer, then to religion. As mania speeds up the links between thoughts may become less direct. The interviewer may detect where links were lost. Difficult to keep pace with, best recorded and then described verbatim. In the above example, the spoken (manifest thoughts) might skip from birds to prayers to bells. . .
Shock Thought block Reduced speech
Inarticulacy Alogia Nonfluent dysphasias
Slurred speech
Dysarthria
Repetition of words
Verbigeration Palilalia Echolalia
Repetition of themes and phrases
Perseveration
Poverty of content
Empty speech
Puns and clang associations Increased rate of speech
Pressure of speech Flight of ideas Tangentiality Knight’s move thinking
Continued
Psychopathology and Consciousness Table 4
257
Continued
Behavior
Possible causes
Definition
Incoherent speech
Derailment
The patient (with schizophrenia) slips between unrelated topics with successive sentences. The patient (with schizophrenia) slips between unrelated topics within the space of a sentence. ‘‘In Panama my fisherman’s waitress swore yellow glasses in a bird’s nest.’’ Word approximations, may be seen in dementia and schizophrenia, for example, gloves referred to as ‘handshoes.’ Sometimes the process is one of elimination: ‘‘I went to. . .you know. . .the building, not the school. . .no, the building, not the shop. . ..after the church. . .the hospital. That’s right!’’ The patient makes up his own words. Characteristic of schizophrenia (but important to exclude malapropisms and poor educational attainment). Also called schizophasia, when speech becomes totally incoherent. Left superior temporal lobe lesions may render a patient incoherent but fluent (Wernicke’s aphasia), in which case they will not appear to perceive their own impairment.
Illogicality Paraphasias
Neologisms Word salad Fluent dysphasis
a
Always attempt to establish the patient’s first language and handedness.
schizophrenia or personality disorders. The risk to others is greatest from those who are paranoid, in the lay sense of the word (Table 1); who abuse illicit substances (especially, crack cocaine and amphetamines); who are nonadherent to conventional medications; and whose personalities might have been prone to violence anyway, even in the absence of a mental illness (men who were ‘antisocial’ and/or ‘paranoid,’ in the lay sense, before they ever became mentally ill). Absence of empathy for others or remorse toward victims are poor prognostic indicators. In terms of relationships, jealousy, delusional, or otherwise may lead to violence against a partner, and erotomanic delusions (of being loved by another, at a distance) may occasionally lead to stalking of celebrities, politicians, doctors, and other figures regarded by the patient as hierarchically significant. There is a particular risk to the families and partners of those who are stalked in this way (the deluded individual may believe that their ‘relationship’ with the beloved might be consummated were it not for the beloved’s family). Ego Boundaries Thoughts and perceptions may sometimes overlap in the phenomenology of schizophrenia (F20). If a man tells us that he is thinking other people’s thoughts and that he feels their point of entry into his skull, is he admitting to thought insertion (a false belief ) or a physical hallucination (a false perception) or both? Clinically, it is best to describe
the phenomenon as clearly as possible. In terms of theory what his problem exemplifies is the porosity of the so-called ‘ego-boundary,’ the notion that we may differentiate our own selves from the outside world. In schizophrenia, these territories overlap repeatedly: patients experience others’ thoughts and think others access theirs; the physical borders of their body are breached and viscera are touched or moved by agents whom they cannot see. Empirically, it seems likely that such disturbances implicate abnormalities of parietal cortical regions; clinically, what is important is that the patient should be afforded space and not be unnecessarily crowded or touched by those caring for them. Physical contact may be misperceived. Misidentification Syndromes The elision of perception and belief is also apparent in the so-called misidentification syndromes. In these, awareness of others’ identities is either lost or bestowed inappropriately. In the so-called ‘illusion of doubles’ (Capgras syndrome) the subject believes that someone known to them (often a spouse or child) has been replaced by someone else, sometimes a real enemy, sometimes a ‘robot’ or alien. The sufferer agrees that the object of their problem ‘looks like’ the person they should be but ‘something has changed.’ Often it is that the person looks less ‘real,’ less herself; there is a resemblance, but that is all. The phenomenon may be frightening for those involved and may
258
Psychopathology and Consciousness
Table 5
Disordered forms and contents of thought and speech
Form
Description
Over valued ideas
Single themes that can come to dominate a person’s life, without being psychotic; often value-laden, for example, the pursuit of thinness in anorexia nervosa, the avoidance of waste in obsessive hoarding. The themes of neuroses may also be of this sort: the fear of illness in hypochondria, over concern with normal physical sensations in somatoform disorder, the cognitive distortions of depression (‘‘People don’t like me’’), and dysmorphophobia (‘‘My nose is too big’’). Recurrent intrusive thoughts, which the sufferer recognizes as his own, the content of which may be absurd distressing, and which he tries to resist/suppress. Common themes include contagion and symmetry; magical thinking may be invoked (‘‘if I walk on the cracks in the pavement my father will get cancer’’). Often accompanied by rituals and compulsions, for example, hand-washing. In the addictions, cravings may behave like obsessive thoughts, though the extent to which they are resisted varies with the stage of addiction (‘‘I need a drink’’). Abnormal beliefs, acquired through abnormal inferences and out of keeping with the subject’s culture. Said to be incorrigible to counterargument. May be primary (rare), coming out of the blue, or secondary to other phenomena, for example, hallucinations or mood states. Characteristic themes: Persecutory delusions – others are against the patient Delusions of reference – others are communicating with him (e.g., the song on the radio is ‘meant’ for him) Grandiose delusions – mania, schizophrenia, GPI Delusions of guilt – depression Nihilistic delusions – Cotard’s delusion Hypochondriacal delusions – depression Religious delusions – may also be grandiose Delusional jealousy – Othello’s syndrome, particularly dangerous, commonly seen in alcoholic males who may kill their partners Erotic delusions – De Clerambault’s syndrome, also high risk Delusions of control – the patient perceives their thoughts and actions to be under the control of external agencies. Some risk of violence, especially if the perceived controller is identified in proximity to the patient Classic first rank symptoms (Table 6) are thought insertion, withdrawal, and broadcast. Alien thoughts are inserted, one’s own thoughts extracted, or known at a distance. Characteristic of schizophrenia. In both schizophrenia and depression there may be a ‘milder’ experience of ‘influence,’ so-called ‘passivity’: the thoughts and feelings are still one’s own but they are influenced or ‘made’ by external agencies. Occasionally, people living in close relationship with each other (and relative isolation from others) may come to share the same delusion. Usually, there is a dominant, primary subject who is mentally ill, and in whom the delusion originated, and a passive, secondary subject who has come to share the delusion. Separation of the pair may lead to spontaneous recovery in the second subject. However, they are likely to have developed an enmeshed relationship so that separation may itself be very traumatic. When two are involved, this has been termed folie a deux. Some small sects may exhibit similar dynamics (often it is the leader who is disturbed while the sect ‘members’ are relatively normal).
Obsessions
Delusions
Beliefs about thought possession
Shared delusions
constitute a risk for the misperceived (the ‘impostor’) as the patient may try to kill them. In Fregoli’s syndrome, the problem has been ‘reversed’: instead of an intimate having been replaced by a stranger, now strangers are replaced by a known (probable) persecutor. The syndrome may involve an elderly woman on a ward, who becomes ill postoperatively and notices that staff have been replaced by a neighbor (in disguise) whom she does not like. Again, there may be assaults upon the misperceived. The neurological term for the inability to recognize a familiar stimulus, in the presence of preserved sensation and movement, is agnosia, classically a
consequence of inferior temporal lobe lesions. Hence, it may be that misidentification syndromes occupy the same perceptual hinterland as ‘prosopagnosia’ seen in Alzheimer’s disease, when a patient no longer recognizes those whom they should, for example, the daughter who cares for them.
Disorders of Perception In order to understand disorders of perception it may be helpful to consider the normal relationship between our thoughts and experiences.
Psychopathology and Consciousness
Top-Down and Bottom-Up Processing Imagine standing in a very crowded room, among many other people talking. While one is engaged in conversation one may be able to hear one’s interlocutor’s voice quite clearly, even amidst the din of a hundred other voices. Hence, we may clearly choose to focus our attention upon one stream of incoming information. This has been termed ‘top-down’ cognition. It implies that higher processes (and centers) in our brains can focus on salient information, despite our presence within a complex environment containing many competing streams of data. We have an element of cognitive control over what we choose to attend to. However, even though we may be focusing upon what our companion is saying there are still moments when we may be suddenly distracted by the mention of our own name by someone else. Thus, even though we were using ‘top-down’ processes to focus upon one stream of data, somewhere in our brains another process was analyzing those extraneous data emerging around us, so that a single salient stimulus (our name) might gain access to consciousness, ‘bottom-up.’ This situation exemplifies the relationship between the contents of our awareness, in consciousness, and the streams of stimuli arising within our environment: we need to focus on the salient and while we can do so volitionally, consciously (‘top-down’) to an extent, we must also rely upon automatic (‘bottom-up’) mechanisms to detect significant changes in other channels of information. When we are tired, or our hearing is failing, then these processes can become difficult to maintain and their data to disambiguate. In the above example, the voice of our interlocutor constitutes a true (veridical) stimulus in our environment. While we focus upon what he is saying we can accurately hear his words, even if the room is noisy. However, should we fail to detect his signal properly, should we mishear his speech, then we may experience an illusion. He says ‘‘Chris is retired now’’; we hear ‘‘Chris is getting tired now’’ or even ‘‘Chris has been fired now.’’ Meanwhile, there is the possibility that unattended information in other streams of data might also generate illusions; especially if the illusory material resembles something salient for us (‘is that Minnie Driver?’).
259
Top-down processing is all the more apparent in mental imagery (e.g., in our ‘mind’s eye’), when we ‘call up’ an image from our past or an imagined future (so-called prospective memory). Because imagery is under top-down control it contrasts with obsessions (thoughts that enter the mind against the subject’s will and keep coming back) and also hallucinations (where the percept, e.g., a voice, also acquires consciousness, unbidden). A central problem common to obsessions and hallucinations is that the subject cannot control them using ‘top-down’ techniques, he cannot ‘make them go away.’ Now, while illusions and inner imagery are of marginal interest in psychiatry, hallucinations carry much more diagnostic significance.
Hallucinations The problem for patients with a number of psychiatric disorders is that perceptions emerge into their consciousnesses, which do not originate from real external stimuli and cannot be controlled. It is as if aberrant bottom-up processing cannot be modulated or ‘turned down’ by top-down processes. Hence, a man with schizophrenia (F20) hears someone speaking about him and he cannot make ‘him’ stop. Hallucinations have the quality of reality. They may be simple (e.g., tones, lights) or complex (faces, voices). They can arise out of the blue or be carried by environmental signals (‘functional hallucinations,’ e.g., voices heard in the roar of passing traffic) or else follow on sensations in another modality (‘reflex hallucinations,’ e.g., a voice that is heard when the skin is touched). Sometimes hallucinations ‘could not possibly’ be based on real veridical experiences, for example, with extracampine visual hallucinations, the patient ‘sees’ out of the back of his head; during out of body experiences, he can look down upon his physical body from above. In autoscopic hallucinations he may see his own ‘double’ walking toward him (such doubles can be mirror images or ‘correctly’ transposed, that is, with the left ear to right side of the figure). Auditory hallucinations are particularly characteristic of functional psychoses (above). Some of those described in schizophrenia have acquired special status (as so-called first rank symptoms; Table 6)
260 Table 6
Psychopathology and Consciousness The first rank symptoms of Schizophrenia
Modality
Symptom
Auditory hallucinations
Third-person comments (including running commentary) upon the patient (‘‘He’s looking at the TV; he’s changing the channel. . .’’). Voices conversing between each other. Patient’s own thoughts being spoken aloud, simultaneously in Gedankenlautwerden, after a delay in echo de la pensee (thought echo) Bodily sensations that the patient attributes to external agencies, characteristically affecting internal organs (e.g., ‘‘Professor Hirsch is probing my spine, I can feel his fingers pressing through me. . .’’) The patient experiences his thoughts, feelings, and movements as being under the influence of external entities. They are still his processes, but they are impeded or interfered with by others. Processes are no longer his: thought insertion (he is thinking others’ thoughts), thought withdrawal (his thoughts are taken away), thought broadcast (his thoughts are known to others at a distance). Rare and diagnostically specific occurrence: includes the so-called delusional perception in which a normal percept, not thematically related to his situation, suddenly precipitates a delusional interpretation by the patient (‘‘I was walking down the street, the traffic lights changed to green, and I knew then that I was the Messiah’’); also the autochthonous idea, a delusion which suddenly appears without precedence (‘‘As I sat, thinking, in the library I suddenly realized that I had been chosen to save the world from the masons’’).
Bodily hallucinations Passivity phenomena Alien phenomena Primary delusions
not because they are confined to this disorder, not because they are the only auditory hallucinations described in it, nor again because every schizophrenia patient experiences them, but because they are ‘so unusual’ that they are highly likely to be reliably ascertained and agreed upon by different psychiatrists. These auditory hallucinations, described by Kurt Schneider, are listed in Table 6. Hence, the first rank symptoms have clinical utility because they are so characteristically abnormal. However, people with schizophrenia may also experience second person voices (speaking to, rather than about, them) and, unfortunately, third-person voices can occasionally arise in people with depression, mania, or epilepsy. Again, we are working with the phenomenological nature of psychiatric diagnoses and attempting to compare the constellation of symptoms occurring in a single case with an empirical, syndromal profile, to find the best ‘match.’ Generally speaking, in depression and mania, auditory hallucinations are more likely to be second-person in form and mood-congruent in content. While the voices heard in schizophrenia tend to be derogatory and intrusive, seemingly aware of what the patient is thinking, those in depression say depressing things (‘‘The world is ending, you have cancer’’), while those heard by the manic may exalt him (‘‘You are a genius, you are cleverer than Ray’’).
There are many less common causes of auditory hallucinations and these tend to lack characteristic features. However, there are some notable exceptions: the long-term heavy drinker may develop alcoholic hallucinosis (F10.52), hearing derogatory voices when drinking, in the absence of delusions (i.e., it appears to be a disorder solely of perception); musical hallucinations are another example: they may arise in otherwise healthy elderly people who experience hearing impairments (these patients usually realize that the music is not ‘real’). Visual hallucinations are more characteristic of organic mental states; particularly those induced by drugs, for example, hallucinogens such as lysergic acid diethylamide (LSD) (F16), or epilepsy or migraine; but they can also arise in the functional psychoses, delirium and dementia. As with musical hallucinations, occurring in the hard of hearing, elderly people who are visually impaired may sometimes develop visual hallucinations, which may be simple (flashing lights, colors) or complex (small figures, in the so-called Charles–Bonnet syndrome). If cognitively intact, the older patient usually recognizes that such phenomena are not ‘real’ (hence, in one sense, these constitute ‘pseudohallucinations’; Table 1). The greatest diagnostic significance of visual hallucinations is that they prompt one to consider ‘organic’ disorders (above). Olfactory and gustatory hallucinations (of smell and taste, respectively) in general suggest temporal or frontal lobe pathology, particularly in the form of
Psychopathology and Consciousness
focal partial epilepsies (classically, temporal lobe epilepsy), where unusual smells, such as burning flesh or rubber, may be accompanied by an epigastric aura (‘butterflies in the stomach’) and facial flushing. In functional psychoses, the schizophrenia patient may also smell burning flesh, with distressing associations (e.g., Hell, the Devil), while the depressed person smells her body rotting or giving off noxious fumes. Physical hallucinations, that is, disturbances of bodily perception, constitute a wide range of sometimes intermingled phenomena, which go by a variety of names in different texts: tactile, haptic, visceral, somatic, and so on. They pose an interesting philosophical problem in that many physical hallucinations do not concern ‘veridical, external stimuli’; in other words, rather like an ‘itch,’ they cannot be verified by a third-person observer. Nevertheless, what tends to attract attention is the proffered explanation for such phenomena (i.e., what the patient tells us about them). In psychiatric terms, the most telling is a hallucination implicating the viscera, especially the sexual organs; this usually denotes schizophrenia (F20). The patient may say that her persecutors are raping her, that they can change the shape of her spine (and this constitutes a first rank symptom; Table 6). Organic disorders of the parietal cortex, especially on the right side, may sometimes give rise to the experience that the left half of the body is occupied, controlled by someone else (‘somatoparaphrenia’). In substance misuse, cocaine is particularly associated with formication, the feeling that there are insects moving under the skin (F14.04). Delusional infestation (with very similar phenomenology) may occasionally arise as an isolated symptom in the elderly (Ekbom’s syndrome; F22–23).
Disorders of Cognition It is a moot point whether we are ever really assessing specific phenomenological domains in isolation from cognition during the psychopathological interview, for it might be argued that cognitive processes underpin and support all of the phenomena described here anyway (e.g., those cognitive processes integral to perception and
261
speech). However, when we talk about cognition we are usually referring to those explicit aspects of intellectual function that the patient might regard as contributory to their ‘thinking,’ for example, attention, concentration, and memory. It is also clear that as we progress from disorders of emotion and perception toward disordered cognition, we are increasingly dependent upon objective data (e.g., measured intellectual performance) and less upon the subject’s point of view. There are several reasons for this: first, the layperson may not necessarily have stopped to assess their own cognitive state (as opposed to their having inevitably experienced their own emotional state); second, the very deficit of cognition that we are investigating might itself have prevented the patient from perceiving that deficit. We rarely ‘know what we do not know’; hence, amnesic patients cannot access what it is that they have lost in order to tell us what it is that is missing. If ‘attention’ is the ability to focus on information (e.g., the examiner’s instructions), and ‘concentration’ the ability to remain focused (over seconds, minutes, or longer), then it is clear that a great many mental disorders might potentially interfere with such abilities. Some reasons may be ‘normal’: the patient is bored, tired, uninterested, or more interested by something else happening elsewhere. Other causes might be transitory and, hence, potentially missed or misleading: intoxication, sedation, concussion, delirium, or hallucination (Table 7). Still others might be longstanding and definitely pathological: attention-deficit disorder, dementia or learning disability (Table 8). It follows that while it may be helpful to detect abnormalities of attention and concentration, in order to gauge the patient’s safety in their environment and their ability to engage with others, the differential diagnosis for a detectable impairment is very broad indeed. Much depends upon the accompanying data. When we consider ‘orientation,’ we mean, ‘does the patient know where they are in ‘‘time, place, and person’’ ’? In most functional mental illnesses, such as the psychoses and neuroses, people retain this type of information (though the institutionalized may occasionally lose track of time; ‘everyday is much the same’). People with chronic schizophrenia sometimes underestimate their own age.
262
Psychopathology and Consciousness
There are many possible causes of delirium (acute organic brain state)
Table 8 There are many possible causes of dementia (chronic organic brain state)
Potential cause
Exemplars
Exemplars
Drug or toxin Infection
Drug or alcohol intoxication or withdrawal. Viral or bacteriological infection affecting the body systemically Viral or bacteriological meningitis or encephalitis TB, HIV, syphilis, etc. Diabetes mellitus, thyroid disease Hypo- or hyperglycemia Hypo- or hypercalcemia Electrolyte disturbance, renal or hepatic failure Porphyria Thiamine deficiency, as in Wernicke–Korsakoff syndrome. Seen most in alcohol dependence, rarely in malabsorption syndromes, hyperemesis gravidarum. Systemic lupus erythematosus, sarcoidosis Behcet’s syndrome CNS tumor (primary or secondary) Para-neoplastic syndrome Endocrine tumor, for example, pheochromocytoma Traumatic brain injury Vascular, for example, transient ischemic attack (TIA), subdural hematoma Seizure Migraine Sleep disorder, for example, narcolepsy
Potential cause Degenerative
Alzheimer’s disease FTDs, including Pick’s disease and Semantic dementia Huntington’s disease Parkinson’s disease, Lewy-body dementia Creutzfeldt–Jacob disease Normal pressure hydrocephalus Multiple sclerosis Multi-infarct dementia Chronic subdural hematoma Cranial arteritis Binswanger’s disease CNS tumor (primary or secondary) Severe single head injury Repeated head injury in boxers (dementia pugilistica). Encephalitis of any sort Neurosyphilis Sustained uremia, hepatic failure, dialysis (aluminum toxicity) Wilson’s disease Hypothyroidism Alcohol Heavy metal poisoning (lead, arsenic, thallium) Postcardiac arrest, chronic respiratory failure, carbon monoxide poisoning, anemia Sustained lack of vitamin B12, folic acid or thiamine
Table 7
Endocrine or metabolic
Nutritional
Autoimmune Neoplastic
Neurological
However, in the ‘organic’ states of delirium and dementia, temporal and spatial information will often be disturbed, less so the subject’s identity (though they may confuse and misidentify others; above). The perception of time is a complex process. In psychopathological terms it is of note that time can seem to speed up in certain states (e.g., mania, temporal lobe epilepsy, and LSD intoxication), and slow down in others (especially depression). Patients suffering the consequence of bilateral medial temporal lobe destruction (e.g., following herpes simplex encephalitis) may exist in a perpetual present: they have no recent memory beyond a few minutes ago; they have lost that interval between the onset of their illness and the present (‘anterograde amnesia’). Following head injuries, poisoning or anesthesia patients may lose awareness of moments preceding the target event
Vascular
Neoplastic Traumatic Infection Metabolic Endocrine Toxic Anoxia Nutritional
(‘retrograde amnesia’). A heavy drinker of alcohol may lose temporal islands in his immediate past (‘blackouts’), for example, he cannot recall getting home the night before. Abnormalities of place are not unusual in organic brain states: patients often do not know where they are, indeed one of the presenting complaints of delirium and dementia may be getting lost, not knowing one’s way around the neighborhood. In reduplicative paramnesias the subject believes their present physical surroundings to be someplace actually known to them from former experience (probably a complex amalgam of disorientation and illusions and/or hallucinations of the former environment); this is classically associated with delirium tremens (alcohol withdrawal; F10.3). In some forms of epilepsy and migraine, the environment never seen before may nevertheless seem familiar (de´ja` vu) or the known environment
Psychopathology and Consciousness
unfamiliar (jamais vu). Sounds may appear inappropriately familiar (deja ecoute) or unfamiliar ( jamais ecoute). It is usually only in advanced dementia that a person loses knowledge of their own identity. If an apparently physically healthy person, who uses language normally and interacts with everyday objects purposefully, says that they do not know who they are, then it is highly likely that the cause is psychogenic (hysterical fugue, F44.1, or malingering, F68.1). Memory, per se, as opposed to orientation, is a further group of processes, the distinctions between which may be neuroanatomically informative. There are competing methods for classifying memory and its disorders, but a clinically useful schema is shown in Table 9. Memory can be conceptualized as the storage of information for action. Hence, it follows that information may be brought out of storage, either by the subject accessing the data themselves (through a process termed ‘retrieval’) or through some external stimulus prompting ‘recall.’ Retrieval is the more difficult to achieve and may fail earlier in
Table 9
263
cognitive disorders. It is generally easier to recall something if prompted with a ‘cue.’ Retrieval can be seen as a relatively ‘top-down’ process, while recall is more ‘bottom-up’ or stimulus-driven. During delirium (by definition) all mental processes may be disturbed (F05). This is also the case in the dementias (F00–03), yet it is clear that at their onsets the dementias may be characterized by relatively discrete psychopathologies that may make them easier to diagnose. For the clinician, there is a noticeable distinction between the dementia of Alzheimer’s disease (F00) where the patient at first exhibits symptoms concerned with forgetting everyday subject matter and spatial coordinates (e.g., ‘getting lost’), and the frontotemporal dementias (FTDs, e.g., Pick’s disease, F02.0) where the first signs might comprise embarrassing, uncharacteristic behavior, for example, a man undoes his fly to urinate at the graveside during a funeral. The psychopathological ‘picture’ reflects something of the brain regions initially implicated in the pathophysiological process: the temporal and parietal regions in Alzheimer’s disease, the frontal regions in FTDs. While most dementia
Memory may be divided into multiple systems
Form
Function
Explicit or declarative memory
That which may be spoken about, explicitly. Includes: Prospective memory ‘Memory of the future,’ really a construction of possible future states (albeit based on known data): thought to implicate prefrontal polar regions. Involved in intentions, plans for future action, strategies. Working memory The contents of consciousness in the immediate present. Classically shown to be able to store a mean of 7 bytes of information (e.g., a telephone number to be dialed). Supported by function of dorsalateral prefrontal executive regions, manipulating data held in posterior association areas ‘online.’ Episodic (autobiographical) memory Remembered events that the subject has experienced in her life. Implicates medial temporal lobe structures, especially the hippocampus. Semantic memory Shared knowledge of the world, including facts, definitions. Implicates inferolateral temporal regions, also parietal cortices. Forms of memory that cannot be described verbally. Include: Procedural memory Memory for how to perform certain tasks, for example, riding a bike, playing the piano. Implicates the basal ganglia especially. Priming A subject may have learned information without being aware that they were learning, in response to environmental stimuli of which they were unaware, or following which they have no conscious memory, that is, their performance in the present is enhanced by prior experience, of which they have no awareness.
Implicit or nondeclarative memory
264
Psychopathology and Consciousness
processes elicit a gradual decline in cognitive function, some are temporally distinct: hence, the classic ‘stepwise’ descent in vascular ‘multiinfarct’ dementia (F01.1), where new episodes of stroke bring incremental deteriorations; and Creutzfeldt–Jacob disease (CJD, F02.1) and ‘newvariant CJD,’ where the illness may affect a younger adult and progress is catastrophic, declining to death in as little as a year. It follows that if memory and cognition are partially ‘localizable’ to certain brain systems then there may be relative sparing of some faculties (initially), while others are lost. In cortical dementias, for example, Alzheimer’s and FTDs, the patient loses episodic memory and visuospatial cognition at first in the former and social control and language function at first, in the latter. However, both may exhibit relatively preserved overlearned manual skills (procedural memory), subserved by the basal ganglia and other ‘lower’ systems. In contrast, a so-called subcortical dementia (such as Parkinson’s, F02.3, or Huntington’s diseases, F02.2) may affect the control of movement early on, also slowing cognition (through involvement of the basal ganglia). So, it is the constellation of symptoms seen that aids diagnosis; we are rarely reliant upon a single symptom. Nevertheless, certain disorders may be restricted to one domain of memory (e.g., the variant of FTD confined to ‘semantic dementia’). Clinically, the most common would be Korsakoff ’s syndrome (F10.6), encountered following the prolonged use of alcohol and the consequences of thiamine deficiency. This is a syndrome in which the subject loses autobiographical memory, exhibiting a dense anterograde amnesia from illness onset forward, and a retrograde amnesia, stretching back from onset ‘toward’ his younger life (he may retain relatively preserved memory for his childhood). Classically, Korsakoff ’s is an amnestic syndrome, not a dementia, because only memory is affected, but it is clear that this is not the sole symptom: sometimes the patient ‘confabulates,’ telling stories to fill in the gaps in his memory. Sometimes the stories can be quite fantastic, and studies have suggested that confabulation indicates frontal impairment (in addition to the usual midbrain pathology of Korsakoff ’s syndrome).
As dementias progress they become less discretely disorders of memory, and come to disturb all elements of mental life. Patients’ personalities may change, their mood states may become labile or apathetic; they may hallucinate and be unable to speak or communicate. Some develop seizures. Behaviors become difficult for carers to cope with: especially screaming, wandering, incontinence, and violence. By this stage, there may be little way of accessing the disordered ‘consciousness.’ Visuospatial and Self-Perception While the emphasis in descriptive psychopathological interviews often focuses upon verbal information, which subjects may describe or remember in words, there is clearly a need to consider nonverbal data, for example, visuospatial information. While in the right-handed subject verbal information is preferentially processed by centers in the left hemisphere, the nonverbal, visuospatial, and acoustic data (e.g., prosody) appear preferentially processed by the right, nondominant hemisphere. Hence, right-sided parietal lesions may leave the patient particularly impaired in a variety of ways concerning the self ’s relation to space. In hemispatial neglect the subject seems unaware of the left side of space; they may eat from only the right side of their plate, neglect to dress the left side of their body, read only half the face of a clock. Initially, the problem may be subtle and revealed only when sensory data are complex (e.g., if there are salient stimuli arising in both halves of the visual field then only those on the right may be perceived). Awareness of the left side of the body may be lost or radically altered: a woman with right parietal epilepsy may report that her left arm is swelling and no longer belongs to her. Following a right parietal stroke a man may report that the left limb in the bed belongs to a relative. Stranger still, following vestibular stimulation (cold water placed in his left ear) he may temporarily regain awareness of his left limb, only to lose it again later. Given that parietal association cortex synthesizes data from many incoming streams of information it is unsurprising that a great many complex disorders may follow its damage.
Psychopathology and Consciousness
If a subject is asked to keep their eyes closed, then somatognosia is a failure to locate an object that touches the skin; finger agnosia is a failure to recognize which of their own fingers was touched; astereognosia describes an inability to identify a three-dimensional form, placed in their hand; while agraphaesthesia is an inability to recognize what has been ‘written’ on the skin under the same conditions. Anosognosia describes a patient’s lack of insight into their own illness. What these phenomena seem to have in common is an impaired appreciation of the state of the body. In left parietal lesions the deficits are more disparate, including apraxia (an inability to copy a purposeful voluntary movement despite the preservation of normal power and sensation – the deficit implicates the programming of motor sequences); when left angular gyrus is affected, ‘Gerstmann’s syndrome’ may arise, in which the subject exhibits dysgraphia, dyscalculia, finger agnosia, and right-left disorientation. While the right parietal cortex seems ‘interested’ in egocentric space – how the world is orientated according to our bodies (to the left and right) – the right hippocampus seems involved in memory for allocentric space (impersonal, map space). Hence, it is particularly implicated in navigation through space.
265
A more nuanced approach to assessing insight was proposed by Tony David. It involves a staged approach: 1. Does the patient recognize that something is wrong? 2. What do they think is the nature of the problem? 3. Do they think they require help with this problem? 4. What form of help do they think is required? By the time one has negotiated these steps one is in a far better position to comment upon the insightfulness of the patient-subject. Indeed, one may find a similar approach to assessing ‘capacity’ in other areas of life: the capacity to make a will, to advise legal counsel, to marry, to decline medical treatment. Capacity and insight relate to specific decisions; we are all better equipped for undertaking certain judgments than others; in any given important decision in life we need to be able to understand what we are being told, believe the information provided (and our informant), know the consequences of different lines of action, weigh them in the balance, and communicate them with our solicitor, doctor, executor. We are all sifting levels of information.
Disorders of Movement Insight It is perhaps fitting to address insight after parietal lobe disorders, given that parietal lesions may deprive patients of awareness of their disabilities (anosognosia). Insight is a multifaceted concept that is deployed in a variety of ways by different disciplines. In terms of descriptive psychopathology, we are primarily concerned with the patient’s understanding of the abnormal phenomena they have described to us. At its most basic, the presence or absence of insight is deployed in the diagnosis of neurosis or psychosis, respectively; indeed one often hears the banal declaration that insight is ‘intact’ or ‘absent,’ a patient being said to possess ‘no insight.’ It is most unlikely that insight can ever be said to be entirely present or absent in anyone, let alone someone who is attempting to survive a major change in their mental state.
All movements (and, indeed, even the absence of movement) are of interest phenomenologically; behavior is indicative of mental life, albeit at some remove. Hence, even the mute patient is ‘describable’ in psychopathological terms. Also, in the normal course of events there are movements that we may take for granted, which are to be remarked upon phenomenologically: the way we use our eyes and hands while talking, the intonations of our voice, the seemingly involuntary shifts we make, on the chair, at the table. Some of these simple forms of information are exaggerated during the psychiatric interview: the depressed person may move very little, offering no expressive aids to speech, while the manic may be standing on the table or trying to climb out of the window. Indeed, the physical space between people is also informative: the frightened man on ‘acid’ (LSD)
266
Psychopathology and Consciousness
cowers in the farthest corner of the room, while the woman elated on ecstasy sits too close, smiles and flirts; an antisocial male may sit, staring fixedly, deploying a direct gaze that feels intimidating. One of the classic texts on forensic psychiatry invites one to consider whether the hairs upon one’s neck stand up in the presence of such a man; the good observer reads their own responses to the situation. When we turn to disordered movement, we usually mean more pronounced and qualitatively distinct disturbances of behavior. These may be divided into voluntary and involuntary movement disorders. Voluntary Movements What can a subject do voluntarily that is abnormal? They can do strange things (jump in the river) or they can do ordinary things in a strange way (stand, waiting for the bus while making a Nazi salute). Patients come to attention in public places usually as a consequence of doing something noteworthy (e.g., standing on the bridge looking at the water; driving the wrong way up the motorway; talking to themselves; exposing themselves or being ‘threatening’). In many of these cases there is nothing abnormal about the movements per se, it is where their action sequence is leading, the goal of such movements that is abnormal (wanting to drown, crash, or masturbate in public). It is also worth recalling that intoxication (with either alcohol or drugs) may disinhibit and ‘magnify’ the expression of dysfunction. However, some voluntary movements do appear unusual. In ‘mannerisms’ a normal task is performed in an abnormal way, a way that might appear symbolic to some (e.g., beckoning for the bus, making a Nazi salute, above). Mannerisms seem to have a purpose (drinking tea from a cup, with one’s little finger raised used to be regarded as quite stylish; certain forms of handshake carry meaning in ‘normal’ circles). In ‘stereotypy’ a voluntary behavior is repeatedly performed, yet it seems purposeless (e.g., rocking to and fro in the chair, rocking one’s leg at the knee while sitting cross-legged). Sometimes the behavior may merely release tension, as when a patient treated with neuroleptic medications experiences ‘akathisia’
(subjective motor restlessness), a state that may be most distressing: never being able to sit still or relax. Sometimes the patient looks restless, but does not feel it; they are rocking to and fro, but cannot say why; when asked to perform a purposeful task while standing, for example, alternating patterns of hand movement, they may ‘tramp’ on the spot (pseudoakathisia). Mostly these behaviors are encountered in people with schizophrenia (F20). It is a moot point whether they are entirely attributable to the disorder, partially driven by its treatment (especially neuroleptic-induced akathisia), or whether the more exotic, historic descriptions of catatonic behavior (F20.2) were facilitated by institutionalized living conditions. Catatonia incorporated a wide variety of signs that are seldom seen today and, when they are seen, rarely coincide as a syndrome: hence, the subject may exhibit ‘waxy flexibility,’ allowing an examining doctor to move his limbs about, into awkward postures, ‘negativism,’ resisting the examiner’s movements, ‘obedience,’ doing what the examiner says even when told not to do so, ‘mitgehen,’ allowing oneself to be moved even when told not to, ‘ambitendence,’ appearing to alternate between mutually exclusive courses of action (hovering on the threshold, attempting to walk into and out of the room at the same time), ‘psychic pillow,’ appearing to sleep with the head raised above the bed, ‘posturing,’ apparently staying in a fixed position for long periods of time. It is hard to consider these behaviors without noting both the straitened circumstances experienced by such patients, incarcerated within asylums, and the emphasis (in these descriptions) placed upon their obeying or not obeying their examiners’ commands. Other behaviors may look abnormal, but are performed normally, for instance compulsive handwashing (F42.1). The behavior is voluntary, but it is performed under some duress, the patient having tried to resist its performance. Similarly, certain habits can attract psychiatric attention: compulsive hair pulling (F63.3), nail biting, binge eating (F50.2), drinking, and so on. In tics, a recurrent, rapid movement of a group of muscles occurs (F95), sometimes without thinking, sometimes in response to an urge to act. In Tourette’s syndrome,
Psychopathology and Consciousness
the tics performed vary from those that are simple, for example, coughing, blinking, head turning, to those that are complex: vocal ejaculations of obscenities (coprolalia) or similarly obscene gestures (copropraxia). Sometimes patients can learn to ‘hide’ the abnormal behavior by morphing it into another gesture or obscuring words with a cough. Involuntary Movements The simplest involuntary movement is a reflex, which may or may not be abnormal (the patella tap, the knee-jerk is normal; while a grasp reflex of the fingers is normal in a baby, but not in an adult). Tremor, also, may be normal or abnormal. Thus we may tremble if nervous or frightened, we may shiver when cold. Pronounced tremors may be seen at ‘rest’ or during action (‘intention tremors’). The former include the normal variant essential tremor (snooker players may allay this with alcohol or betablocker drugs), and the pathological tremors seen in Parkinson’s disease (‘pill-rolling tremor’) or iatrogenically induced, following neuroleptic medication (e.g., haloperidol). Tremors confined to intention are classically associated with cerebellar disease (e.g., postalcohol dependence). More pronounced involuntary movements (‘dyskinesias’) may attract attention to the subject and be very embarrassing. ‘Choreiform’ movements are proximal (involving shoulder and hip joints), jerky and ‘dance-like.’ They are seen in Huntington’s disease (F02.2), Sydenham’s chorea (following streptococcal infection in children) and, rarely, antipsychotic medication. ‘Athetoid’ movements are distal (involving hands, wrists, feet, ankles), sinuous and snake-like; they may be encountered in cerebral palsy, poststroke, and, rarely, postneuroleptics. Choreiform and athetoid limb movements appear to lack purpose. However, some involuntary hand movements may appear purposeful even if they are not: in the ‘anarchic hand’ syndrome, a brain lesion ‘releases’ the contralateral hand to grasp objects in the environment. To the subject the hand appears to have ‘a mind of its own.’ In a right-handed subject a left premotor lesion may release a grasping right hand, while a right premotor or callosal lesion may release an interfering left hand; the subject
267
exhibits ‘intermanual conflict,’ the left hand may undo buttons while the normal right hand tries to do them up, the left may put down the cup, which the right is attempting to raise. Tardive dyskinesias (so-called because they come on slowly after the initiation of neuroleptic medication) may affect the face most noticeably. A patient exhibits abnormal chewing movements, protrusion of the tongue, and head and neck turning. While ‘top-down’ processes may hold the process in check while the subject is alert and unoccupied, distraction with a difficult task (e.g., naming the months of the year in reverse order) may unmask the abnormal movements (an important element of clinical assessment). Similar dyskinesias may affect the fingers and wrists, repetitive movements being more noticeable while the patient is walking. Dystonias are abnormalities of posture caused by abnormalities of motor tone. In torticollis, the head is turned to one side because the contralateral neck muscles are in spasm. In writer’s cramp the muscles of the hand are affected; in blepharospasm, those external to the eyes. In Parkinson’s disease the increased tone of the muscles gives rise to rigidity, which can be felt upon bending the limbs. These signs may also be found in those receiving neuroleptics, and chronic administration may give rise to unusual truncal postures: Pisa syndrome, leaning to one side, Rabbit syndrome, flexing the trunk and arms together. It is important to note that although disorders of movement may be most apparent to observers and less noticeable to the patient, some may be very distressing indeed. Hence, acute dystonia, a painful arching of the neck and back, following acute administration of neuroleptic medication, and associated with occulogyric crises (in which the eyes roll upwards under the upper eyelids), is acutely distressing; while akathisia, a sense of motor restlessness, increases the risk of suicide.
Disorders of Personality While it might be relatively easy to assess a patient’s speech and movement upon first meeting them, it is a great deal more difficult to assess their
268
Psychopathology and Consciousness
personality. Indeed, even after repeated contacts and recourse to coinformants one might still not know another person’s personality. Descriptive psychopathology tends to adopt a typological approach to personality, categorizing abnormal personalities into ‘types,’ little more than caricatures really, listing criteria that must be satisfied, though often these are rather blunt terms. More realistic is a dimensional approach to those aspects of feeling, behaving, and relating, which constitute a person. So it would probably be more valid to regard all humans as located on a continuum when it comes to extraversion/introversion, paranoia, empathy, moodiness, and obsessionality, so that each personality might occupy a point in a highly multidimensional space (there being many facets to a personality). However, clinically, we are required to provide sketches, often on the basis of inadequate information. What is a personality? It is a pattern of temperament, behavior, and relation to others that is probably established by teenage years and which remains generally stable over the long term. Under approximately equal genetic and environmental influence, it seems that we become ‘more genetic’ as we age, ‘more like ourselves,’ a finding that has been attributed to the human ability to create our own environments as we become autonomous. It is also clear that, in the philosophical sense, we are subject to ‘moral luck’: we did not choose our genes or our early environments, so ‘moral responsibility’ is probably at best a relative condition: some have more or better choices than others, but few might be said to be ‘free!’ In terms of psychopathology, a personality disorder represents an abnormal pattern of temperament, behavior, and relation to others, which is usually in place by teenage years, as a consequence of which, either the person or their society suffers. It is a controversial topic. Some might argue that it constitutes the medicalization of ‘deviance.’ Some might exclude it from psychiatry altogether. Currently, the UK government places great emphasis upon attempts to treat those with what it calls ‘dangerous and severe personality disorders.’ Evidence of response to treatment is limited. Table 10 shows those personality disorders currently recognized by the prominent diagnostic systems. They may be ‘clustered’ into those people
who seem cut off from others (Cluster A), those who have tempestuous relationships with them (Cluster B), and those for whom relationships are unsatisfactory, emotionally (Cluster C). This is a rough and ready categorization, but it is interesting to note that those personalities with whom psychiatrists probably have the most contact are those harming themselves (borderline personality disorder, F60.3) and others (antisocial personality disorder, F69.2); both located in Cluster B. Though other personalities may suffer, they tend not to present to psychiatric care as frequently.
Difficult Situations A recurring theme of this article has been the central problem in psychopathology, which is aligned with the ‘problem’ of consciousness itself: we do not know what others are thinking. Just as scientists and philosophers cannot explain how a mechanistic, and apparently deterministic, universe generates awareness, or devise a means by which one person could truly know what it is like to be another (or indeed a bat!), thereby limiting discourse to an attempted correlation between ‘third-person’ observations (of an object) and ‘first-person’ experiences (of a subject), so also psychopathology attempts to describe and diagnose what it is that someone else is experiencing. In this regard, as I have stated already (above) there is something distinguishing psychiatry from other branches of medicine: we are uniquely concerned with the phenomenological, not merely as a signpost on the way to reaching a diagnosis, but as our primary concern: it is, after all, the mental state that we are attempting to ‘treat’: we aim to change phenomenology. Admittedly, we have the ‘clues’ offered by the subject’s speech and behavior, but it is obvious that we operate at several removes (and indeed dimensions) from the object of our inquiry, (We are a first-person subject cogitating upon our thirdperson observations of another first-person subject, possibly supplemented by the evidence of other subjects in our midst). Given that most psychiatric diagnoses describe what people think and the reasons for their actions this represents
Psychopathology and Consciousness Table 10
Clinically encountered personality disorders
Disorder Cluster A Paranoid Schizoid Schizotypal Cluster B Antisocial Borderline
Histrionic Narcissistic Cluster C Obsessive–compulsive (anankastic) Avoidant Dependent
269
Features Suspicious of others, sensitive to slights, often feels ‘put down’; bears grudges. Suffers as a consequence of perceived exclusion. Cold affect, prefers own company, lives in ‘own world’ mentally, insensitive to praise or criticism. Not thought to suffer from isolation. Controversial concept, possible mild variant of schizophrenia; subject may be particularly interested in occult, UFOs, alternative or eccentric lifestyles, prone to illusions, magical thinking, unusual patterns of speech, may seem cold and aloof. Recurrent breaker of society’s rules, impulsive, reactive aggression, lacks ‘sense of responsibility,’ has difficulty sustaining relationships with others. At severe end, psychopaths may engage in instrumental violence, sadism, lying, cheating, and lack remorse for conduct or empathy for victims. Unstable sense of self, frequent decompensations, often connected with relationships, prone to feelings of abandonment, recurrent self-harm, may experience short bouts of very low mood, suicidal ideation, low self-esteem, confusion over sexuality, problems with impulse control: binge eating, use of drugs and alcohol, stealing, promiscuity, gambling. Superficial expression and experience of emotions (‘theatricality’), but appears shallow, needs to be the center of attention, given to displays of decompensation. Suggestible, preoccupied with physical attractiveness. Feels greater self-worth than others, sense of entitlement, resentful of others, fantasizes about success, may lead ‘Billy Liar’ lifestyle, spending money he does not have, pretending to be a very important personage. Lacks empathy and uses others. Rigid character, stubborn, pedantic, preoccupation with perfectionism that limits completion of tasks, needs routine, unable to adapt to change. May appear ‘sticky’ in relationships and thinking (things must be done ‘his way’), unable to ‘let things go.’ Sensitive to criticism, avoids social contact out of fear of rejection. Suffers as a consequence. Anxious. Limits contact to those where ‘sure to be liked.’ Overly reliant upon others, requires much support. Afraid of being left alone, avoids critique of those relied upon, has others make decisions.
something of an impasse. We are always operating at the level of inference. It follows then that there may be several ways in which this system of psychopathological communication may break down. Some are obvious: we may speak different languages, come from different cultures, and carry different assumptions about the world. We may be ‘there’ not because we want to be, but because someone else has requested the meeting: the benefits agency, the insurance company, the custody sergeant, the duty doctor, the prison staff, the defense counsel, or the judge. There may be many reasons why the subject doubts or mistrusts the psychopathologist. Therefore, there is always a risk of misunderstanding and of deception. Indeed, the frailties of psychiatric diagnoses were exposed by the classic experiment of D.L. Rosenhan who, with others, gained admission to psychiatric hospitals in the United States in the 1970s, while claiming to ‘hear voices’ (they did not hear them, they were well).
They concluded that there was a problem with psychiatric diagnosis, because they were believed, uncritically. More recently, Peter Tyrer and colleagues have been concerned by what they term ‘instrumental psychosis,’ the apparent relapse into psychoses of people known to have had such illnesses in the past, but for whom the current ‘episode’ is, what one might call, ‘convenient,’ for example, it circumvents a court appearance. Is there anything surprising about any of this? No, not really. It is the price that we pay for attempting to understand phenomena that only a ‘subject’ can ever truly know. Similar issues are played out in casualty departments at night, whenever a duty psychiatrist attempts to discern whether another human being ‘really’ intended to ‘kill herself ’: whether she knew that she would be found, whether she knew how many tablets might have proven fatal, whether there was an argument, whether she told anyone what she was thinking of doing, who it was that
270
Psychopathology and Consciousness
called the ambulance, whether there was a note and what it said, did she feel remorse? We have to infer the answers, but we must also be humane. More important than being strictly accurate or ‘intellectually correct’ is the desire to save life; to admit a patient ‘unnecessarily,’ to ‘err on the safe side,’ might not concur with what we actually believe did happen (or what we are encouraged to do by health service managers) but we are there to save a life, not to conduct an experiment. We played out these issues also in the discussion of the word ‘functional,’ where we pointed to the difficulty in determining whether physical states, behaviors that resemble illnesses, but for which there is no discernible ‘organic’ cause, are attributable to the unconscious processes hypothesized in hysteria or the conscious intention of a subject who is deceiving us, malingering. We may try to detect the difference, while diagnostic systems maintain such distinctions, but it is clear that objective proof of one state above another is largely hypothetical. So, is there such a thing as ‘silent schizophrenia’? Only if psychopathology ceases to be about mental states, about what a consciousness perceives. Until then, schizophrenia is an experience, experienced by a subject, the subject whom we meet, in the room at the clinical interview. See also: Phenomenology of Consciousness; Psychodynamic Theories of the Unconscious.
Suggested Readings Amador X and David A (eds.) (2004) Insight and Psychosis: Awareness of Illness in Schizophrenia and Related Disorders. Oxford: Oxford University Press. Berrios GE (1996) The History of Mental Symptoms: Descriptive Psychopathology Since the Nineteenth Century. Cambridge: Cambridge University Press.
Breuer J and Freud S (1974) Studies on Hysteria, Strachey J (trans.), The Penguin Freud Library, vol. 3. London: Penguin. Cutting J (1987) The phenomenology of acute organic psychosis: Comparison with acute schizophrenia. British Journal of Psychiatry 151: 324–332. Enoch D and Trethowan W (1991) Uncommon Psychiatric Syndromes, 3rd edn. Oxford: Butterworth-Heinemann Ltd. Evans D (2001) Emotion: A Very Short Introduction. Oxford: Oxford University Press. Gelder M, Mayou R, and Cowen P (eds.) (2001) Shorter Oxford Textbook of Psychiatry, 4th edn. Oxford: Oxford University Press. Hare RD (1999) Without Conscience: The Disturbing World of the Psychopaths Among Us. New York: Guilford Press. Jaspers K (1963) General Psychopathology. Hoenig J and Hamilton MW (trans.). Manchester: Manchester University Press. Leudar I and Thomas P (2000) Voices of Reason, Voices of Insanity: Studies of Verbal Hallucinations. London: Routledge. Pawar A and Spence SA (2003) Defining thought broadcast: Semi-structured literature review. British Journal of Psychiatry 183: 287–291. McKenna P and Oh T (2005) Schizophrenia Speech: Making Sense of Bathroots and Ponds that Fall in Doorways. Cambridge: Cambridge University Press. Mullins S and Spence SA (2003) Re-examining thought insertion: Semi-structured literature review and conceptual analysis. British Journal of Psychiatry 182: 293–298. Rosenhan DL (1973) On being sane in insane places. Science 179: 250–258. Spence SA (1999) Hysterical paralyses as disorders of action. Cognitive Neuropsychiatry 4: 203–226. Spence SA (2006) All in the mind? The neural correlates of unexplained physical symptoms. Advances in Psychiatric Treatment 12: 349–358. Spence SA, Crimlisk HL, Cope H, Ron MA, and Grasby PM (2000) Discrete neurophysiological correlates in prefrontal cortex during hysterical and feigned disorder of movement. Lancet 355: 1243–1244. Spence SA and Halligan PW (eds.) (2002) Pathologies of Body, Self and Space. Hove: Psychology Press. Tyrer P, Babidge N, Emmanuel J, Yarger N, and Ranger M (2001) Instrumental psychosis: The Good Soldier Svejk syndrome. Journal of the Royal Society of Medicine 94: 22–25. World Health Organization (1992) The ICD-10 Classification of Mental and Behavioural Disorders. Geneva: World Health Organization.
Psychopathology and Consciousness
271
Biographical Sketch
Sean A Spence is a professor of general adult psychiatry at the University of Sheffield, where he holds an MRC Career Establishment Grant. He was previously De Witt-Wallace visiting research fellow at Cornell University, New York (1999), and MRC clinical training fellow at the MRC Cyclotron Unit, Hammersmith Hospital (1995–98). He has won the Royal College of Psychiatrists Research Prize and Medal (1997) and Royal Society of Medicine Section of Psychiatry Essay Prize (1997). He studied medicine at Guy’s Hospital, London, acquired an intercalated BSc in psychology, and won the Gillespie Prize in psychological medicine (1986). His principal research interest is in the regulation of voluntary behavior (volition) in healthy subjects and those affected by neuropsychiatric disease. He is a coauthor of Managing Negative Symptoms of Schizophrenia (Science Press, 2001), a coeditor of Pathologies of Body, Self and Space (Psychology Press, 2002), and of Voices in the Brain: The Cognitive Neuropsychiatry of Auditory Verbal Hallucinations (Psychology Press, 2004). In 2001 the 2nd Reunion de investigacion en psicopatologia, in Cadiz, was devoted to his volitional studies. In 2007, a television series entitled ‘Lie Lab’ was devoted to his work on fMRI of deception (Channel 4 UK). His clinical work is with the homeless.
Religious Experience: Psychology and Neurology H Roggenkamp, M R Waldman, and A B Newberg, University of Pennsylvania, Philadelphia, PA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Autonomic nervous system – A part of the nervous system that connects the brain to the body and regulates various organ systems to prepare for an arousal or calming response. Deafferentation – Blocking input, usually neuronal, into one or more structures in the brain. Electroencephalography – A method for measuring the electrical changes that occur in the brain. Functional magnetic resonance imaging (fMRI) – A technique for performing scans to show activation of different parts of the brain using magnetic fields rather than radioactivity. Neurochemicals/Neurotransmitters – Refers to different substances such as serotonin and dopamine that are chemicals or molecules that are required for the functioning of the brain. Phenomenology – Relating to the subjective experiences of the world that people have. Positron emission tomography (PET) – A technique for performing scans of various physiological processes in the brain that requires initial administration of a radioactive material, similar to single photon emission computed tomography (SPECT). Posterior superior parietal lobe – An area of the brain located toward the back (posterior) and top (superior) that is involved in spatial orientation of the body. Prefrontal cortex – Part of the brain’s frontal lobe that is highly involved in focusing attention and executive functions. Single photon emission computed tomography (SPECT) – A technique for
performing scans of various physiological processes in the brain that requires initial administration of a radioactive material, similar to positron emission tomography (PET). Thalamus – A key central structure in the brain that is a crucial relay between sensory organs and the brain and between different brain structures to each other.
Introduction Spiritual Experience as a Means for Studying Consciousness Religion has served as a vital influence on the lives of humans since the beginning of modern man and a plethora of practices incorporate people’s ongoing quest to be closer to what they perceive as God. Religious experiences, particularly those that are transformative, have often been described as involving an alteration in the usual sense of consciousness. The study of religious and spiritual experiences is ultimately the study of complex mental processes. Thus, the study of religious and spiritual experiences is also potentially one of the most important areas of research that may be pursued by science in the next decade. This may not be an understatement since such experiences offer a fascinating window into human consciousness and psychology, the relationship between mental states and body physiology, emotional and cognitive processing, and the biological correlates of religious and spiritual experiences. At one end of the continuum of religious experiences is the state that might be called ‘Absolute Unitary Being’ which is described in the mystical literature of all the world’s great religions. When a person is in that state he or she loses all sense of
273
274
Religious Experience: Psychology and Neurology
discrete being, even to the point where the difference between self and other is obliterated. There is no sense of the passing of time, and all that remains is a perfect timeless undifferentiated consciousness. When such a state is suffused with positive affect, there is a tendency to describe the experience, after the fact, as personal. Such experiences are often described as a perfect union with God (the ‘Unio mystica’ of the Christian tradition) or the perfect manifestation of God (the attainment of ‘moksha’ in the Hindu tradition). When such experiences are accompanied by neutral affect they tend to be described, after the fact, as impersonal. These states are described in concepts such as the Void or Nirvana of Buddhism or the Absolute of a number of philosophical/mystical traditions. There is no question that whether the experience is interpreted personally as God or impersonally as the Absolute, it nevertheless possesses a quality of altered consciousness – a transcendent wholeness without any temporal or spatial division. There seem to be neurobiological correlates of such states of consciousness, such as the inhibition of sensory input into the posterior superior parietal lobe (PSPL), especially on the right. This area of the brain is responsible for the orientation of objects in three-dimensional space. If it is denied of all sensory input, as a result of mechanisms generated during practices such as profound meditation, the result may be a sense of pure space. Since space has no objective reality (unless it relates things to each other), the subjective experience is one of total spacelessness, or of total perfect unity. Research on brain areas involved during states of absolute unity in the context of meditation will be discussed in more detail later in the article. Typically, some form of meditation or concentration is used to bring the individual into this altered state of consciousness. Although there are many styles and forms of meditation, we have typically divided such practices into three basic categories. The first category is one in which the subjects simply attempt to clear all thought from their sphere of attention. This form of meditation is an attempt to reach a subjective state characterized by a sense of no space, no time, and no thought. Further, this state is cognitively
experienced as fully integrated and unified such that there is no sense of a self and other. The second category is one in which the subjects focus their attention on a particular object, image, phrase, or word, and includes practices such as transcendental meditation (TM) and various forms of Tibetan Buddhism. This form of meditation is designed to lead to a subjective experience of absorption with the object of focus. The third category of meditation can be referred to as mindfulness, in which the person observes without judgment the uninterrupted flow of thoughts and feelings as they consciously rise and fall in the mind. This form of meditation creates what is known as an ‘observing self,’ which allows the person to feel emotionally neutral. These three forms of meditation tend to overlap, as concentrative practices can also incorporate mindfulness and clearing of thoughts. Meditation can be practiced in many different conditions. Qigong, an ancient Chinese meditation practice, coordinates breathing and focusing on certain parts of the body, creating a very dynamic physical form of meditation. Qigong is practiced sitting, standing, and during movement, which makes it similar to Yogic meditation. Speaking in tongues (glossolalia) is a religious experience that often involves singing and dancing, further extending the range of activities that can be undertaken to induce altered states of consciousness associated with religious experience. There is another distinction in which some meditation is guided by following along with a leader who is verbally directing the practitioner either in person or on tape. Others practice the meditation on their own volition. This difference can be seen in specific differences in cerebral activation while practicing volitional or guided meditation. Despite differences in practice methodology, phenomenological analysis suggests that the end result of many religious and spiritual practices is similar. However, this result might be described differently using unique characteristics depending on the culture and individual. Therefore, it seems reasonable that while the initial neurophysiological activation occurring during any given practice may differ, there should eventually be a convergence.
Religious Experience: Psychology and Neurology
Consciousness and Religious Experience can be Explored via Brain Imaging and Activation Studies Relevant Imaging and Activation Modalities For any study of human consciousness, one would have to begin by defining the specific operationalized paradigm through which the experience itself would be explored. Such studies can range from the evaluation of various sensory experiences to determine when, where, and how such experiences enter consciousness. Religious practices such as meditation, which are specifically designed to alter consciousness, can be explored scientifically as well. Different states of consciousness can theoretically be studied using various brain imaging techniques. Depending on what elements of consciousness are the focus of a given study, each of the functional brain imaging techniques provides different methodological advantages and disadvantages. The currently available functional brain imaging techniques include functional magnetic resonance imaging (fMRI), single photon emission computed tomography (SPECT), and positron emission tomography (PET). fMRI requires a large magnet in order to measure changes in cerebral blood flow (CBF) and oxygen content. SPECT and PET require the injection of various types of radioactive tracers that can measure a variety of neurophysiological functions including blood flow, metabolism, and neurotransmitter changes. fMRI has improved spatial resolution over SPECT, and probably PET, as well as the ability to perform immediate anatomic correlation. Thus, functional MRI can produce functional images that can be directly mapped onto the anatomical structures, thereby helping to pinpoint areas of the brain associated with various activation tasks (PET and SPECTcan also be co-registered with structural MRI images, but this often involves complex computer programs to overlay the functional image with the anatomical ones). Another advantage of fMRI is the excellent temporal resolution, so that multiple scans can be made over the course of a single imaging session. However, the major disadvantage of fMRI is that the individual is required to lie down in a confined space and it makes a significant amount of noise, both of which could be very disturbing to
275
various types of practices such as meditation or could interfere with other religious practices that require specific movements or dancing. PET imaging provides better resolution than SPECT, and has the important ability to make quantitative measures of activity. This could be very important since certain tasks or experiences may either diminish or augment activity in some areas and not others, and this may be missed if absolute quantitation is not performed. However, if one strives to make the environment relatively distraction free, it is sometimes beneficial to perform these studies after hours, which may complicate the use of PET because radiopharmaceuticals such as fluorodeoxyglucose (FDG) may not be readily available. Furthermore, PET is the most expensive of the imaging modalities. One other potential problem is that the most common tracer used in PET imaging, FDG, has an uptake period of approximately 20–30 min, which may prevent the ability to capture a momentary state. SPECT imaging is the most readily available and rivals MRI in low cost per study and also has the advantage, as well as PET, in being able to study subjects outside of the scanner. For example, SPECT studies can enable the subjects to meditate or pray until they experience a ‘peak’ in their practice. At this point, the subject can be injected with a radioactive tracer through an indwelling intravenous catheter while they continue to practice. The tracer is fixed in the brain at the time of injection so that when the images are acquired approximately 20 min later, the images reflect the CBF during ‘peak’ meditation. Different states can then be compared. These studies have helped toward the development of complex neurobiological models of practices such as meditation, which can dramatically alter a person’s consciousness. Current models include not only a variety of structures including the prefrontal cortex (PFC), parietal lobe, limbic system, thalamus, and hypothalamus, but now include neurotransmitters such as glutamate, gamma aminobutyric acid (GABA), dopamine, and serotonin, as well as hormonal, autonomic, and immune changes (see Table 1 – neurochemical table). Electroencephalography (EEG) is another method for evaluating the brain’s activity, although
276
Religious Experience: Psychology and Neurology
Neurochemically related changes in serum concentration observed during meditation techniques and the central nervous system (CNS) structures typically involved in their production
Table 1
Neurochemical
Observed change
CNS structure
Arginine vasopressin GABA
Increased
Supraoptic nucleus
Increased
Melatonin Serotonin Cortisol
Increased Increased Decreased
Norepinephrine b-Endorphin
Decreased Rhythm changed; levels unaltered
Thalamus, other inhibitory structures Pineal gland Dorsal raphe Paraventricular nucleus Locus ceruleus Arcuate nucleus
it is not typically considered an imaging technique. EEG reflects cortical activity through recording electrical signals via electrodes placed on the scalp, and primarily measures neuronal activity immediately beneath the cortex. Alpha rhythm waves are 8–12 Hz frequency signals that are related to relaxation and lack of active cognitive processes. Beta wave signals, characterized by rhythms from 18 to 30 Hz, occur when one is alert and focused; frequencies from 30 to 70 Hz are termed gamma activity. Gamma waves seem to be present when the brain is working to integrate a variety of stimuli into a coherent whole and during peak performance. Between the ranges of 4 and 8 Hz one sees theta activity, which is correlated with rapid eye movement sleep, problem solving, overall attention, and hypnosis. Delta activity is seen especially in infants in the first two years of life, and is a low frequency wave of 0.5–4 Hz. It tends to be associated with sleep as well. EEG recordings reflect the difference of voltage between the signals at two electrodes. Artifacts can confound EEG results very easily: EEG voltages are extremely small, and must be amplified by a factor of one million in order to be measured. Therefore, it is easy to mistake electrical interference that originates from the electrodes, the recorder, or electric lights, as activity in the brain. EEG is best used when paired with neuroimaging modalities such as PET, as it cannot measure activity in deeper areas of the brain such as the brainstem; it primarily shows activity of pyramidal neurons of the cortex.
Several newer methods are also being developed such as optical imaging or near infrared imaging in which a light probe is placed on the skin surface and the change in light received by an adjacent detector reflects changes in the brain’s blood flow. This technique has excellent temporal and good spatial resolution and is very portable. Its main disadvantage is that it cannot observe deep structures. Problems with Relating Neurobiology to Religious Experiences There are several generic methodological issues with brain imaging studies of various elements of conscious experience. The most problematic is not so much what is measured on the scan, but whether the scan corresponds to the subjective state that one is trying to measure. In the example of studying religious experiences, it typically is not possible to ‘interrupt’ the practitioner during the practice, and therefore it can never be known if the scan corresponds to a specific subjective state of consciousness or a specific experience. Studies that attempt to measure changes in the brain associated with sensorial experience presenting in consciousness require the subjects to indicate when the experience does in fact enter into their consciousness. The problem here is that there is a necessary delay between when they experience something and when they actually respond. Furthermore, the response itself can alter brain physiology. Thus, neurobiological research into the subjective nature of consciousness is quite difficult and any findings need to be interpreted carefully. On the other hand, neuroimaging techniques continue to make major advances and provide the best window into the underlying neurobiological processes associated with consciousness. Besides practical questions regarding imaging modalities, there are additional concerns regarding inferences about brain activation data and religious experience. Up to the present, consciousness has been understood as referring to consciousness of something. The existence of a state like pure consciousness or absolute unity, consciousness devoid of content which meditators often report experiencing, has generally not even been entertained as a problem. Such mystical experiences are devoid of the perception of
Religious Experience: Psychology and Neurology
discrete reality, and have no sense of the passage of time, no sense of the extension of space, and no sense of the self-other dichotomy. What is particularly interesting about the state is that neither during the experiencing of pure consciousness, nor upon subsequent recollection, is this state ever perceived as subjective. Although it is attained by going deeply within the subject, once it is attained, it is perceived as neither subjective nor objective. In other words, the state consists of an absolute sense of unity without thought, without words, without sensation, and not even being sensed to inhere in a subject. Such a state may have a specific neurobiological correlate, supported by preliminary brain imaging research of meditative
states, but more data is necessary to explore the implications of this state more deeply. Further, it may still remain to be seen whether the biological correlates explain the experience or are merely evaluating how the brain responds to an altered consciousness state.
Brain Activation in Religious Experiences The following review is presented as a working hypothesis regarding the brain’s function during various religious and spiritual practices and experiences. This hypothesis incorporates recent Right PFC
Arcuate nucleus of the hypothalamus
Left PFC
+ BE
+ glutamate
+ glutamate
+ BE via arcuate nuc.
– NAALADase
+ NAAG
Glutamate
Nucleus accumbens
Left thalamic reticular nucleus
Right thalamic reticular nucleus
+ GABA
+ GABA
Left PSPL
Lateral geniculate lat. posterior nucl.
Right PSPL
Glutamate
Striatum
Glutamate Right hippocampus
Left hippocampus
+ acetylcholine
Anterior cingulate
+ NAAG
NAALADase
Nucleus basalis
277
Glutamate Left amygdala
Glutamate Right amygdala
Glutamate Lateral hypothalamus
Breakthrough
D. Raphe
Pineal
+ ST
+ MT
Glutamate Ventromedial hypothalamus
Parasympathetic
GABA
Depamine Substantia nigra GABA Ventral thalamus
Supraoptic nucleus + AVP
+ DMT Paragigan to cellular nucleus Sympathetic
+ Increase – Decrease Stimulate Inhibit
Locus ceruleus
Paraventricular nucleus
– NE – Cortisol
Figure 1 Schematic overview of the neurophysiological network, possibly associated with meditative states and experiences.
278
Religious Experience: Psychology and Neurology
neuroimaging, neurochemical, hormonal, and physiological studies. The purpose of this review is to provide a foundation from which many different types of religious experiences and practices can be considered and compared. As shown in Figure 1, the schematic representation of the involved pathways begins with the PFC and suggests a number of complex interactions with the thalamus, PSPL, limbic system, and autonomic nervous system (ANS). It should be noted that not all structures that might be involved in religious and spiritual experiences are included, especially since new structures such as the insula and claustrum are just now being hypothesized to be involved. Furthermore, a number of both excitatory and inhibitory neurotransmitters can now be proposed to play a role in such practices and experiences. Dopamine, serotonin, acetylcholine, and several other molecules may be associated with various phenomenological aspects of such experiences and these are also considered in this model. It would be anticipated that depending upon the specific practice, the ritual, tradition, and individual involved in the specific mechanisms might be somewhat different. However, focusing on the phenomenology of such experiences, this model provides information regarding the diversity of experiences, both sensory, cognitive, and affective, that can be associated with religious and spiritual experiences. The model was initially developed utilizing information from studies primarily on meditative practices due to the relatively large amount of data available. However, this model can likely be applied to many different types of practices and experiences as they result in altered states of consciousness. Activation of the Prefrontal and Cingulate Cortex Since some approaches toward attaining spiritual experiences involve practices in which the person on their own volition attempts to enter into such a state, we will begin with the parts of the brain that might be associated with the human will. Brain imaging studies suggest that willful acts and tasks that require sustained attention are initiated via activity in the PFC, particularly in the right hemisphere. The cingulate gyrus has also been shown to be involved in focusing attention, probably
in conjunction with the PFC. Since transformative types of practices such as meditation require intense focus of attention, it seems that there should be activation of the PFC as well as the cingulate gyrus. This notion is supported by the increased activity observed in these regions on several of the brain imaging studies of volitional types of meditation including that from our laboratory. In a study of eight Tibetan Buddhist meditators, the subjects had an intravenous line placed and were injected with a CBF tracer while at rest, in order to acquire a ‘baseline’ image. They then meditated for approximately 1 h when they were again injected with the tracer while they continued to meditate. The tracer was fixed in the brain at the time of injection so that when the images were acquired approximately 20 min later, they reflected the CBF during the meditation state. These images demonstrated increased activity in the PFC bilaterally (greater on the right) and the cingulate gyrus during meditation. Therefore, meditation appears to start by activating the prefrontal and cingulate cortex associated with the will or intent to clear the mind of thoughts or to focus on an object. Studies of other forms of meditation, such as Yoga Nidra, Kundalini Yoga, TM, and Christian prayer show an increase in activation in the frontal and prefrontal cortices as well. Of course, meditation does not necessarily lead to transformation, so that frontal lobe activity alone does not seem to provide a complete understanding of these experiences. In fact, some transformative experiences are associated with feelings in which the self feels ‘overtaken’ by the experience itself. The self feels as if it becomes absorbed or lost into the experience. If the frontal lobe is involved in a willful activity, then a transformative experience associated with a feeling of losing the self or of the original self being overcome by something spiritual, might be associated with a lack of activity in the frontal lobes. A decrease in frontal lobe activity is seen in Pentecostal Christian glossolalia practitioners, a practice involving speaking in tongues. It would seem that glosssolalia practitioners may not be focusing their attention when speaking in tongues; they report feeling that they can no longer control what is happening. More research is needed that compares different practices and their effects on frontal lobe activity.
Religious Experience: Psychology and Neurology
Thalamic Activation Several animal studies have shown that the PFC, when activated, innervates the reticular nucleus of the thalamus, particularly as part of a more global attentional network. Such activation may be accomplished by the PFC’s production and distribution of the excitatory neurotransmitter glutamate which the PFC neurons use to communicate among themselves and to innervate other brain structures. The thalamus itself governs the flow of sensory information to cortical processing areas via its interactions with the lateral geniculate and lateral posterior nuclei and also likely uses the glutamate system in order to activate neurons in other structures. The lateral geniculate nucleus receives raw visual data from the optic tract and routes it to the striate cortex for processing. The lateral posterior nucleus of the thalamus provides the posterior superior parietal lobule (PSPL) with the sensory information it needs to determine the body’s spatial orientation. When excited, the reticular nucleus releases the inhibitory neurotransmitter GABA onto the lateral posterior and geniculate nuclei, cutting off input to the PSPL and visual centers in proportion to the reticular activation. During practices such as meditation, because of the increased activity in the PFC, particularly on the right, there should be a concomitant increase in the activity in the reticular nucleus of the thalamus. While brain imaging studies of meditation have not had the resolution to distinguish the reticular nuclei, our recent SPECT study did demonstrate a general increase in thalamic activity that was proportional to the activity levels in the PFC. This is consistent with, but does not confirm the specific interaction between the PFC and reticular nuclei. If the activation of the right PFC causes increased activity in the reticular nucleus during meditation, the result may be decreased sensory input entering into the PSPL. Several studies have demonstrated an increase in serum GABA during meditation, possibly reflecting increased central GABA activity. This functional deafferentation related to increased GABA would mean that fewer distracting outside stimuli would arrive at the visual cortex and PSPL enhancing the sense of focus. During meditative experiences, similar mechanisms might be in place
279
although the interaction between the frontal lobe and the thalamus might be more complex such that some states may be associated with mutual increases, some with mutual decreases, and some with one turned on and the other off. The mechanism of the later is less clear at this time. It should also be noted that the dopaminergic system, via the basal ganglia, is believed to participate in regulating the glutamatergic system and the interactions between the PFC and subcortical structures. A recent PET study utilizing 11C-Raclopride to measure the dopaminergic tone during Yoga Nidra meditation demonstrated a significant increase in dopamine levels during the meditation practice. They hypothesized that this increase may be associated with the gating of cortical-subcortical interactions that leads to an overall decrease in readiness for action that is associated with this particular type of meditation. Similarly, there might eventually be a reciprocal decrease in PFC activity associated with the inhibitory function of the thalamus. This decrease may effectively take the frontal lobes ‘off line’ and result in the sense of losing control. It is also well known that the dopaminergic system is involved in euphoric states and thus may be associated with transformative experiences of high emotional states. Future studies will be necessary to elaborate on the role of dopamine in spiritual transformation as well as the interactions between dopamine and other neurotransmitter systems. Posterior Superior Parietal Lobe Deafferentation The PSPL is heavily involved in the analysis and integration of higher-order visual, auditory, and somaesthetic information. It is also involved in a complex attentional network that includes the PFC and thalamus. Through the reception of auditory and visual input from the thalamus, the PSPL is able to help generate a three-dimensional image of the body in space, provide a sense of spatial coordinates in which the body is oriented, help distinguish between objects, and exert influences in regard to objects that may be directly grasped and manipulated. These functions of the PSPL might be critical for distinguishing between the self and the external world. It should be noted
280
Religious Experience: Psychology and Neurology
that recent studies have suggested that the superior temporal lobe may play a more important role in body spatial representation, although this has not been confirmed by other reports. Regardless of what the actual relationship between the parietal and temporal lobes is, in terms of spatial representation, these areas are likely to play a key role in the altered sense of self and space associated with these unusual states of consciousness. If, for example, there is deafferentation of the PSPL by the reticular nucleus’s GABAergic effects, the person may begin to lose their usual ability to spatially define the self and help to orient the self. This is an experience frequently associated with meditative practices as well as with transformative types of experiences. Deafferentation of the PSPL has also been supported by several imaging studies demonstrating decreased activity in this region during intense meditation. Hippocampal and Amygdalar Activation In addition to the complex cortical-thalamic activity, religious experiences might also be associated with altered activity in the limbic system, especially since stimulation of limbic structures is associated with profound emotional responses. The hippocampus acts to modulate and moderate cortical arousal and responsiveness, via rich and extensive interconnections with the PFC, other neocortical areas, the amygdala, and the hypothalamus. Hippocampal stimulation has been shown to diminish cortical responsiveness and arousal; however, if cortical arousal is initially at a low level, then hippocampal stimulation tends to augment cortical activity. The ability of the hippocampus to stimulate or inhibit neuronal activity in other structures likely relies upon the glutamate and GABA systems respectively. It has been previously suggested that the blocking of sensory information (i.e., deafferentation) of the PSPL might be associated with the loss of the sense of self and/or the orientation of that self to other objects in the world. Since this loss of orientation of the self is often described in religious experiences, the deafferentation of the PSPL might be an important mechanism underlying such experiences. If partial deafferentation of the right PSPL occurs during a transformative experience,
the result may be stimulation of the right hippocampus, because of the inverse modulation of the hippocampus in relation to cortical activity. If, in addition, there is simultaneous direct stimulation of the right hippocampus via the thalamus (as part of the known attentional network) and mediated by glutamate, then a powerful recruitment of stimulation of the right hippocampus could occur. Right hippocampal activity may ultimately enhance the stimulatory function of the PFC on the thalamus via the nucleus accumbens, which gates the neural input from the PFC to the thalamus via the neuromodulatory effects of dopamine. The hippocampus greatly influences the amygdala, such that they complement and interact in the generation of attention, emotion, and certain types of imagery. It seems that much of the prefrontal modulation of emotion is via the hippocampus and its connections with the amygdala. Because of this reciprocal interaction between the amygdala and hippocampus, the activation of the right hippocampus likely stimulates the right lateral amygdala as well. The results of the fMRI study by Lazar et al. support the notion of increased activity in the regions of the amygdala and hippocampus during practices such as meditation. However, whether such functional changes are associated with religious experiences remains to be seen.
Language Areas An increase in the inferior parietal region has been reported in Franciscan nuns during their focus on a bible passage, which is an area commonly associated with language functioning. This type of activation is lacking in meditative practices that do not involve words. One example is Tibetan Buddhists’ focus on a visual stimulus, which is unlike the nun’s verbal focus. Interestingly, Pentecostal Christians speaking in tongues showed no increase in the inferior parietal region and a decrease in frontal lobe activity, despite the fact that they ‘‘speak’’ during their practice. This finding makes sense when considering that linguistic analyses of glossolalia find that it does not correspond to any linguistic structure. It would seem that the process of speaking in tongues is distinct from speaking a language. More research
Religious Experience: Psychology and Neurology
of this practice and other verbal practices such as Sufism is needed. Hypothalamic and Autonomic Nervous System Changes The hypothalamus is extensively interconnected with the limbic system. Stimulation of the right lateral amygdala has been shown to result in stimulation of the ventromedial portion of the hypothalamus with subsequent stimulation of the peripheral parasympathetic system. Increased parasympathetic activity should be associated with the subjective sensation first of relaxation, and eventually, of a more profound quiescence. Activation of the parasympathetic system would also cause a reduction in heart rate and respiratory rate. All of these physiological responses have been observed during meditation. Religious and spiritual experiences can also have a number of strong emotional responses ranging from bliss to ecstasy. Blissful elements of such experiences may be mediated in part by activity in the parasympathetic system. Typically, when breathing and heart rate slow down, the paragigantocellular nucleus of the medulla ceases to innervate the locus ceruleus (LC) of the pons. The LC produces and distributes norepinephrine (NE), a neuromodulator that increases the susceptibility of brain regions to sensory input by amplifying strong stimuli, while simultaneously gating out weaker activations and cellular ‘noise’ that fall below the activation threshold. Decreased stimulation of the LC results in a decrease in the level of NE. The breakdown products of catecholamines such as NE and epinephrine have generally been found to be reduced in the urine and plasma during meditation, which may simply reflect the systemic change in autonomic balance. However, it is not inconsistent with a cerebral decrease in NE levels as well. During a meditative practice, the reduced firing of the paragigantocellular nucleus probably cuts back its innervation of the LC, which densely and specifically supplies the PSPL and the lateral posterior nucleus with NE. Thus, a reduction in NE would decrease the impact of sensory input on the PSPL, contributing to its deafferentation. Thus, religious and spiritual experiences associated with an altered state of consciousness could be facilitated by such neuronal interactions.
281
The LC would also deliver less NE to the hypothalamic paraventricular nucleus. The paraventricular nucleus of the hypothalamus typically secretes corticotropin-releasing hormone (CRH) in response to innervation by NE from the LC. This CRH stimulates the anterior pituitary to release adrenocorticotropic hormone (ACTH). ACTH, which in turn stimulates the adrenal cortex to produce cortisol, one of the body’s stress hormones. Decreasing NE from the LC during meditation would likely decrease the production of CRH by the paraventricular nucleus and ultimately decrease cortisol levels. A drop in blood pressure associated with parasympathetic activity would be expected to relax the arterial baroreceptors leading the caudal ventral medulla to decrease its GABAergic inhibition of the supraoptic nucleus of the hypothalamus. This lack of inhibition can provoke the supraoptic nucleus to release the vasoconstrictor arginine vasopressin (AVP), thereby tightening the arteries and returning blood pressure to normal. AVP has also been shown to contribute to the general maintenance of positive affect, decrease self-perceived fatigue and arousal, and significantly improve the consolidation of new memories and learning. During meditative practice, plasma AVP has been shown to increase dramatically, although it is not known whether there will be a similar increase associated with more transformative experiences. AVP could help to enhance the memory of a particular experience, perhaps explaining the subjective phenomenon that meditative experiences are remembered and described in very vivid terms. Prefrontal Cortex Effects on Other Neurochemical Systems As PFC activity increases, it produces increasing levels of free synaptic glutamate in the brain. Increased glutamate can stimulate the hypothalamic arcuate nucleus to release b-endorphin (BE). BE is an opioid produced primarily by the arcuate nucleus of the medial hypothalamus and distributed to the brain’s subcortical areas. BE is known to depress respiration, reduce fear, reduce pain, and produce sensations of joy and euphoria. That such effects have been described during meditation may implicate some degree of BE
282
Religious Experience: Psychology and Neurology
release related to the increased PFC activity. Further, the joy and euphoric feelings associated with spiritual experiences might similarly implicate the endogenous opioid system. Meditation has been found to disrupt diurnal rhythms of BE and ACTH, while not affecting diurnal cortisol rhythms. Thus, the relationship between opioid receptors and various spiritual experiences is not clear, especially in light of one very limited study demonstrating that blocking the opiate receptors with naloxone did not affect the experience or EEG associated with meditation. Glutamate activates N-methyl D-aspartate receptors (NMDAr), but excess glutamate can kill these neurons through excitotoxic processes. It is possible that if glutamate levels approach excitotoxic concentrations during intense experiences, the brain might limit its production of N-acetylatedalpha-linked-acidic dipeptidase, which converts the endogenous NMDAr antagonist N-acetylaspartylglutamate (NAAG) into glutamate. The resultant increase in NAAG would protect the cells from excitotoxic damage. There is an important side effect, however, since the NMDAr inhibitor, NAAG, is functionally analogous to the disassociative hallucinogens ketamine, phencyclidine, and nitrous oxide. These NMDAr antagonists produce a variety of states that may be characterized as either schizophrenomimetic or mystical, such as out-of-body and near-death experiences. Whether such substances are elaborated within the brain during the short interval associated with a religious and spiritual experiences is not yet clear. Autonomic-Cortical Activity In the early 1970s, Gellhorn and Kiely developed a model of the physiological processes involved in meditation based almost exclusively on ANS activity, which while somewhat limited, indicated the importance of the ANS during such experiences. These authors suggested that intense stimulation of either the sympathetic or parasympathetic system, if continued, could ultimately result in simultaneous discharge of both systems (what might be considered a ‘breakthrough’ of the other system). We have suggested that this breakthrough is associated with the most intense, and potentially most transformative types of religious and spiritual
experiences. Several studies have demonstrated predominant parasympathetic activity during meditation, associated with decreased heart rate and blood pressure, decreased respiratory rate, and decreased oxygen metabolism. However, a recent study of two separate meditative techniques suggested a mutual activation of parasympathetic and sympathetic systems by demonstrating an increase in the variability of heart rate during meditation. The increased variation in heart rate was hypothesized to reflect activation of both arms of the ANS. Several other studies have also suggested that there may be some type of mutual excitation or rapid oscillation between the sympathetic and parasympathetic arms of the ANS during meditation practices. This notion also fits the characteristic description of meditative states in which there is a sense of overwhelming calmness as well as significant alertness. Also, the notion of mutual activation of both arms of the ANS is consistent with recent developments in the study of autonomic interactions. However, whether such complex interactions actually occur during meditative practices remains to be fully elucidated. Serotonergic Activity Activation of the ANS can result in intense stimulation of structures in the lateral hypothalamus and median forebrain bundle that are known to produce both ecstatic and blissful feelings when directly stimulated. Stimulation of the lateral hypothalamus can also result in changes in serotonergic activity. In fact, several studies have shown that after meditation, the breakdown products of serotonin in urine are significantly increased suggesting an overall elevation in serotonin during meditation. Serotonin is a neuromodulator that densely supplies the visual centers of the temporal lobe, where it strongly influences the flow of visual associations generated by this area. The cells of the dorsal raphe produce and distribute serotonin when innervated by the lateral hypothalamus (and also when activated by the PFC). Moderately increased levels of serotonin appear to correlate with the positive affect, while low serotonin often signifies depression. This relationship has clearly been demonstrated with regard to the effects of selective serotonin reuptake inhibitor medications, which are widely used for the treatment of depression.
Religious Experience: Psychology and Neurology
Increased serotonin levels can affect several other neurochemical systems. An increase in serotonin has a modulatory effect on dopamine, suggesting a link between the serotonergic and dopaminergic system that may enhance feelings of euphoria, frequently described during religious and spiritual states. Serotonin, in conjunction with the increased glutamate, has been shown to stimulate the nucleus basalis to release acetylcholine, which has important modulatory influences throughout the cortex. Increased acetylcholine in the frontal lobes has been shown to augment the attentional system, and in the parietal lobes to enhance orienting without altering the sensory input. Increased serotonin combined with lateral hypothalamic innervation of the pineal gland may lead the latter to increase production of the neurohormone melatonin from the conversion of serotonin. Melatonin has been shown to depress the central nervous system (CNS) and reduce pain sensitivity. During meditation, blood plasma melatonin has been found to increase sharply, which may contribute to feelings of calmness and decreased awareness of pain. EEG Studies of Spiritual Experiences EEG can be used to study neuroelectrical changes that occur due to various spiritual practices such as meditation or prayer. For example, over sixty studies have been performed since at least 1957 on the effects of meditation, based on EEG output; practices studied include Kriya Yoga, Transcendental Meditation (TM), Zen Buddhism, Tibetan Buddhism, and Kundalini Yoga. For many years this method was the only tool available to researchers for studying the activity of the mind – the first studies using functional neuroimaging modalities such as PET, SPECT, and fMRI to study meditation were performed in the 1990s. Thus, there is a large amount of data on brain wave changes and meditation, and in this data there is some consideration of the fact that no two practices are necessarily the same. Different practices can induce varying EEG results. Overall, when changes in alpha waves (8–12 Hz) have been described during meditation practices, the result is an increase in alpha wave power. In addition, this band is stronger at rest in meditators versus nonmeditators. Interestingly, the location of
283
increased alpha power varies across meditation practices. Advanced Qigong meditators have been shown to have increased alpha power only over the frontal cortex, with decreased alpha power over the occipital cortex. TM, Zen meditation, various yoga practices, and even relaxation training have all been observed to affect alpha waves. Alpha wave blocking can occur in meditation as well. In resting states, alpha power is reduced after eyes that were closed are opened, in which case the occipital cortex shows a decrease in alpha waves. This signifies going from a state of relaxation to one of cognitive processing. Therefore, there is a connection between alpha waves and cortical processing as alpha decreases are typically seen after a stimulus is administered. Studies from the 1960s examining Indian yogis while meditating showed no alpha blocking in response to stimuli such as placing their hands in cold water. A more recent study of TM meditators showed no changes in alpha blocking when musical tones were presented. These results suggest that meditation training enables one to maintain levels of relaxation and mental ‘emptiness’ despite stimuli that are meant to be disruptive. Increased theta activity (4–8 Hz) can be seen especially in advanced long-term meditators, but might be correlated with maintaining sustained attention, which is generated by the anterior cingulated and medial/dorsolateral prefrontal cortex, and may not necessarily be part of the spiritual experience. An increase in theta waves can be seen in TM, certain types of Yoga meditations, and Tibetan Buddhism, and is especially associated with advanced practitioners. When studying two different types of Qigong meditation it was found that those practicing the more concentrative type showed frontal midline theta activity, while the more passive, mindfulness based type did not illicit such brain activation in practitioners. It is further interesting to note the hypnosis is often associated with a similar increase in theta activity. Trances can be observed in religious rituals and in patients with psychiatric conditions. EEG recordings of a trance called Kerauhan, which is experienced by Indonesians participating in the Calonarang ritual, showed enhanced alpha and theta bands of spontaneous EEG activity. These trance results differ from results of individuals who undergo trances as a
284
Religious Experience: Psychology and Neurology
result of schizophrenic or dissociative disorders. Schizophrenic trance states tend to show a decrease in alpha rhythms. A Tibetan Buddhist practice called gTum-mo generates heat in the body and is characterized by increased beta activity; in addition, increases in finger and toe temperatures were seen at up to 8.3 C in this practice. More recent EEG studies of Tibetan Buddhist meditation have shown increases in gamma power, while a study of Zen meditators showed mostly decreased gamma coherence. Gamma activity may play a role in affective regulation, but more research is needed to elucidate this relationship more clearly. Cortical synchronization is another aspect of cortical physiological function that EEG can measure. While Sahaja yoga meditators and TM practitioners have been found to show less hemispheric lateralization when compared to controls, subjects trained in mindfulness meditation have shown an increase in right-sided alpha power while not meditating. Subjects in the previously mentioned Tibetan gTum-mo condition showed greater asymmetries during their practice, with an increase in right hemispheric activity. Therefore, no generalizations can be made regarding hemispheric lateralization and meditative practices. In addition, lateralization has been studied outside of the realm of meditation, as left hemispheric dominance has been associated with happier states and traits, and left and right hemisphere interactions are involved with the approach-withdrawal process.
Other Relevant Topics in Researching Religion and Spiritual Experiences The Role of Psychotropic Drugs in Spiritual Experiences The effects of dugs on consciousness are discussed in another article of this encyclopedia. However, drugs have been used to induce mystical states, and for the purposes of ritual and religious experience throughout human history. Several studies have linked dimethyl tryptamine (DMT) to a variety of mystical states, including out-of-body experiences, distortion of time and space, and interaction with supernatural entities. Hyperstimulation of the
pineal at this step, then, could also lead to DMT production that can be associated with a wide variety of mystical-type experiences associated with hallucinogens. It has been observed that under circumstances of heightened activation, pineal enzymes can also endogenously synthesize DMT. Another hallucinogen, D-lysergic diethylamide (LSD), is thought to work directly on the serotonergic system and use the connections that the system has, in order produce wide-ranging effects, including stimulation of the noradrenergic system. Tryptamine-based psychedelic drugs such as psilocybin and LSD bind to cortical serotonin (ST2) receptors (especially in the temporal lobes), which can result in a hallucinogenic effect. The mechanism by which this appears to occur is that serotonin inhibits the lateral geniculate nucleus, greatly reducing the amount of visual information that can pass through. If combined with reticular nucleus inhibition of the lateral geniculate, serotonin may increase the fluidity of temporal visual associations in the absence of sensory input, possibly resulting in the internally generated imagery that has been described during certain meditative states and ritual experiences. As mentioned, psilocybin has similar effects on the serotonergic system as LSD. In a study of 36 volunteers who participated in a double blind psilocybin study, 22 of them reported having a ‘complete’ mystical experience, while only four reported having a mystical experience with the control drug (methylphenidate). In addition, 67% of the volunteers rated the psilocybin experience as either the single most meaningful experience of their lives, or in their top five meaningful lifetime experiences. However, a recent PET study of the brain during psilocybin use was interpreted as a ‘psilocybin model of psychosis.’ Interestingly, many of the brain areas activated during psilocybin use are the same as those activated during spiritual experiences. Further research of psychotropic drugs is needed. A Sample of Relevant Religious and Spiritual Practices In order to fully understand the implications of the research described above, it is important to have some insight into the meditation practices studied.
Religious Experience: Psychology and Neurology
One form of Kundalini Yoga meditation includes volitional practices such as one in which practitioners passively observe their breathing and silently repeat the phrases ‘sat nam’ during their inhalation and ‘wahe guru’ during exhalation. It is important to note that this practice focuses meditators both on a physical point (their breathing) and a mental point (the phrases). This type of meditation activates neural structures involved in attention (frontal and parietal cortex), which relate to the mental concentration and arousal/ autonomic control (pregenual anterior cingulate, amygdala, midbrain, and hypothalamus), relating to physical concentration. A practice of Tibetan Buddhist meditators involves focusing on an image in their mind’s eye, and concentrating on increasingly absorbing themselves in this image. Practitioners report clarity of thought and losing a sense of time and space, making this form of meditation, similar to the Kundalini Yoga practice mentioned above, one that focuses intently on mental processes, however, without physical focus. Increases in CBF in the inferior and orbital frontal cortices, dorsolateral prefrontal cortices, the sensorimotor and dorsomedial cortices, the midbrain, cingulate gyri, and the thalamus are often seen here. Note that brain areas important for autonomic arousal are not active here. A Yoga Nidra guided meditation practice involves a relaxed meditative state in which the consciousness of the sensory world and the consciousness of action are subjectively dissociated, meaning the mind ‘withdraws’ from wishing to act. Meditators in this tradition tend to report a loss of conscious control of action, and vivid imagery. While Yoga Nidra meditators are instructed to be relaxed, their focus is not physiological in the sense of the focus on breath in Kundalini meditation. Yoga Nidra meditators tend to have increased dopaminergic tone in the ventral striatum during an altered consciousness state. This makes sense when examining circuits in the brain that run from the frontal to the subcortical regions of the brain, in which the ventral striatum is wired, that regulate behavior. Damage to one of these loops can cause anterior cingulate syndrome, which is characterized by an increased sense of apathy and a decrease in speech, motor behavior, and emotionality. What follows from the existence of an anterior cingulate
285
syndrome is that altered activity in the ventral striatum during meditation could, in some form, mirror the effects of an altered frontal subcortical loop. This research suggests that functions known to be mediated by a certain area can be altered through thought processes established in meditation, as seen in a change in the local activity. As is evident, meditation practices vary widely, and have very different effects on brain activation. Psychological Studies of Spiritual Experiences and Religion Meditation, and more generally, spiritual practices are quickly spreading into many areas of psychological and health treatments. It appears that even short-term practice of meditation can significantly alter an individuals functioning, both physiologically and psychologically. One of the most agreed upon changes that occur is an increase in attention. This includes increased performance on tests that present distracting stimuli and the ability to sustain attention. However, all types of spiritual practices do not show the same results. Research on the effects of spirituality and religion on health outcomes is abundant and not the focus of this article. However, meditation is being used to treat psychological effects of disease, and has gained a reputation as a treatment for anxiety and stress. Its effects on the CNS seem to be able to translate into decreasing individuals’ perception of their stress levels. Programs that teach mindfulnessbased stress reduction are being developed in hospitals and recommended for both the healthy and sick. Another mindfulness-based practice has been integrated into various forms of cognitive therapy to treat depression and anxiety and prevent relapse into these psychological conditions. Mantra recitation has been used in HIV patients to improve their overall quality of life, existential well-being and spiritual faith, even outside of affecting their physiological functioning. In addition to those suffering from illness, caregivers tend to turn to religion and spirituality for coping purposes, and religiosity seems to contribute to their own well-being. There remains some debate over the idea of religion as a universal buffer against psychological stress, primarily because the health benefits, which are only slightly significant, may be due to the
286
Religious Experience: Psychology and Neurology
individual, not to the religious activity itself. Religion has also been associated with fear of death and guilt, and when someone whose religious beliefs differ from those around him or her, religion can lead to stress and tension. Many have suggested a link between religion and happiness, however, because a clear definition of happiness does not exist, this relationship needs to be qualified more exactly. Overall, more research is needed on the effect of religion and psychological well-being.
Future Issues with Studying Spiritual Experiences It is important to note that while neurophysiological mechanisms may be correlated with a spiritual experience they do not explain why the brain is capable of generating these experiences. Some would assert that this ability proves the existence of a higher power, but this is an opinion that should be considered with great care. There is no reasonable hypothesis to explain conscious awareness or spiritual experience arising out of an electrical input/output system, no matter what its complexity. It is unknown if animals with less complicated nervous systems have spiritual experiences or even have basic consciousness. This and other problems are explored in other articles of this encyclopedia. Critics of research on religious experience and spiritual experience assert that the field is seeking to measure the immeasurable, and some religious groups argue that the research attempts to reduce religious experience to a function of neurons and neurotransmitters. This leads to the question of whether the deepest spiritual and mystical experiences are externally real or not, that is, solely generated and existing in the minds of the meditators and mystics, or actually part of some external reality/force in an objective sense. However, it might be argued that physiological analyses need not challenge the reality of religious experiences
or of a possible reality causing them, any more than it challenges the reality of the physical world, since ultimately, all human experiences are interpreted in some manner by the brain. Another issue with the current research is that the subjects studied tend to be individuals whose practices are on the extreme end of the religious spectrum. Such a small percentage of people feel that they attain these intense spiritual states, that these studies do not address the neurobiology of a typical religious experience. The everyday believer who might go to a church, synagogue or mosque on their holy day, and consider that to be his/her religious activity, is more difficult to evaluate scientifically because of the subtle nature of these experiences. Thus, substantial advances in scientific methodology and an expansion of research to include the enormous variety of religious and spiritual experiences, not just the strongest ones, will be required to fully elucidate the nature of the psychology and neurology of such experiences. See also: Altered and Exceptional States of Consciousness; The Neurochemistry of Consciousness.
Suggested Readings Austin J (1999) Zen and the Brain. Cambridge, MA: MIT Press. Beauregard M and O’Leary D (2007) The Spiritual Brain. New York: HarperOne. Cahn BR and Polich J (2006) Meditation states and traits: EEG, ERP, and neuroimaging studies. Psychological Bulletin 132(2): 180–211. Cardena E, Lynn SJ, and Krippner S (2000) Varieties of Anomalous Experience. Washington, DC: APA. McNamara P (2006) Where God and Science Meet: How Brain and Evolutionary Studies Alter Our Understanding of Religion. Westport, CT: Praeger Publishers. Newberg AB and Iversen J (2003) The neural basis of the complex mental task of meditation: Neurotransmitter and neurochemical considerations. Medical Hypothesis 61(2): 282–291. Newberg AB and Waldman MR (2007) Born to Believe. New York: Free Press.
Religious Experience: Psychology and Neurology
287
Biographical Sketch
Hannah Roggenkamp, BA is a research assistant in the Department of Radiology, University of Pennsylvania.
Mark Waldman, BA is a therapist and an associate fellow at the Center for Spirituality and the Mind, University of Pennsylvania.
Andrew Newberg, MD is associate professor in the Department of Radiology and Psychiatry, and director of the Center for Spirituality and the Mind, University of Pennsylvania.
Self: Body Awareness and Self-Awareness J L Bermu´dez, Washington University in St. Louis, St. Louis, MO, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Body awareness, high level – Representations of the body that feed directly into and are informed by central cognitive/ affective processes. Body awareness, low level – Awareness of the structure, layout, and momentby-moment distribution of body parts derived from mechanisms of somatic proprioception and somatosensation. Immunity to error through misidentification – Information sources (such as introspection or proprioception) are immune to error through misidentification when the information that they provide can only be about the subject receiving it – when there is no possibility for misidentifying the source of the information. Self-awareness – Self-awareness is a special type of awareness of the self. It is derived from information sources that are immune to error through misidentification and that have immediate and direct implications for action. Somatic proprioception – Awareness of limb position and bodily configuration derived from the vestibular (balance) system, the sense of touch, and receptors in the joints, muscles, and tendons.
Introduction Human beings, and indeed most animals, are aware of their bodies in distinctive ways. We have ways of finding out about our own bodies that are very different from the ways we have of finding out about other physical objects in the world. There
are distinctive information channels that allow us directly to monitor both the body’s internal states and how it is oriented in space. Some of these information channels are conscious and others unconscious. They all contribute, however, to a distinctive type of experience, viz, the experience of oneself as an embodied agent. It is to this distinctive type of experience that theorists are referring when they talk about body awareness. Body awareness is of fundamental importance for the survival of the organism. It allows us to monitor homeostatic states, such as hunger and thirst, as well as to detect and anticipate bodily damage. It is of fundamental importance in generating and controlling action. Moreover, body awareness has an important role to play in self-awareness. For the purposes of this article, I take self-awareness to be the same as self-consciousness and am understanding it in a specialized sense (explained in the ‘Glossary’ and in the ‘Self-awareness and the elusiveness thesis’ and ‘Body awareness as self-awareness’ sections). This article begins by outlining the different types of body awareness. High-level forms of body awareness are discussed in ‘High-level body awareness’ and lower-level types of body awareness in ‘Low-level body awareness’. The lowerlevel types will be primarily emphasized in the article, and the section ‘The mechanisms of lowerlevel body awareness’ outlines some of what is known about the physiological mechanisms governing lower-level body awareness. In ‘Body awareness in visual perception’ section, we consider a type of body awareness derived from visual perception. The sections entitled ‘Self-awareness and the elusiveness thesis’ and ‘Body awareness as self-awareness’ explain why body awareness counts as a form of self-awareness. Finally, in ‘The space of body awareness’ we explore the spatiality of body awareness and the differences between how we experience bodily space and how we experience extra-bodily space.
289
290
Self: Body Awareness and Self-Awareness
High-Level Body Awareness To understand the complex phenomenon of bodily awareness, we need to begin with a very general distinction between high-level and low-level representations of the body. High-level representations of the body feed directly into, and are informed by, central cognitive/affective processes. In contrast, low-level representations of the body (which we will consider in the next section) feed directly into action. The general distinction is sometimes made between the body-image (highlevel) and body-schema (low-level). In this section and the next, I extend this basic distinction into a fuller taxonomy of the different types of body awareness. Within the general category of high-level representations of the body, we can distinguish at least three different types of representation or bodies of information: Conceptual awareness of the body: The set of beliefs we all have about the structure and nature of our body: how the body fits together, the functions of particular body-parts, their approximate locations, and the sort of things that can go wrong with them. Semantic awareness of the names of body-parts: Knowledge that interfaces with nonsemantic ways of identifying events in the body to allow us to report on what is going on in our bodies. Affective awareness of the body: Representations of the body associated with emotional responses to the body. Conceptual representations of the body are not particularly interesting from the perspective of self awareness. There seems little reason to think that such conceptual representations will be any different in kind from the set of commonsense beliefs that we all have about the physical and social world. They may be based upon particular forms of body awareness, but it does not seem appropriate to describe them as actually being a form of body awareness. The remaining two types of higher-level representation are more interesting and more clearly types of body awareness. One way of seeing this is to reflect on how these higher-level representations can become distorted. There are identifiable
pathologies specific to both semantic and affective representations of the body. The pathologies associated with affective representations of the body are familiar. Bulimia and anorexia are good examples – forms of emotional response based on distorted representations of the body. These are plainly types of body awareness, as is the (relatively) undistorted affective response to the body that normal subjects have. There are also identifiable pathologies associated with semantic representations of the body. Patients with autotopagnosia have difficulty in naming body-parts or pointing to body-parts identified by name or by the application of some stimulus, either on their own bodies or on a schematic diagram of the body. The problems here are not purely semantic. Semantic representation of the body is not simply a matter of knowing the names of body-parts. Although superficially similar deficits can be found in some aphasic patients, autotopagnosic patients do not have a localized word category deficit. They lack a particular way of representing bodily locations, as we see from the fact that the problem carries across to pointing to body-parts identified by the application of a stimulus. Again, what is at stake here is the distortion of what in normal subjects is a mode of body awareness – an awareness of the body that manifests itself in the ability to identify specific body-parts.
Low-Level Body Awareness Turning to lower-level representations of the body, here too we find a range of phenomena and associated information channels that need to be distinguished. The first is information about the structure and limits of the body. This type of bodyrelative information has a number of distinctive pathologies, which have been studied by Ronald Melzack, Vilayanur Ramachandran, and others. The best known is the phenomenon of phantom limb found in many patients with amputated limbs, as well as some with amelia, the congenital absence of limbs. This first category of body-relative information performs two tasks. First, it is responsible for the felt location of sensations. Sensations are referred to specific body-parts in virtue of a body of information about the structure of the body.
Self: Body Awareness and Self-Awareness
Second, the same body of information informs the motor system about the body-parts that are available to be employed in action. Information about the structure and limits of the body is relatively stable, although experiments by James Lackner and others have shown that it can be distorted (as in the aptly named Pinocchio illusion, in which subjects who are touching their nose and have their biceps tendon stimulated with a vibrator feel their noses growing). This structural type of body-relative information should be distinguished from semantic representations of the body (as discussed in the section ‘Highlevel body awareness’). In deafferented patients they are dissociated in both directions. Deafferented patients have lost peripheral sensations in certain parts of the body. The patient GL studied by Jacques Paillard suffers from almost complete deafferentation from the mouth down, although she retains some sensitivity to thermal stimuli. If a thermal stimulus is delivered to a point on her arm that she is prevented from seeing then, although she is unable to point to the location of the stimulus on her body, she is able to identify the location verbally and on a schematic body diagram. In my terms, she possesses semantic information without body-relative information. The dissociation also holds in the opposite direction. Another of Paillard’s patient had a parietal lesion that resulted in central deafferentation of the forearm. Although she could not verbally identify and report on a tactile stimulus delivered to her deafferented hand in a blindfolded condition, she was able to point to the location of the stimulus. (Unlike the very similar and well-documented phenomenon of blindsight, there was no need to force subjects to make a choice.) The dissociation here may well be between an action-based representation of the body and an objective representation of the body. An actionbased representation of the body represents body location in a way that feeds directly into action, whether that action is body-directed or worlddirected. It is this that is lost in GL, but preserved in the patient with the deafferented forearm. In what I am calling an objective representation of the body, on the other hand, the body does not feature purely as a potentiality for action, but rather as a physical object whose parts stand in certain determinate relations to each other.)
291
There is a second type of lower-level representation of the body. This is a moment-to-moment representation of the spatial position of the various parts of the body. This moment-to-moment representation of bodily position is essential for the initiation and control of action, and needs to be constantly updated by feedback from moving limbs. This representation has been called the short-term bodyimage by the philosopher Brian O’Shaughnessy, but the name is misleading, suggesting that there is a single way in which the disposition of body-parts is represented, whereas in fact the spatial location of any given body-part can be coded in three different and independent ways. The first type of coding is relative to objects in the distal environment. Consider a simple action, such as reaching one’s hand out for an object. The success of this action depends upon an accurate computation of the trajectory from the initial position of the hand to the position of the relevant object. This requires the position of the hand and the position of the object to be computed relative to the same frame of reference. I call this object-relative spatial coding. It is most likely that object-relative spatial coding takes place on an egocentric frame of reference – that is to say, a frame of reference whose origin is some body-part. The reason for calling this type of coding objectrelative is that it deals primarily with the spatial relations between body-parts and objects in the distal environment. But many actions are directed toward the body rather than to objects independent of the body. Some of these actions are voluntary, as when I clasp my head in my hands in horror. Some are involuntary, as when I scratch an itch. Many more are somewhere between the two, as when I cross my legs or rub my eyes. Clearly, the possibility of any of these sorts of action rests upon information about the location of the body-parts in question relative to each other. We can call this sort of information body-internal spatial coding. It is information about the moment-by-moment position of body-parts relative to each other. Body-internal spatial coding is required, not just for body-directed action, but also for many types of action directed towards objects in the distal environment. Psychological studies of action often concentrate on very simple actions, such as grasping objects with one hand. But the vast majority of actions
292
Self: Body Awareness and Self-Awareness
require the coordination of several body-parts. When I play volleyball, for example, I need to know not just where each of my hands is relative to the ball as it comes over the net, but also where each of my hands is relative to the other hand. Both body-internal and object-relative spatial coding is required. A third type of information about the momentto-moment disposition of the body is just as important for the initiation and control of action as the first two. This is information about the orientation of the body as a whole in objective space, primarily involving information about the orientation of the body with respect to supporting surfaces and to the gravitational field. This information comes from the calibration of information from a number of sources. The three principal sources of orientational information are vision, the vestibular system in the inner ear, and the proprioceptive/kineasthetic system (at least two of which must be properly functioning for orientational information to be accurate). I call this orientational coding.
The Mechanisms of Lower-Level Body Awareness The various types of lower-level body awareness have intricate and highly developed physical underpinnings that are relatively well understood. Physiologists and neurophysiologists have devoted considerable attention to understanding the mechanisms of proprioception (awareness of limb position and bodily configuration) and somatosensation (bodily sensation). We have a good understanding of how bodily sensations originate in specialized receptors distributed across the surface of the skin and within the deep tissues. Some of these receptors are sensitive to skin and body temperature. Others are pain detectors (nociceptors). There are receptors specialized for mechanic stimuli of various kinds, such as pressure and vibration. Information about muscle stretch comes from muscle spindles. Other receptors monitor stresses and forces at the joints and in the tendons. Information from all of these receptors and nerve endings is carried by the spinal cord to the brain. Once again, the mechanisms here are relatively well understood. It is thought, for example, that there are three different pathways in the spinal cord.
One pathway carries information stemming from discriminative touch (which is a label for a complex set of tactile ways of finding out about the shape and texture of physical objects). Another carries information about pain and temperature. The third carries proprioceptive information. Each of these pathways ends up at a different brain area. The discriminative touch pathway travels to the cerebral cortex while the proprioceptive pathway terminates in the cerebellum. The properties of these brain areas have been well studied. We know, for example, that tactile information is processed in the somatosensory cortex, which is located in the parietal lobe. The somatosensory cortex is somatotopically organized, with specific regions representing specific parts of the body. The cortical space assigned to information from each bodily region is a function of the fineness of tactile discrimination within that region (which is itself of course a function of the number of receptors there). Neuropsychologists, neuroimagers, and computational neuroscientists have made considerable progress in understanding how somatosensory and proprioceptive information is processed in the brain and how that processing can be disturbed by brain injury.
Body Awareness in Visual Perception We can of course perceive the body visually in the same way as we perceive any other object. Looking at ourselves in the mirror would be an example. This does not count as body awareness in any interesting sense. But there are several ways in which vision provides us with a distinctive form of body awareness. These operate through 1. self-specifying structural invariants in the field of vision, 2. visual kinesthesis, and 3. the perception of affordances. Visual body awareness is derived from various types of self-specifying information available in visual perception. These have been studied most systematically by the perceptual psychologist J. J. Gibson. Unfortunately, Gibson’s work in this area is not as widely accepted as it should be, no doubt due to his polemical and controversial criticisms of
Self: Body Awareness and Self-Awareness
standard information-processing approaches to vision. Gibson argued that vision does not involve the type of information-processing standardly studied by cognitive scientists and cognitive neuroscientists. We should understand vision in terms of direct sensitivity to information that is present in the flow of light around the perceiver. Gibson’s theory of ecological optics is based on the idea that our perceptual systems resonate to this information. Fortunately, it is possible to learn from Gibson’s analysis of the self-specifying information in vision without taking literally his ideas about the direct pickup of information. We begin with the self-specifying structural invariants. The field of vision contains bodily objects that hide, or occlude, the environment. The nose is an obvious example. It is distinctively present in just about every visual experience. The cheekbones, and perhaps the eyebrows, occupy a slightly less-dominant position in the field of vision. And so too, to a still lesser extent, do the bodily extremities, hands, arms, feet, and legs. They protrude into the field of vision from below in a way that occludes the environment, and yet which differs from the way in which one nonbodily physical object in the field of vision might occlude another. They are, as Gibson points out, quite peculiar objects. All objects, bodily and nonbodily, can present a range of solid angles in the field of vision (where by a solid angle is meant an angle with its apex at the eye and its base at some perceived object), and the size of those angles will of course vary according to the distance of the object from the point of observation. The further away the object is, the smaller the angle will be. This gives rise to a clear, and phenomenologically very salient, difference between bodily and nonbodily physical objects. The solid angles subtended by occluding body-parts cannot be reduced below a certain minimum. Perceived body-parts are, according to Gibson, ‘subjective objects’ in the content of visual perception. The mass of constantly changing visual information generated by the subject’s motion poses an immense challenge to the perceptual systems. How can the visual experiences generated by motion be decoded so that subjects perceive that they are moving through the world? Gibson’s notion of visual kinesthesis is his answer to this traditional problem. Whereas many theorists have assumed that motion perception can only be
293
explained by hypothesizing mechanisms that parse cues in the neutral sensations into information about movement and information about static objects, the crucial idea behind visual kinesthesis is that the patterns of flow in the optic array and the relations between the variant and invariant features make available information about the movement of the perceiver, as well as about the environment. As an example of such a visually kinesthetic invariant, consider that the optical flow in any field of vision starts from a center that is itself stationary. This stationary center specifies the point that is being approached, when the perceiver is moving. The aiming point of locomotion is at the vanishing point of optical flow. These kinesthetic invariants count as forms of bodily awareness because they directly provide information about the body’s trajectory through space. The theory of ecological optics identifies a third form of self-specifying information in the field of vision. This is due to the direct perception of a class of higher-order invariants that Gibson terms affordances. It is in the theory of affordances that we find the most sustained development of the ecological view that the fundamentals of perceptual experience are dictated by the organism’s need to navigate and act in its environment. The uncontroversial premise from which the theory of affordances starts is that objects and surfaces in the environment have properties relevant to the abilities of particular animals, in virtue of which they allow different animals to act and react in different ways. According to Gibson, information specifying affordances is available in the structure of light to be picked up by the creature as it moves around the world. The possibilities that the environment affords are not learnt through experience, and nor are they inferred. They are directly perceived as higher-order invariants. The perception of affordances is a form of body awareness, because it contains information about environmental possibilities for action and reaction.
Self-Awareness and the Elusiveness Thesis The preceding sections have been devoted to exploring the nature and mechanisms of body
294
Self: Body Awareness and Self-Awareness
awareness. We turn now to the idea that body awareness is a form of self-awareness. In this section, I give some philosophical background that will help in thinking about body awareness and self-awareness. The first step is to distinguish two different types of awareness – direct awareness and propositional awareness. One can be aware of something (as when I catch sight of someone walking up the garden path) or one can be aware that a particular state of affairs is the case (as when the sound of the doorbell alerts me to the fact that a visitor is at the door). In direct awareness, the object of awareness is a particular thing. In propositional awareness, the direct object of awareness is a proposition or state of affairs (a complex of particular things, properties, and/or relations). Direct awareness is not sensitive to how one thinks about the object of which one is directly aware. All that is required for me to be directly aware of something is that I be able to discriminate it. I do not need to know what it is, or to conceptualize it in any way. Propositional awareness, however, is not like this. I can be aware that a state of affairs holds when it is conceptualized in one way, but be unaware that it holds under a different conceptualization. I might be propositionally aware that Bob Dylan is balding, in virtue of seeing that the person on the stage in front of me is losing his hair and know that that person is Bob Dylan, without being propositionally aware that Robert Zimmerman is balding, since I have no idea that Bob Dylan is Robert Zimmerman. In contrast, if I am directly aware of Bob Dylan, then I am directly aware of Robert Zimmerman, even if I have no idea who either of them is. Applying this distinction gives us two ways of thinking about self-awareness. Self-awareness can be understood either in terms of direct awareness of the self or in terms of propositional awareness that the self has such-and-such a property, or stands in such-and-such relations. Although it might seem obvious that we can be directly aware of ourselves, some philosophers have maintained that there can be no such thing as direct self-awareness at the level of direct awareness. The most famous example of this is David Hume’s strong claim in his Treatise of Human Nature that the self cannot be directly perceived.
For my part, when I enter most intimately into what I call myself, I always stumble on some perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. I never catch myself at any time without a perception, and never can observe anything but the perception.
Hume’s claim, in effect, is that we cannot be directly aware of the self in introspection (taking ‘‘introspection’’ here to pick out the means by which we enjoy first-person access to our own mental states). This claim, which seems itself to be derived from introspection, is generally known as Hume’s elusiveness thesis. At the time at which Hume was writing, the consensus view among philosophers and scientists was that the self is a purely psychological entity. This explains the weight attached to the elusiveness thesis. If the self is a purely psychological entity, then introspection is the only possible source for direct awareness of the self, and the impossibility of introspective awareness of the self would cast significant doubt on whether any such purely psychological entity exists (particularly for empiricists such as Hume, since empiricists typically hold that all knowledge must come either through the senses or from introspection). Since there are few if any surviving dualists, Hume’s elusiveness thesis does not now have the same force that it once did. In fact, one might reasonably think that the forms of body awareness we have been discussing in this article all count as forms of direct awareness of the embodied self – and hence as counterexamples to the elusiveness thesis. And then, one might go on to think, body awareness is a form of self-awareness (self-consciousness). Nonetheless, things are not quite as straightforward as this. Not all forms of direct awareness of the self count as forms of self-awareness. If I catch sight of myself in a shop window but do not realize that the person in the window is me, then I am certainly directly aware of myself, but I am not self-aware in any interesting sense. So we need to ask whether body awareness is like the shop window case, or whether it counts as a genuine form of self-awareness. The first step in answering this is to get a clearer idea of what counts as genuine self-awareness. Most philosophers would think that there are at least two minimal conditions on genuine
Self: Body Awareness and Self-Awareness
self-awareness over and above its involving a direct awareness of the self. The first emerges when we think about an obvious feature of the shop window scenario. One of the things that makes the shop window case so obviously not an example of genuine self-awareness is the fact that I fail to identify myself. I correctly identify certain bodily properties, but fail to recognize that they are properties of my own body and misattribute them to somebody else. We do not want to collapse direct awareness into propositional awareness, and so it cannot be a requirement upon direct awareness of the self that it involves any sort of judgment (such as the judgment: ‘‘I am that person’’). Nonetheless, we can reasonably demand that for direct awareness of the self to count as genuine self-awareness it should be derived from information channels that do not yield information about anybody’s bodily properties except one’s own (just as introspection does not yield information about anybody’s psychological properties except one’s own). In philosophy this is often put in rather cumbersome terms. In what has become the standard terminology, what I am calling genuine self-awareness is based upon information sources that have the property of being immune to error through misidentification relative to the first-person pronoun. A second requirement upon genuine selfawareness is that it feeds directly into action. Again, we can appreciate this by noting how it is absent from the shop window case. Suppose I see that the person in the window is about to walk into a lamppost. In the scenario I am envisaging I may laugh, or I may be concerned. But, since I do not realize that I am the person who is about to walk into a lamppost, I will not take evasive action. Again, we do not want to make it a requirement upon genuine selfawareness that it involves any sort of judgment (such as: ‘‘I am about to walk into the lamppost’’). But it seems reasonable to demand that, for a form of direct awareness of the self to count as genuine self-awareness, it should be based on forms of information that can feed directly into action
Body Awareness as Self-Awareness It is plain that body awareness provides direct awareness of the embodied self. But does it provide
295
a form of self-awareness? At the end of the previous section, we identified two basic requirements for a form of awareness of the self to count as genuine self-awareness. In this section, we begin by seeing how these two basic requirements are satisfied by body awareness. We then go on to explore a further sense in which body awareness yields awareness of the self. The first requirement for direct awareness of the self to count as genuine self-awareness is that it should be based upon sources of information that do not yield information about anybody’s bodily properties except my own (just as introspection does not yield information about anybody’s psychological properties except my own). Here we need to focus primarily on two of the types of body awareness distinguished earlier – one operating through the mechanisms of somatic proprioception and somatosensation, and the other operating through self-specifying information in vision. Both of these satisfy the first requirement. Somatic proprioception is the most straightforward. It seems to follow from the simple fact that I somatically proprioceive particular bodily properties that those bodily properties are my own. We are quite simply ‘‘wired up’’ in such a way that proprioceptively derived information has the property of being immune to error through misidentification. Of course, this is not a logical truth. And it may not even be a nomological truth (that is, a truth dictated by the laws of nature). It is certainly conceivable that we could be hooked up to other people’s bodies in such a way that we receive proprioceptive information from their bodies as well as (or even instead of) from our own. But since, as things currently stand, that possibility is not a ‘relevant alternative,’ it does not need to be taken into account. I am always justified in taking proprioceptively derived information to be information about my own body. The same holds for self-specifying information derived through vision. The forms of selfspecifying information identified earlier collectively create, for each of us, a distinctive and unique visual perspective on the world – and of course on ourselves, as physical objects navigating the world. This direct awareness of ourselves obtained through vision satisfies the first requirement for genuine self-awareness because one’s perceptually derived information about where
296
Self: Body Awareness and Self-Awareness
one is relative to objects in the perceived environment could not possibly be about anyone but oneself. Of course, as the mirror and shop window cases show, not all visually derived awareness of the self counts as self-awareness. We need to restrict ourselves to the types of information identified earlier, such as that gained through visual kinesthesis and self-specifying structural invariants. Moving on to the second requirement for genuine self-awareness, it is clear that direct awareness of the self through somatic proprioception and somatosensation does have immediate and direct implications for action. This holds true both of self-directed action (such as scratching itches) and of action whose target is not the body. One index of this is the role of the various somatic information systems in providing feedback about limb position and balance, as was brought out in our earlier discussion of the different types of lowlevel body awareness. The action-guiding role of self-specifying information in vision is equally clear. Gibson’s analysis of vision focused primarily on how action is guided by information available in the optic flow – information both about the perceiver’s movement relative to landmarks in the environment and about the possibilities for action and reaction that the environment affords. The important point is that the implications for action are direct – precisely because this is information that cannot be but about the perceiver. This is the crucial contrast with information derived from mirrors, shop windows, and so forth. This information has implications for action only after one has identified the person seen as oneself. The direct awareness of the self obtained from the two types of body awareness that we have discussed do seem, therefore, to count as genuine forms of self-awareness, according to the two criteria we have identified. But there is, in addition, a further sense in which body awareness counts as a form of self-awareness. A crucial element in selfawareness is what developmental psychologists call self-world dualism. All subjects properly described as self-aware must be able to register the distinction between themselves and the world. This is a distinction that can be registered in a variety of different ways, and at different degrees of sophistication. Body awareness provides
a way, perhaps the most primitive way, of registering the distinction between self and nonself. This is a weaker distinction than the distinction between self and world, of course, but it is certainly a necessary component of it. There are two key elements to the distinction between self and nonself yielded by body awareness. The first element is an awareness of the limits of the body. This comes from a number of sources. The sense of touch has an important part to play, as the felt boundaries of the body define the limits between self and nonself. This offers a way of grasping the body as a spatially extended and bounded object. But self-specifying information in vision also has a role to play. As we saw earlier, body-parts appear in vision in a special way (as what Gibson evocatively calls ‘subjective objects’), since their size can vary in only a limited way. A second element in understanding the distinction between self and nonself is being able to distinguish between what is, and what is not, responsive to the will. Here somatic proprioception is particularly important. It is the feedback gained through kinesthesia, joint position sense and the vestibular system, which explains how one is aware that the body is responding to motor commands. It is true that much of the body is not at all responsive to the will – the internal organs are obvious examples. But the scope of the will does encompass all bodily surfaces and extremities (whether directly or indirectly). Although not every portion of the bodily surface can be moved at will, it is nonetheless the case that every portion of the bodily surface can be experienced as moving in response to an act of the will. Take an arbitrary area on the top of the head, for example. Although I cannot move that area at will (in isolation), I can nonetheless experience it as moving when I move my head as a whole. The limits of the will mark the distinction between the self and the nonself just as much as does the skin, although in a different way.
The Space of Body Awareness Almost all existing discussions of the spatiality of body awareness have presupposed that exteroceptive
Self: Body Awareness and Self-Awareness
perception, body awareness, and the intentions controlling basic bodily actions must all code spatial information on comparable frames of reference (where a frame of reference allows locations to be identified relative to axes centered on an object). This is a natural assumption, given that action clearly requires integrating motor intentions and commands with perceptual information and several different types of lower-level information about the body (particularly what we earlier termed the short-term body-image). Since the spatial locations of perceived objects and objects featuring in the contents of intentions are given relative to axes whose origin lies in the body – in an egocentric frame of reference – it is natural to suggest that the axes determining particular proprioceptive frames of reference are centered on particular body-parts, just as are the axes determining the frames of reference for perceptual content and basic intentions. The picture that emerges, therefore, is of a number of different representations of space, within each of which we find representations both of bodily and of nonbodily location. So, for example, we might imagine reaching behavior to be controlled by an egocentric frame of reference centered at some location on the hand – a frame of reference relative to which both bodily location (such as the mosquito bite on my arm) and nonbodily location (such as the cup on the table) can be identified. Despite its appealing economy, however, this account is ultimately unacceptable, because of a fundamental disanalogy between the bodily space of lower-level body awareness and the egocentric space of perception and action. In the case of vision or exteroceptive touch, there is a perceptual field bounded in a way that determines a particular point as its origin. Since the visual field is essentially the solid angle of light picked up by the visual system, the origin of the visual field is the apex of that solid angle. Similarly, the origin of the frame of reference for exploratory touch could be a point in the center of the palm of the relevant hand. But our awareness of our own bodies is not like this at all. It is not clear what possible reason there could be for offering one part of the body as the origin of the frame of reference that fixes locations in bodily space. There are certain spatial notions that are not applicable to body awareness. For any two objects
297
that are visually perceived, it makes obvious sense to ask both of the following questions: (a) Which of these two objects is further away? (b) Do these objects lie in the same direction? The possibility of asking and answering these questions is closely bound up with the fact that visual perception has an origin-based frame of reference. Question (a) basically asks whether a line between the origin and one object would be longer or shorter than a corresponding line between the origin and the other object. Question (b) is just the question whether, if a line were drawn from the origin to the object that is furthest away, it would pass through the nearer object. Neither question makes sense with respect to body awareness. One cannot ask whether this proprioceptively detected hand movement is farther away than this itch, nor whether this pain is in the same direction as that pain. What I am really asking when I ask which of two objects is further away is which of the two objects is further away from me, and a similar tacit self-reference is included when I ask whether two objects are in the same direction. But through somatic proprioception one learns about events taking place within the confines of the body, and there is no privileged part of the body that counts as me for the purpose of discussing the spatial relations they bear to each other. To get a firmer grip on the distinctiveness of the frame of reference of bodily awareness, one need only contrast the bodily experience of normal subjects with that of completely deafferented subjects, such as Jonathan Cole’s patient IW. The momentto-moment information about their bodies that deafferented patients possess is almost exclusively derived from vision. Their awareness of their own body is continuous with their experience of the extra-bodily world. They are aware of their bodies only from the same third-person person perspective that they have on nonbodily physical objects. The frame of reference for their bodily awareness does indeed have an origin – the eyes – and for this reason both of the two questions mentioned make perfect sense. But this is not at all how we experience our bodies from a first-person perspective. The conclusion to draw from this is that the spatial content of bodily awareness cannot be specified within a Cartesian frame of reference
298
Self: Body Awareness and Self-Awareness
that takes the form of axes centerd on an origin. But then how is it to be specified? We can start from the basic thought that an account of the spatiality of bodily awareness must provide criteria for sameness of place. In the case of somatic proprioception, this means criteria for sameness of bodily location. But there are several different forms of criteria for sameness of bodily location. Consider the following two situations: (i) I have a pain at a point in my right ankle when I am standing up and my right foot is resting on the ground in front of me. (ii) I have a pain at the same point in my ankle when I am sitting down and my right ankle is resting on my left knee. According to one set of criteria the pain is in the same bodily location in (i) and (ii) – that is to say, it is at a given point in my right ankle. According to another set of criteria, however, the pain is in different bodily locations in (i) and (ii), because my ankle has moved relative to other body-parts. Let me term these A-location and B-location, respectively. Note, moreover, that B-location is independent of the actual location of the pain in objective space. The B-location of the pain in (ii) would be the same if I happened to be sitting in the same posture five feet to the left. Both A-location and B-location need to be specified relative to a frame of reference. In thinking about this we need to bear in mind that the human body has both moveable and (relatively) immoveable body-parts. On a large scale the human body can be viewed as an immoveable torso to which are appended moveable limbs – the head, arms, and legs. Within the moveable limbs there are smallscale body-parts that can be directly moved in response to the will (such as the fingers, the toes, and the lower jaw) and others that cannot (such as the base of the skull). A joint is a body-part that affords the possibility of moving a further bodypart, such as the neck, the elbow, or the ankle. In the human body, the relatively immoveable torso is linked by joints to five moveable limbs (the head, two legs, and two arms), each of which is further segmented by means of further joints. These joints provide the fixed points in terms of which the particular A-location and B-location of individual body-parts at a time can be given.
A particular bodily A-location is given relative to the joints that bound the body-part within which it is located. A particular point in the forearm is specified relative to the elbow and the wrist. It will be the point that lies on the surface of the skin at such-and-such a distance and direction from the wrist and such-and-such a distance and direction from the elbow. This mode of determining A-location secures the defining feature of A-location, which is that a given point within a given body-part will have the same A-location irrespective of how the body as a whole moves, or of how the relevant body-part moves relative to other body-parts. The A-location of a given point within a given body-part will remain constant in both those movements, because neither of those movements will bring about any changes in its distance and direction from the relevant joints. The general model for identifying B-locations is as follows. A particular constant A-location is determined relative to the joints that bound the body-part within which it falls. A-location will either fall within the (relatively) immoveable torso or it will fall within a moveable limb. If it falls within the (relatively) immoveable torso, then its B-location will also be fixed relative to the joints that bound the torso (neck, shoulders, and leg sockets) – that is to say, A-location and B-location will coincide. If, however, that A-location falls within a moveable limb, then its B-location will be fixed recursively relative to the joints that lie between it and the immoveable torso. The B-location will be specified in terms of the angles of the joints that lie between it and the immovable torso. Some of these joint angles will be rotational (as with the elbow joint, for example). Others will be translational (as with the middle finger joint). This way of specifying A-location and Blocation seems to do justice to how we experience our bodies. In particular,
We do not experience peripheral body-parts in isolation, but rather as attached to other bodyparts. Part of what it is to experience my hand as being located at a certain place is to experience that disposition of arm-segments in virtue of which it is at that place. It is part of the phenomenology of bodily awareness that sensations are always experienced
Self: Body Awareness and Self-Awareness
within the limits of the body. This is exactly what one would expect given the coding in terms of A-location and B-location. There are no points in (nonpathological) body-space that do not fall within the body. Although B-location is specified recursively in terms of the series of joint angles between a given A-location and the immovable torso, the torso does not function as the origin of a Cartesian frame of reference. These aspects of the phenomenology of body awareness (of how we experience our bodies) are important factors in explaining why body awareness is a genuine form of self-awareness. The distinctiveness of the spatiality of body awareness is an important part of what underwrites the agent’s sense of the distinction between self and nonself.
Conclusion Philosophers dealing with bodily awareness have to balance two considerations that pull in opposite directions. There is, first, the obvious fact that the body is a physical object in the world – which generates the plausible thought that awareness of the body must be somehow awareness of the body as an object in the world. Second, and equally obvious, is the fact that the body is (at least from a firstperson perspective) quite unlike any other physical object – generating the equally plausible thought that awareness of the body is somehow fundamentally different from awareness of the body as an object in the world. The points about body awareness that have been discussed in this article offer a way of reconciling the apparent conflict. The types of body awareness that we have discussed jointly
299
provide an awareness of the self as a spatially extended and bounded physical object that is distinctive in being responsive to the will. See also: The MindBody Problem; Philosophical Accounts of Self-Awareness and Introspection; Self: Personal Identity; Self: The Unity of Self, SelfConsistency.
Suggested Readings* Bermudez JL, Marcel AJ, and Eilan N (eds.) (1995) The Body and the Self. Cambridge, MA: MIT Press. Bermudez JL (1998) The Paradox of Self-Consciousness. Cambridge, MA: MIT Press. Cassam Q (1997) Self and World. Oxford: Oxford University Press. Cole J and Paillard J (1995) Living without touch and peripheral information about body position and movement: Information from deaferented subjects. In: Bermudez JL, Marcel AJ, and Eilan N (eds.) The Body and the Self. Cambridge, MA: MIT Press. Evans G (1982) The Varieties of Reference. Oxford: Oxford University Press. Gallagher S (2005) How the Body Shapes the Mind. New York: Oxford University Press. Gibson JJ (1979) The Ecological Approach to Visual Perception. Boston: Houghton Mifflin. Knoblich G, Thornton I, Grosjean M, and Shiffrar M (2005) Human Body Perception from the Inside Out. Oxford: Oxford University Press. Lackner JR (1988) Some proprioceptive influences on the perceptual representation of body shape and orientation. Brain 111: 281–297. Melzack R (1992) Phantom limbs. Scientific American 266: 120–126. O’Shaugnessy B (1995) Proprioception and the body image. In: Bermudez JL, Marcel AJ, and Eilan N (eds.) The Body and the Self. Cambridge, MA: MIT Press. Ramachandran VS and Blakeslee S (1998) Phantoms in the Brain: Probing the Mysteries of the Human Mind. New York: William Morrow. Shoemaker S (1967) Self-reference and self-awareness. Journal of Philosophy 65(19): 555–567.
* This article draws upon material from my book The Paradox of SelfConsciousness (Cambridge MA: MIT Press. 1998) and from my article ‘The phenomenology of bodily awareness’ in Phenomenology and Philosophy of Mind, edited by D. Woodruff Smith and A. L. Thomasson (Oxford: Oxford University Press. 2005).
300
Self: Body Awareness and Self-Awareness
Biographical Sketch
Jose´ Luis Bermu´dez is a professor of philosophy at Washington University of St. Louis, where he is the director of the Philosophy–Neuroscience–Psychology program and the director of the Center for Programs in Arts and Sciences. Dr. Bermu´dez graduated from Cambridge University in 1988 and received his PhD in 1992. Before coming to Washington University in St. Louis, he was professor and chair of the philosophy department at the University of Stirling, UK. Professor Bermu´dez’s research interests are primarily in interdisciplinary philosophy of mind and philosophy of psychology. Topics of recent interest include the nature of mental content, models of psychological explanation, the role and origins of self-consciousness, and the possibility of thought without language. His books include The Paradox of Self-Consciousness (MIT, 1998), Thinking without Words (Oxford, 2003), Philosophy of Psychology: A Contemporary Introduction (Routledge, 2005), and Decision Theory and Rationality (Oxford, 2009).
Self: Personal Identity E T Olson, University of Sheffield, Sheffield, UK ã 2009 Elsevier Inc. All rights reserved.
Glossary Cerebrum – The largest and uppermost part of the brain, responsible for higher cognitive functions. Dissociative-identity disorder – A rare condition in which a human being acts as if he or she were inhabited alternately by a number of people, with different personalities, memories, and cognitive abilities. Functionalist theory of mind – A theory that characterizes mental states in terms of their typical causes and effects, including their interactions with other mental states. Hemispherectomy – The surgical removal of a cerebral hemisphere. Persistence conditions – The conditions logically necessary and sufficient for a thing to continue existing, as opposed to ceasing to exist. Personal ontology – The basic metaphysical nature of human people, such as whether we are material or immaterial. Persistent vegetative state – A long-term condition, caused by brain damage, in which one’s life-sustaining functions continue spontaneously but one is permanently unconscious and unable to act purposefully. Temporal part – A part of an object that has the same spatial extent as that object has during a period of time, and the same temporary properties, but a shorter temporal extent: for example, the first half of a football match.
Introduction Personal identity deals with philosophical questions about ourselves that arise by virtue of our being people (or persons). Some of these questions
are familiar ones that occur to all of us: What am I? When did I begin? What will happen to me when I die? Others are more abstruse. Many philosophers, following Locke and Hume, give consciousness a central role in answering questions of personal identity. They say that many of these questions are nothing more than questions about the unity and continuity of consciousness, or at least that facts about consciousness and related psychological matters suffice to settle the facts about personal identity. Others disagree, saying that consciousness is irrelevant to most questions of personal identity. (A terminological note: this article is about the self in the sense that my self is simply myself – me, the author – and not about other senses of the word self.)
The Problems of Personal Identity There is no single problem of personal identity, but rather a wide range of loosely connected questions. Here are some of the most central and important ones. The Persistence Question One question is what it takes for a person (or a human person) to persist or survive from one time to another. What sorts of adventures could you possibly survive, in the broadest sense of the word possible? What sort of event would necessarily bring your existence to an end? What determines which past or future being is you? What, in other words, does our persistence through time consist in? Suppose you point to a boy or girl in an old class photograph and say, ‘‘That’s me.’’ What makes you that one, rather than one of the others? What is it about the way that person then relates to you that makes her you? For that matter, what makes it the case that anyone at all who existed back then is you? Is a certain boy or girl you by virtue of 301
302
Self: Personal Identity
psychological connections to you? Is it, for instance, because you are now able to remember things that happened to her back then? Or should we look instead to brute physical or biological relations? Might that person be you because she is the same biological organism as you are, or because you and she have the same body? Or does the answer lie somewhere else entirely? This is the question of personal identity over time, or the persistence question. It applies to our survival into the future as well as our existence in the past. Historically this question arises when people wonder whether we might continue to exist after we die (in Plato’s Phaedo, for instance). Whether this could happen depends on whether biological death necessarily brings one’s existence to an end. Imagine that after your death there really will be someone, in the next world or in this one, who is in some ways like you. How would that being have to relate to you for it actually to be you, rather than someone else? What would the higher powers have to do to keep you in existence after your death? Or is there anything they could do? The answer to these questions depends on the answer to the persistence question. The Population Question We can think of the persistence question as asking which of the characters introduced at the beginning of a story have survived to become the characters at the end of it. But we can also ask how many characters are on the stage at any one time. What determines how many of us there are now? If there are some six billion people on the earth at present, what facts – biological, psychological, or what have you – make this the case? The question is not what causes there to be a certain number of people at a given time, but rather what there being that number of people consists in. It is like asking what sort of configuration of pieces amounts to black’s winning a game of chess, rather than like asking what sorts of moves might lead to its winning. This is the population question. (It is sometimes called the question of synchronic identity, as opposed to the diachronic identity of the persistence question.) One possible answer is that the number of people at any given time is simply the number of
human organisms there are then, which is determined by brute biology. (We may have to discount human organisms in an immature or radically defective state – human embryos and human vegetables, for instance – as many deny that they count as genuine people.) Another answer, proposed by Hume and further developed in Kant, is that the number of people is determined by facts about psychological unity and disunity. These answers may agree in ordinary cases, but diverge in unusual ones. Other answers are also possible. Commissurotomy – the severing of the corpus callosum connecting the cerebral hemispheres – can produce a certain degree of mental disunity, illustrated in extreme cases by such peculiar behavior as simultaneously pulling one’s trousers up with one hand and pulling them down with the other. Some even say that commissurotomy can produce two separate streams of consciousness. There is dispute about how much mental disunity there really is in these cases. But many philosophers say that if there were enough disunity, of the right sort, there would be two people – two thinking, conscious beings – sharing a single organism, each with its own unified consciousness and its own set of beliefs, experiences, and actions. Dissociative-identity disorder (multiple personality) raises similar issues: we can ask whether Dr. Jekyll and Mr. Hyde would be two people or one. (At least this is so if one of them continues to exist while the other is active; otherwise this case is a puzzle for the persistence question.) Does the existence of two or more separate streams of consciousness entail that there are two or more conscious beings? In these cases, or in imaginary but possible extensions of them, there can be psychological disunity of the sort that is ordinarily present in cases where there are two or more people: there may be two independent sets of beliefs and experiences, and two separate streams of consciousness. Yet at the same time there is biological unity of the sort we find when there is just one person: there is always just one biological organism present, and one brain. If the number of people or thinking beings is determined by facts about psychological unity (e.g., unity of consciousness), there will be two or more people in these cases. If it is determined by nonpsychological facts, such as biological
Self: Personal Identity
unity, there will be one. Philosophers disagree about this, though the psychological-unity account is currently more popular.
The Personal-Ontology Question More generally, we can ask about our basic metaphysical nature. What are the most general and fundamental characteristics of human people? This is one of the more abstruse questions about personal identity, but it is of central importance. We can call it the personal-ontology question. It is really a collection of more specific questions, such as these: What, at the most basic level, are we made of ? Are we made up entirely of matter, as stones are? Or are we made up at least partly of something other than matter? If we are made up of matter, what matter is it? Is it just the matter that makes up our bodies? Do we extend all the way out to our skin and no further, or might we be larger or smaller than our bodies? Where, in other words, do our spatial boundaries lie? More fundamentally, what fixes those boundaries? If we do extend right out to our skin, why that far and no further? Does this have to do with the extent of a certain sort of conscious awareness, as Locke thought, or does the answer lie in brute biology? Are we ontologically independent beings? Or is each of us a mere state or aspect of something else? Think of a knot in a rope. It is not ontologically independent. It cannot exist without the rope: you can’t take it away in your pocket and leave the entire rope behind. It is a state or aspect of the rope. The same goes for events, such as the rope’s gradually wearing out. The rope itself, though, does not seem to be a state or an aspect of anything else: nothing appears to stand to the rope as the rope stands to the knot. It seems to be an ontologically independent being – what metaphysicians call a substance. The question, then, is whether we are substances, like ropes, or whether we are states or aspects or events, like knots. Is there something – an organism or a mass of matter, perhaps – that stands to you as a rope stands to a knot? An answer to these questions and others like them would tell us our basic metaphysical nature.
303
The persistence and population questions also fall under the personal-ontology question. Here are the main proposed answers to this question: We might be biological organisms (and thus material substances). This view, held by Aristotle, is currently defended under the name of animalism. Most philosophers, both past and future, however, have rejected it. We might be immaterial substances or souls, as Plato, Descartes, and Leibniz held. This view faces notorious difficulties in explaining how the soul relates to the body – what enables my soul to animate this particular organism and no other, for instance – and has few advocates today. Many contemporary philosophers say that we are material things but not organisms. One such view is that we are spatial parts of organisms, such as brains. A more popular view is that we are temporal parts of organisms: you are spatially the same size as your animal body but temporally shorter, in that the animal extends further into the past or future than you do. Yet another currently popular view is that we are nonorganisms made of just the same matter as our animal bodies. The thinking behind this is that the same matter can make up two different objects at once. Specifically, the matter making up a typical human organism also makes up a certain nonorganism, and such things are what we are. Hume proposed that we are bundles of mental states and events. Our parts are not cells or atoms, he thought, but memories and dreams. We are not substances, but events or processes: Hume compares a person to a theater production. Advocates of this view need to explain which mental states make up a person, or a conscious being more generally: what ties the bundles together, as it were. They must also defend the baffling claim that a bundle of mental states is the sort of thing that can think or be conscious. A few philosophers espouse the paradoxical view that we don’t really exist at all. When we say ‘I,’ we fail to refer to anything. Your atoms may be real enough; perhaps even the thoughts and experiences we call yours are real; but those atoms or mental states are not parts or states of any conscious being. To answer the personal-ontology question, we need to know what sort of thing it is – if
304
Self: Personal Identity
anything – that thinks and is conscious. When I wonder what I am, what is it that wonders? What Matters in Identity Then there is the question of the practical importance of facts about our persistence through time. We ordinarily care deeply about what happens to us in the future, differently from the way we care about the future welfare of others. Why is this? More specifically, what reason do we have for this special selfish concern? What justifies it? This is the question of what matters in identity. Imagine that surgeons are going to destroy all of you but your brain and then transplant that organ into my head, and that neither of us has any choice about this. The resulting person, we may suppose, will have all your memories and other psychological features and none of mine, and will think he is you. To make things simple, suppose that we are both entirely selfish. Which of us will have a reason to care about the welfare of the resulting being? Suppose he will be in terrible pain after the operation unless one of us pays a large sum in advance. Which of us will have a reason to pay? And why? (In the same way, we can ask whether the resulting being will be morally responsible for my actions or for yours – or perhaps both, or neither.) The answer may seem to turn simply on whether the resulting person would be you or I: if he is you, you have a reason to care about his welfare and I don’t; if he is me, the reverse is true. And the reason for this may seem blindingly obvious: I care in a special way about what happens to myself in the future simply because he is me. Each person has a special reason to care about her own future welfare, and hers alone. The only one whose future welfare I cannot rationally ignore is myself. Likewise, only I myself can be morally responsible for my past actions. What matters in identity, we might say, is identity. But some deny this. They say that I could have an entirely selfish reason to care about someone else’s welfare, for his own sake. The reason why I care, selfishly, about a certain person’s future welfare is not that he is me, but that he relates to me in some other way. It may be, for instance, that
I care selfishly about a certain future person because he has inherited my mental life, and can remember my actions and no one else’s. Ordinarily, the only person who has inherited my mental life and can remember my actions is me, and so I ordinarily care selfishly only about myself. But the fact that he is me is not the reason why I care about him. I care about him because of his psychological connections with me. If someone else, other than me, were connected with me in that psychological way, he would have what matters in identity, and I ought to transfer my selfish concern to him. Perhaps that person would then also be responsible for my actions. We will return to this theme in a later section.
Understanding the Persistence Question Let us turn now to the persistence question: what is necessary and sufficient for a past or future being to be you, or what it takes for a human person to continue existing from one time to another. The concept of identity over time, or persistence, is a notorious source of confusion, and the persistence question is often misunderstood. The question asks what it takes for a being existing at another time to be numerically identical with you. For this and that to be numerically identical is for them to be one and the same: one thing rather than two. Numerical identity is different from qualitative identity or exact similarity: identical twins may be qualitatively identical (nearly), but not numerically identical. A past or future person need not, at that past or future time, be exactly like you are now in order to be you, that is, in order to be numerically identical with you. You don’t remain qualitatively the same throughout your life. You change. Of course, someone might say (as Hume did) that a past or future being could not be you unless he or she were then qualitatively just like you are now. This would mean that no person could survive any change whatever: even blinking your eyes would be fatal, resulting in your ceasing to exist and being replaced by someone else. In that case there would be no point in asking the persistence
Self: Personal Identity
question. Virtually all discussions of personal identity over time assume that it is possible for a person to change. This point enables us to distinguish the persistence question from another one that sounds similar. We sometimes ask what it takes for someone to remain the same person from one time to another – as opposed to simply asking what it is for someone to continue existing. If I were to change in certain ways – if I lost a great deal of memory, say, or my personality changed dramatically – we might describe this by saying that I should no longer be the person I was before, but a different one. The question of what it takes for someone to remain the same person in this sense is not the persistence question. It is not about numerical identity at all. If it were, it would answer itself, for I could not possibly come to be a numerically different person from the one I am now. Nothing can start out as one thing and end up as another thing: a thing can change qualitatively, but it cannot change numerically. This has nothing to do with personal identity in particular, but reflects the logic of identity in general. Those who say that certain events could make you a different person from the one you were before mean that you would still exist, but would have changed qualitatively in some profound and important way. They mean that you might come to be a radically different kind of person. If the resulting person were not numerically identical with you, it would not be the case that you yourself were a different person; rather, you would have ceased to exist and been replaced by someone else.
Proposed Answers to the Persistence Question There have been three main sorts of answers to the persistence question. The most popular, the psychological-continuity view, says that some psychological relation is both necessary and sufficient for one to persist. You are that future being that in some sense inherits its mental features – beliefs, memories, preferences, the capacity for rational thought and consciousness, or the like – from
305
you; and you are that past being whose mental features you have inherited in this way. This view comes in different versions, varying in the sort of inheritance that figures in them: whether a future being, in order to be you, needs to acquire its mental features from you via a continuously functioning brain, for instance, or whether some less direct transfer, such as Star Trek teleportation, might suffice. They also vary in the sort of mental features that need to be inherited – whether mental contents, such as memories, or core mental capacities such as the capacity for thought and consciousness, for instance. And they vary in how much of your mental life they require to be preserved in order for you to survive. Presumably it would not suffice, for some future person to be me, for him to inherit only one of my belief-states. But how much is enough? Yet another disagreement has to do with fission cases, of which more later. Most philosophers writing on the persistence question since the mid-twentieth century have followed Locke in endorsing some version of the psychologicalcontinuity view. The second answer is the brute-physical view or bodily criterion, according to which our identity through time consists in some brute physical relation. You are that past or future being that has your body, or that is the same biological organism as you are, or the like. Whether you survive or perish has nothing to do with psychological facts. This is typically held in combination with the view that we are biological organisms (animalism), for organisms, including human animals, appear to persist by virtue of brute physical facts. Here too there are variants. Hybrid views are also possible: one might propose that we need both mental and physical continuity to survive, or that either would suffice without the other. For present purposes we can treat these as versions of the psychologicalcontinuity view. Both the psychological-continuity and brutephysical views agree that there is something that it takes for us to persist – that our identity through time consists in or necessarily follows from something other than itself. A third view denies this. Mental and physical continuity are evidence for
306
Self: Personal Identity
our persistence, it says, but do not always guarantee it, and are not required. No sort of continuity is both absolutely necessary and completely sufficient for us to persist. There are no informative and nontrivial persistence conditions for people. This is called the simple view or anticriterialism. It is often, though not always, combined with the view that we are immaterial souls.
The Psychological-Continuity View The psychological-continuity view is intuitively appealing. Suppose your brain, or your cerebrum, were transplanted into my head, resulting in a being with most of your mental features and few if any of mine. He would, of course, believe that he was you and not me. Many people find it obvious that he would be right, precisely because of his psychological connections with you. (The brutephysical view, by contrast, seems to imply that he would be me, and that you would stay behind with an empty head if your cerebrum were transplanted. Most people find this strongly counterintuitive.) But it is notoriously difficult to get from this conviction to a plausible answer to the persistence question. What psychological relation might our identity through time consist in? Memory seems to play an important role: the one who got your transplanted brain would seem to be you at least partly because he or she would remember your life. So one version of the psychological-continuity view is that a past or future being is you just in the case that you can now remember an experience that being had then, or that being can then remember an experience you are having now. More precisely, Necessarily, a person x existing at time t is identical to a being y existing at another time, t *, if and only if x can remember, at t, an experience y has at t* or y can remember, at t*, an experience x has at t.
This sort of view, the memory criterion, is often attributed to Locke (though it is doubtful whether he or anyone else actually held it). The memory criterion faces two well-known problems. First, suppose a young student is arrested for drunken excesses. As a middle-aged neuroscientist, she retains a vivid memory of this event. In
her dotage, however, she remembers her science career, but has entirely forgotten the arrest and all the other events of her youth. According to the memory criterion, the young student would be the middle-aged scientist, the scientist would be the old woman, but the old woman would not be the young student. This is an impossible result: if x and y are one and y and z are one, x and z cannot be two. The second problem is that it seems to belong to the very idea of remembering an experience that you can remember only your own. You can no more remember someone else’s experiences than you could be a married bachelor. To remember being arrested (or the experience of it) is to remember yourself being arrested. That makes it trivial and uninformative to say that you are the person whose experiences you can remember – that memory continuity is sufficient for personal identity. It is uninformative because you could not know whether someone genuinely remembered a past experience without already knowing whether she was the one who had it. Suppose we want to know whether Blott, who exists now, is the same as Clott, whom we know to have existed at some time in the past. The memory criterion tells us that Blott is Clott if Blott can now remember an experience of Clott’s that occurred at that past time. But even if Blott seems to remember one of Clott’s experiences from that time, this counts as genuine memory only if Blott really is Clott. We should already have to know who is who before we could apply the theory that is supposed to tell us. (Note, however, that this is a problem only for the claim that memory connections are sufficient for identity, not for the claim that they are necessary. There is nothing trivial or uninformative in saying that a future being can be you only if he or she can then remember an experience you are having now.) One response to the first problem is to modify the memory criterion by switching from direct to indirect memory connections: the old woman is the young student because in her old age she can recall experiences the scientist had at a time when the scientist remembered the student’s life. The second problem is traditionally (and controversially) met by inventing a new concept, quasimemory, which is just like memory but without the identity requirement: even if it is self-contradictory to say that
Self: Personal Identity
I remember an experience of someone else’s, I could still quasiremember it. Neither move gets us far, however, for even the modified memory criterion faces a more obvious objection: there are many times in my past that I cannot remember or quasiremember at all, and to which I am not linked even indirectly by an overlapping chain of memories. There is no time when I could recall anything that happened to me while I was dreamlessly sleeping last night. The memory criterion has the absurd implication that I did not exist at any time when I was completely unconscious. A better solution appeals to causal dependence. We can define two notions, psychological connectedness and psychological continuity. A being is psychologically connected, at some future time, with me as I am now just if he is in the psychological states he is in then in large part because of the psychological states I am in now. Having a current memory (or quasimemory) of an earlier experience is one sort of psychological connection – the experience causes the memory of it – but there are others. Importantly, one’s current mental states can be caused in part by mental states one was in at a time when one was unconscious: for example, most of your current beliefs are the same ones you had while you slept last night. We can then define the second notion thus: I am now psychologically continuous with a past or future being just if my current mental states relate to those he is in then by a chain of psychological connections. This enables us to avoid the most obvious objections to the memory criterion by saying that a person who exists at one time is identical with something existing at another time if and only if the first is, at the first time, psychologically continuous with the second as she is at the second time.
Fission A serious difficulty for the psychologicalcontinuity view is the fact that you could be psychologically continuous with two future people at once. If your cerebrum were transplanted, the resulting person would be psychologically continuous with you (even if, as recent neuroscience has
307
shown, there would also be important psychological differences). If we destroyed one of your cerebral hemispheres, the resulting being would also be psychologically continuous with you. (Hemispherectomy – even the removal of the left hemisphere, which controls speech – is considered a drastic but acceptable treatment for otherwiseinoperable brain tumors.) What if we did both at once, destroying one of your cerebral hemispheres and transplanting the other? Then too, the one who got the transplanted hemisphere would be psychologically continuous with you, and according to the psychological-continuity view he or she would be you. But now let both hemispheres be transplanted, each into a different empty head. The resulting beings – call them Lefty and Righty – will each be psychologically continuous with you. The psychological-continuity view as we have stated it implies that any future being who is psychologically continuous with you must be you. It follows that you are Lefty and also that you are Righty. But that cannot be, for Lefty and Righty are two, and one thing cannot be numerically identical with two things. (If you and Lefty are one, and you and Righty are one, Lefty and Righty cannot be two.) This is the fission problem. Psychological-continuity theorists have proposed two different solutions to this problem. One, sometimes called the multiple-occupancy view, says that if there is fission in your future, then there are, so to speak, two of you even now. What we think of as you is really two people, who are now exactly similar and located in the same place, doing the same things and thinking the same thoughts. The surgeons merely separate them. The multiple-occupancy view is almost invariably combined with the thesis that people and other persisting things are made up of temporal parts. For each person there is, for instance, such a thing as her first half, which is just like the person only briefer – something analogous to the first half of a football match. On this account, the multipleoccupancy view is that Lefty and Righty coincide before the operation by sharing their preoperative temporal parts, and diverge later by having different temporal parts located afterward. Lefty and Righty are like two roads that coincide for a stretch and then fork, sharing some of their spatial parts
308
Self: Personal Identity
but not others. At the places where the roads overlap, they will look just like one road. Likewise, the idea goes, at the times before the operation when Lefty and Righty share their temporal parts, they will look just like one person – even to themselves. Whether people really are made up of temporal parts, however, is a disputed metaphysical question. The other solution to the fission problem abandons the intuitive claim that psychological continuity by itself suffices for one to persist. It says, rather, that you are identical with a past or future being who is psychologically continuous with you only if no other being is then psychologically continuous with you. (There is no circularity in this. We need not know the answer to the persistence question in order to know how many people there are at any one time; that comes under the population question.) This means that neither Lefty nor Righty is you. They both come into existence when your cerebrum is divided. If both your cerebral hemispheres are transplanted, you cease to exist – though you would survive if only one were transplanted and the other destroyed. This proposal, the nonbranching view, has the surprising consequence that if your brain is divided, you will survive if only one half is preserved, but you will die if both halves are. That is just the opposite of what most of us expect: if your survival depends on the functioning of your cerebrum (because that is what underlies psychological continuity), then the more of that organ we preserve, the greater ought to be your chance of survival. In fact the nonbranching view implies you would perish if one of your hemispheres were transplanted and the other left in place: you can survive hemispherectomy only if the excised hemisphere is immediately destroyed. The nonbranching view makes the question of what matters in identity especially acute. Faced with the prospect of having one of your cerebral hemispheres transplanted, there would seem to be no reason to prefer that the other be destroyed. Most of us would rather have both preserved, even if they end up in different heads. Yet on the nonbranching view that is to prefer death over continued existence. Some philosophers infer from this that that is precisely what we ought to prefer. Insofar as we are rational, we don’t want to continue existing. Or, at least we don’t want it for
its own sake. I only want there to be, in the future, someone psychologically continuous with me, whether or not he is strictly me. Likewise, even the most selfish person has a reason to care about the welfare of the beings who would result from her undergoing cerebral fission, even if, as the nonbranching view implies, neither would be her. In the fission case, the sorts of practical concerns you ordinarily have for yourself seem to apply to someone who isn’t strictly you. This suggests more generally that facts about who is numerically identical with whom have no practical importance. What matters practically is, rather, who is psychologically continuous with whom.
The Beginning and End of Life Advocates of the psychological-continuity view commonly discuss its implications about who is who in imaginary science-fiction scenarios. But the view also has important implications about when we begin and end in real life. We ordinarily suppose that each of us was once an embryo and then a fetus. We suppose that the fetus my mother once carried was born, grew up, and later wrote this article. (This does not imply that a fetus in the womb is already a person, but at most that it is a potential person: something that can come to be a person. We might start out as nonpeople and only later become people, just as we are first children and later adults.) But the psychological-continuity view implies that none of us was ever an embryo. Whatever exactly psychological continuity amounts to, you can never be psychologically continuous with a being that has no psychological states at all: you cannot inherit your current psychological states from a being that has none to pass on. And the embryo from which you developed had no psychological states. If psychological continuity is necessary for you to persist, as the psychological-continuity view says, then you were never an embryo. The embryo from which you developed was not you. Not only was it not a person; it was not even something that could come to be a person. You did not come into being when the embryo did, but some time later. How much later? The psychological-continuity view implies that you could not have existed at a
Self: Personal Identity
time when you had no psychological states or properties at all. The earliest you could have begun to exist is when the fetus you developed from acquired its first mental states. No one is certain when that was, but embryologists agree that the fetus has no mental states – not even the capacity to feel pain – before mid-gestation at the earliest. So the psychological-continuity view implies that none of us was ever a 4-month-old fetus. The fetus you see in an ultrasound scan is not something that can one day come to be a person. According to the psychological-continuity view, when you and I come into existence depends on what sort of psychological continuity our identity over time consists in. Many psychological-continuity theorists say that we do not come into being when the first mental capacities develop in the course of fetal development, but only with the advent of more sophisticated mental properties, such as the capacity for self-consciousness. On their view, we do not begin to exist until we are born, or perhaps several months or even years afterward. The psychological-continuity view also has important implications about when we come to an end. It implies, for instance, that you could not exist in a persistent vegetative state. A being in such a state has no mental states, and thus has not in any way inherited yours. There is complete psychological discontinuity between you as you are now and the human vegetable that would result if you were to lapse into such a state. If all your mental states were destroyed, you would cease to exist, even if the rest of you kept on working as before. The resulting being, despite being alive by most definitions of the word (and in all current legal systems), would not be you. If the doctors withdraw the tubes that feed it, they do not kill anyone, or allow anyone to die. Some being or other dies when the feeding tubes are withdrawn, of course, but not anything that was ever a person. This too goes against what many of us believe. The reason we tend to think that each of us was once a fetus and might end up in a vegetative state is that in each of these cases there is a good deal of brute physical continuity of an appropriate sort: the same sort of physical continuity we normally exhibit from one day to the next. So these considerations seem to support the brute-physical view as against the psychological-continuity view. More generally,
309
the brute-physical view appears to give the right verdicts about who is who in all real-life cases.
The Persistence Question and Personal Ontology We have been evaluating answers to the persistence question by considering their implications about who is who in cases where they disagree. We saw that the psychological-continuity view gives attractive results in science fiction stories (such as the cerebrum transplant), but has implausible consequences about when we begin and end in real life. With the brute-physical view it is the other way round. But this sort of debate over cases is not the only relevant consideration. A good answer to the persistence question must also be compatible with an acceptable answer to the personal-ontology question. What sort of things could we be if the psychological-continuity view were true? That is, what sort of things could persist by virtue of psychological continuity? The psychological-continuity view appears to rule out our being organisms. It says that our persistence consists in some sort of psychological continuity. As we have seen, this means that you would go along with your transplanted cerebrum, because the one who ended up with that organ, and no one else, would be psychologically continuous with you. And it implies that if you were to lapse into a persistent vegetative state, you would cease to exist, because no one would then be psychologically continuous with you. But the persistence of a human organism does not consist in any sort of psychological continuity. If we transplanted your cerebrum, the human organism – your body – would not go along with that organ. It would stay behind with an empty head. If you were an organism, then you would stay behind with an empty head, contrary to the psychological-continuity view. Likewise, no human organism ceases to exist by lapsing into a persistent vegetative state. If you were an organism, you could survive as a human vegetable. What the psychological-continuity view says about our persistence through time is not true of human organisms. So if that view is true, we could not be organisms.
310
Self: Personal Identity
Psychological-continuity theorists must deny that we are organisms, and find another account of what we are. Some say that we are temporal parts of organisms. (If your cerebrum were transplanted, you would made up of temporal parts of two organisms.) Others say that we are material things made of the same matter as organisms, or bundles of mental states. But the view that we are not organisms faces an awkward problem.
The Too-Many-Minds Problem Even if you are not an organism, your body is. That organism – a human animal – thinks and is conscious. In fact, it would seem to be psychologically indistinguishable from you. So if you are not that animal, but something else, it follows that there is a conscious, intelligent being other than you, now sitting in your chair and thinking your thoughts. This means that there are at least twice as many thinking beings as the census reports: for each of us, there is another thinking, conscious being, namely the animal we call one’s body. Worse, you ought to wonder which of the two thinkers is you. You may believe that you are the nonanimal (because you accept the psychological-continuity view, perhaps). But the animal has the same grounds for believing that it is a nonanimal as you have for supposing that you are. Yet it is mistaken. For all you know, you might be the one making this mistake. If you were the animal and not the person, you would never be any the wiser. An analogy may help here. Imagine a threedimensional duplicating machine. When you step into the in box, it reads off your information and assembles a perfect duplicate of you in the out box. The process causes temporary unconsciousness, but is otherwise harmless. Two beings wake up, one in each box. The boxes are indistinguishable. Because each being will have the same apparent memories and perceive identical surroundings, each will think that he or she is you, and will have the same evidence for this belief. But only one will be right. How could you ever know, afterward, whether you were the original or the newly created duplicate? In the same way, the psychological-continuity view appears to leave you without any grounds for
supposing that you are the being who persists by virtue of psychological continuity, rather than the animal. So even if the psychological-continuity view is true, it seems that you could never know whether it applied to you: for all you can tell, you may instead be an organism with brute physical persistence conditions. This is the ‘too-manyminds’ or ‘too-many-thinkers’ problem. It threatens to make any view according to which we are not organisms look absurd. Only animalism, the view that we are organisms (and that there are no beings who persist by virtue of psychological continuity), appears to escape it. Psychological-continuity theorists have proposed two ways of solving the problem. One is to argue that despite appearances, human animals are not psychologically indistinguishable from us. Although our animal bodies share our brains, are physically just like us, and show all the outward signs of consciousness and intelligence, they themselves do not think and are not conscious. If human organisms cannot be conscious, then presumably no biological organism of any sort could have any mental properties at all. Why not? It may be, as Descartes and Leibniz argued, because organisms are material things. Only an immaterial thing could think or be conscious. You and I must therefore be immaterial. Though this would solve the too-many-thinkers problem, it raises many others, and few philosophers nowadays accept it. One notable attempt to explain why organisms should be unable to think that is compatible with our being material things, appeals to the nature of mental states and properties. It says that whatever thinks or is conscious must persist by virtue of psychological continuity. That is because it belongs to the nature of a mental state that it tend to have certain characteristic causes and effects in the being that is in the state, and not in any other being. (This is a version of the functionalist theory of mind.) For instance, your preference for chocolate over vanilla must tend to cause you, and no one else, to choose chocolate. Now if an organism were to have such a preference, that state might cause another being to choose chocolate, for an organism’s cerebrum might be transplanted into another organism. That would violate the proposed account of mental
Self: Personal Identity
states. It follows, on that account, that no organism could have a preference; and similar reasoning goes for other mental states. The persistence conditions of organisms are incompatible with their having mental properties. But a material thing that would go along with its transplanted cerebrum – a being of which the psychological-continuity view was true – could have mental states. It would follow that you and I, who obviously have mental states, persist by virtue of psychological continuity, and thus are not organisms. This would solve the ‘toomany-thinkers’ problem, and at the same time show that the psychological-continuity view is true. It is, however, a minority view. The other alternative for psychologicalcontinuity theorists is to concede that human organisms think as we do, so that you are one of two beings now thinking your thoughts, but try to explain how we can still know that we are not those organisms. One strategy for doing this focuses on the nature of personhood and firstperson reference. It proposes that not just any being with mental properties of the sort that you and I have – rationality and self-consciousness, for instance – counts as a person. A person must also persist by virtue of psychological continuity. It follows from this that human animals, despite being psychologically just like us, are not people. The proposal goes on to say that personal pronouns such as ‘I’ refer only to people. This means that when your animal body says or thinks ‘I,’ it does not refer to itself. Rather, it refers to you, the person who says it at the same time. When the animal says ‘I am a person,’ it does not thereby express the false belief that it is a person, but rather the true belief that you are. It follows that the animal is not mistaken about which thing it is, because it has no first-person beliefs about itself at all. And you are not mistaken either. You can infer that you are a person from the linguistic facts that you are whatever you refer to when you say ‘I,’ and that ‘I’ never refers to anything but a person. You can know that you are not the animal thinking your thoughts because it is not a person, and personal pronouns never refer to nonpeople. Although this proposal avoids the surprising claim that organisms cannot have mental properties, it gives a highly counterintuitive view of what
311
it is to be a person. And it still implies that there are twice as many intelligent, conscious beings as we thought. It remains a controversial view.
Conclusion The answer to the persistence and personal-ontology questions is likely to depend on more general metaphysical matters. Which account of personal identity you find attractive will depend on your view about the metaphysics of material things in general. Consider, for example, the view that all persisting things are made up of temporal parts, mentioned earlier. As we saw, psychological-continuity theorists who hold this view can say that if your cerebrum were divided and each half transplanted into a different head, you would survive twice over, as it were. They need not accept the nonbranching view, according to which you would cease to exist, but would survive if just one cerebral hemisphere were preserved. So temporal-parts theorists will find the psychological-continuity view more attractive than those who reject temporal parts. The temporal-parts theory also implies that animalists face their own version of the toomany-thinkers problem. Suppose you are an animal, and that you are made up of temporal parts. Then there is a temporal part of you that is just like you are now, except that it extends in time only from midnight last night until midnight tonight: your today-part. It is now sitting in your chair, fully conscious and thinking your thoughts. How could you ever know that you are not your today-part, and have only hours to live? This looks just as troubling for the animalist as the problem of how you can know you are not your animal body is for the psychological-continuity theorist. And any solution the animalist can give to her version of the problem will suit the psychological-continuity theorist equally well for hers. (Temporal-parts theorists solve the too-manythinkers problem by appealing to the account of personhood and first-person reference sketched in the previous section.) So if the temporal-parts theory is true, the problems facing the psychological-continuity view appear considerably less than they would
312
Self: Personal Identity
otherwise be. Whether persisting things really do have temporal parts, however, is not a question about personal identity. It is a general question about the fundamental nature of the world. The main problems of personal identity cannot be debated in isolation. See also: Autobiographical Memory and Consciousness; The MindBody Problem; Philosophical Accounts of Self-Awareness and Introspection; Self: Body Awareness and Self-Awareness; Self: The Unity of Self, Self-Consistency.
Suggested Readings Baker LR (2000) Persons and Bodies: A Constitution View. Cambridge: Cambridge University Press. Garrett B (1998) Personal Identity and Self-Consciousness. London and New York: Routledge. Hudson H (2001) A Materialist Metaphysics of the Human Person. Ithaca, NY: Cornell University Press.
Kolak D and Martin R (eds.) (1991) Self and Identity: Contemporary Philosophical Issues. New York: Macmillan. Martin R (2006) The Rise and Fall of Soul and Self: An Intellectual History of Personal Identity. New York: Columbia University Press. Martin R and Barresi J (eds.) (2003) Personal Identity. Oxford: Blackwell. Noonan H (2003) Personal Identity. 2nd edn. London and New York: Routledge. Olson ET (1997) The Human Animal: Personal Identity Without Psychology. New York: Oxford University Press. Olson ET (2007) What are We? A Study in Personal Ontology. New York: Oxford University Press. Parfit D (1984) Reasons and Persons. Oxford: Oxford University Press. Perry J (ed.) (1975) Personal Identity. Berkeley, CA: University of California Press. Shoemaker S (1999) Self, body and coincidence. Proceedings of the Aristotelian Society Supplementary Volume 73: 287–306. Shoemaker S and Swinburne R (1984) Personal Identity. Oxford: Blackwell. Unger P (1990) Identity, Consciousness and Value. New York: Oxford University Press.
Biographical Sketch
Eric Olson was born in 1963 and grew up in Washington State, USA. He was educated at Reed College and Syracuse University. From 1995 to 2003 he was a lecturer in philosophy at the University of Cambridge, and is currently a professor of philosophy at the University of Sheffield.
Self: The Unity of Self, Self-Consistency R R Vallacher, Florida Atlantic University, Boca Raton, FL, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Emergence – Emergence occurs when the individual elements of a dynamical system achieve organization by means of their mutual influence. The higher-order properties that result from the mutual adjustment among lower-level elements provide coordination for the lower-level elements. The higher-level state is said to be emergent because it was not inherent in the properties of the lower-level elements and because it was not imposed on the system from forces outside the system. Emergence provides for substantial growth in the complexity of a system’s processes and properties. Because of emergence, very complex systems can often be described by very simple models. Self-concept certainty – The degree to which people have an internally consistent and confident assessment of themselves. Certainty can vary across different aspects of self-concept, although there is a tendency toward global certainty versus uncertainty that transcends different regions of self-structure. Self-concept certainty is the subjective experience of self-concept coherence and unity. Self-concept differentiation – The tendency for people to form different assessments of themselves in different roles or with respect to different characteristics and competencies. Differentiation results from the tendency toward progressive integration in self-structure. Because complete integration is impossible to attain, the press toward integration stalls at a level beneath global self-understanding. The differentiated aspects of self-concept reflect locally coherent and unified regions of the person’s self-structure.
Self-esteem – Global self-evaluation that transcends specific aspects of a person’s self-concept. Self-esteem provides the highest level of unity in self-understanding. It is the most investigated dimension of self-concept and has been related to many social psychological phenomena. However, self-esteem often lacks the specificity required for action and decision making with respect to particular domains of personal and interpersonal functioning. Self-evaluation maintenance – The tendency by people to protect their self-evaluation, either in global terms or with respect to specific dimensions of self-understanding. Self-evaluation maintenance is manifest as a wide variety of specific cognitive and behavioral mechanisms. These mechanisms are interchangeable and can be substituted for one another depending on their relative effectiveness in a given context. Because most people have a favorable self-concept, self-evaluation maintenance often represents an attempt to maintain high self-esteem. For people with an unfavorable self-assessment, though, self-evaluation maintenance may be manifest as an attempt to maintain low self-esteem – a tendency referred to as self-verification. Self-handicapping – A special case of self-evaluation maintenance characterized by a concern with avoiding circumstances that pose a risk to lowering the assessment of one’s self-perceived competencies. Self-handicapping is commonly conceptualized as any action or choice of performance setting that maximizes the opportunity for internalizing success or externalizing failure on a task relevant to one’s self-concept.
313
314
Self: The Unity of Self, Self-Consistency
Threatened egotism – A special case of self-evaluation maintenance characterized by intolerance of criticism and a potential for aggression against those who are perceived to provide criticism. This tendency is observed among people with high self-esteem but low self-concept certainty. It thus represents a defensive reaction to information that would undermine a person’s self-concept coherence.
Introduction The self is the largest structure in the individual’s mental system. Virtually every experience from infancy to the present moment has some degree of relevance for a person’s understanding of him or herself. The self is also the most chronically accessible mental structure. Even the most important people and issues may remain out of awareness for extended periods of time, but a person’s sense of self is never more than a stranger’s glance from becoming the focus of conscious attention. In view of the amount, diversity, and accessibility of selfrelevant information, the task of integrating this information to form a stable and unified self-concept would seem to be a daunting, if not an impossible task. Yet, people manage to develop a fairly coherent representation of themselves that provides a sense of personal continuity despite changing circumstances and that functions as a frame of reference for decision making and other forms of self-regulation.
Self-Concept Coherence Most theories of self-concept formation emphasize the internalization of self-relevant information derived from social feedback, social comparison, and various forms of socialization. This internalized representation stabilizes the person’s global self-view and enables him or her to respond selectively to new information experienced in daily life. The specifics of the internalization process, however, are not well defined in most accounts. In recent years, greater specificity has been provided by models built on principles of complexity and
dynamical systems. Support for this class of models is provided by computer simulations and, to a lesser extent, by new lines of empirical research. In this approach, the generation of a relatively coherent self-concept conforms to the same general rules that are responsible for the emergence of higher order states in a wide variety of complex systems in nature and the physical sciences. Self-Organization A person’s self-concept represents a large and diverse set of elements of self-relevant information derived in the course of his or her social interactions and other experiences. In a single day, for example, a person may receive explicit or implicit feedback about his or her actions or appearance from several people, with each instance of feedback conveying a somewhat different perspective regarding the person’s characteristics. From day to day, week to week, and year to year, there is a steady onslaught of such feedback that is recognized and processed in some fashion by the person. If the myriad elements of self-relevant information existed in isolation from one another, each specific self-relevant thought that arose in consciousness might subvert the platform for action established by the preceding self-relevant thought. The person’s sense of self would be highly fragmented, providing an erratic frame of reference, at best, for interpreting subsequent experiences and deciding among action possibilities. The elements of self-relevant information do not exist in isolation, but rather are compared, adjusted, and assembled into higher-order units, each providing a semblance of integration in self-understanding. The self-concept, in other words, represents a relatively unified structure within which specific thoughts, images, and memories pertaining to the self become mutually supportive. The process of self-organization, a fundamental feature of contemporary accounts of complexity and dynamical systems, provides a plausible mechanism for the integration of self-relevant information. In a dynamical system consisting of interconnected elements, the state of each element adjusts to the current state of other elements to which it is connected. Because of this mutual influence, elements that are initially different achieve commonality with respect to some parameter.
Self: The Unity of Self, Self-Consistency
As a result, higher-order units are created that represent, and provide coordination for, the set of individual elements. No outside agent or external influence is required for such order to emerge. Hence, the process is referred to as selforganization. The attainment of self-concept coherence by means of self-organization is possible because the elements of self-structure, though diverse by many criteria, can be scaled with respect to the common parameter of evaluation. Self-relevant information runs the gamut of possible self-evaluation, from memories of misdeeds and perceived character flaws to memorable accomplishments and firmly held values. The elements of self-structure do not have a fixed valence, however. Rather, the selfevaluation associated with an element of selfrelevant information is malleable and open to influence from the valence of other elements. An element incongruent with neighboring (thematically related) elements may change its valence or change the valence of its neighbors, so as to establish evaluative coherence with the related elements. ‘Distractable,’ for example, may initially be viewed as a negative feature of the self, but take on positive valence in the context of other selfperceived qualities that together convey an image of oneself as a creative scientist. Whereas individual thoughts about the self may be malleable, the self-organization of many thoughts into higher-order units (e.g., personality traits, competencies, and social roles) provides a coherent sense of the self that promotes stability in thought, emotion, and action. When confronted with new and incongruent self-relevant information, the mutual influences among the elements comprising a higher-order unit support one another and thus promote resistance to the potential disconfirmation. In more mechanistic terms, a stable and coherent self-structure emerges because the separate elements of self-relevant information influence each other via multiple feedback loops. Congruent elements provide cross-validation for each other, whereas incongruent elements set in motion mechanisms designed to eliminate the incoherence or isolate the incongruent elements. In this scenario, no higher order agent or homunculus is necessary to dictate the integration of selfrelevant information.
315
Self-Concept Differentiation Once a higher-order unit of self-understanding emerges, it takes on the status of an element that is amenable to integration with other higher-order units. If such integration occurs, the resultant higher-order unit functions as an element that can be then integrated with other functional units, and so on, in a scenario of progressive integration. A person, for example, may initially integrate specific thoughts about him or herself in an academic context in developing a coherent selfview as a ‘good test-taker.’ This emergent self-view might then be integrated with other locally coherent self-views (‘good study habits,’ ‘good termpaper writer,’ etc.) to promote a yet higher-order self-concept as a ‘good student.’ The ‘student’ selfview might then become integrated with coherent self-views in other domains (e.g., employee, friend, son, or daughter) to generate a yet higher-order sense of self. In principle, the tendency for progressive integration could result in a globally coherent self-concept, with elements of self-relevant information across the entire self-structure providing cross-validation for one another. In reality, complete unity in self-concept is unlikely to be observed. Rather, the press for progressive integration is likely to stall at some level, promoting islands of local coherence. Some higher-order aspects of the self may be difficult or impossible to integrate, or they may rarely become salient at the same time and in the same context. A person may have a coherent and positive view of him or herself as a parent, for example, but this self-view may be compartmentalized and not come into contact with other self-views, such as athlete or science fiction reader. Because of the limits to progressive integration, the self-structure becomes differentiated rather than globally unified, with different sets of elements stabilizing on different values of self-evaluation. A person may have a coherent and positive view of him or herself as a parent, for example, and an equally coherent but less flattering view of him or herself as an athlete. The emergence of a differentiated (locally coherent) self-structure has been modeled on a cellular automata platform. Cellular automata are programmable dynamical systems. In this approach, a finite
316
Self: The Unity of Self, Self-Consistency
set of elements is specified, each of which can adopt a finite number of discrete states. The elements are arranged in a specific spatial configuration that usually takes the form of a two-dimensional lattice or grid. The location of each element on this grid specifies the element’s neighborhood. The elements evolve in discrete units of time, such that the state of an element at t + 1 depends on the states of the neighboring elements at time t. The dynamics of cellular automata depend on the nature of the updating rule and on the format of the grid dictating the neighborhood structure. Repeated iterations of the updating rule invariably produce locally coherent clusters of elements sharing a common state. The application of cellular automata to the emergence of self-structure is straightforward. A person’s self-concept is conceptualized as a system consisting of n elements, each reflecting a specific aspect of the self, which are represented as cells arranged on a two-dimensional grid (see Figure 1). The physical proximity of any two elements represents their degree of relatedness. Each element is characterized with respect to its current evaluation, which is either positive
(a)
(b) Figure 1 The emergence of coherence in self-
structure by means of self-organization. Reproduced from Nowak A, Vallacher RR, Tesser A, and Borkowski W (2000) Society of self: The emergence of collective properties in self-structure. Psychological Review 107, 39–61, with permission from American Psychological Association (APA).
(denoted by light gray) or negative (dark gray). Some elements are more important than others and have greater weight in self-evaluation. An element’s importance, which remains constant in the course of simulation, is denoted by its height. Each element influences and is influenced by its eight neighboring elements (four on the adjacent sides and four on the connecting diagonals). In the course of simulation, an element chosen at random tries to adjust to its neighboring elements by checking how much influence it receives from them. This process involves weighting the valence of each neighbor by the neighbor’s importance. The weighted sum of evaluations of the neighboring elements that results from this process is then compared to the current valence of the element. If the element’s valence agrees with the overall evaluation suggested by its neighbors, the valence does not change. If the valence of the element differs from the overall evaluation suggested by its neighbors, the element changes evaluation only if the combined weight of evaluation from the neighboring elements is greater than the element’s own weighted evaluation. This means that is relatively easy for neighboring elements to change the evaluation of a relatively unimportant element, but it is difficult to change the evaluation of a more important piece of self-relevant information. This process is repeated for another randomly chosen element, and then again for another element, and so on, until each element has been chosen. In the next simulation step, each element has a chance to adjust its state again. The simulation steps are repeated until the system reaches an asymptote, indicating no further changes in the state of elements (i.e., static equilibrium) or a stable pattern of changes in the system (i.e., dynamic equilibrium). The process of mutual adjustment among elements promotes the emergence of clusters, as depicted in Figure 1. Self-relevant information that is randomly distributed at the outset (Figure 1(a)) forms well-defined domains composed of elements that share a common valence (Figure 1(b)). The selfsystem also becomes more polarized in overall evaluation, with more negative elements switching to positive valence than vice versa. In a disordered system, the proportion of positive and negative elements in a region roughly corresponds to the
Self: The Unity of Self, Self-Consistency
proportion of positive and negative elements in the entire structure. Hence, any given element is likely to be surrounded by more positive than negative elements and thus is likely to experience greater influence in the positive direction. Once the self-structure has become clustered, however, most elements are surrounded by elements of the same valence, so that only the elements on the border of a cluster are subjected to conflicting influences. The emergence of locally coherent regions tends to stabilize the self-system. The joint influence of all the elements in a region support the current state of any single element, which promotes resistance to change and enables an element to return to its original value if it is changed. In an incoherent region, however, the current state of an element is supported by some elements but undermined by others. Hence, when an element is influenced from the outside, some of the surrounding elements help the element resist the influence, whereas other elements work in the direction of the influence. If an element is overwhelmed by outside influence, there will be little tendency for it to return to its original state because some of the neighboring elements are likely to support the new state. Global Self-Evaluation Despite the limits to global self-understanding, people are able to characterize themselves with respect to an overall self-evaluation. Even if a person makes a clear distinction between his or her competence in parenting and athletics, for example, he or she is likely to experience little hesitation in answering questions pertaining to his or her general self-worth compared to other people. Global self-evaluation – commonly referred to as self-esteem – is in fact the most investigated aspect of self-concept. Numerous tests have been developed to measure self-esteem and each has shown that people can be reliably scored on this trait. People differ in their respective level of selfesteem (although most people consider themselves ‘better than average’) and their level of self-esteem is relatively stable over time. There are two ways of reconciling self-concept differentiation and the predilection for global selfevaluation. One account suggests that people simply average the self-evaluations associated with
317
specific regions of the self to come up with a global sense of self-worth. A person might feel very positive about his or her intellectual ability, moderately positive about his or her social skills, and have a negative assessment of his or her athletic competence. If the person simply averaged these selfassessments (excellent, good, and bad), his or her self-esteem would be moderately positive. Some self-aspects are more important than others, so variations on this perspective suggest that greater weight is given to the more important aspects when generating a global self-evaluation. The person depicted above should express fairly high selfesteem if he or she attached greater importance to intellectual prowess but fairly low self-esteem if he or she attached greater weight to athletic prowess. In the other account, self-esteem is assumed to be independent of specific self-views, reflecting instead a basic appraisal of oneself that derives from the nature of one’s social experiences and relationships. A secure childhood, for example, may establish an unquestioned view of oneself as accepted and desirable, regardless of how one’s specific competencies are evaluated. Conversely, a childhood associated with neglect and parental inconsistency may create a foundation for self-doubt that persists throughout life, even if the person develops competencies in many areas of life. A recent variation on this idea is the sociometer hypothesis, which assumes that humans have a fundamental need to belong to social groups and to have meaningful social relations. A person’s global self-evaluation reflects the extent to which these social belongingness needs are fulfilled: the more completely the person’s needs are met, the higher the person’s self-esteem. In support of the sociometer hypothesis, high and low self-esteem people do not differ in many ways other than in their respective social skills and social relations. High self-esteem people are not necessarily brighter, more athletic, more skilled, more physically attractive, or even healthier than are people with lower self-esteem. These findings, however, represent generalizations across large samples of people. For any one person, one or more of these dimensions could be vital to his or her self-esteem. But if these self-defining dimensions differ across people, one would not expect to find a connection in the population between any single self-view and people’s level of self-esteem.
318
Self: The Unity of Self, Self-Consistency
Self-Concept Certainty Although there is a tendency toward integration in self-concept, not everyone achieves the same level of differentiation. Some people achieve a fairly integrated self-structure composed of coherent regions corresponding to roles, broad traits, competencies, and values. But other people do not attain this level of integration and have an overly differentiated and fragmented self-concept. People with an optimally differentiated view of themselves have a coherent view of themselves in different roles and settings, and this stable way of thinking about themselves provides a clear sense of how they should behave in different contexts. People with a fragmented sense of self, however, cannot maintain a stable view of themselves, are torn by indecision and ambivalence, and are very susceptible to the feedback about themselves provided by others. Everyone has negative thoughts and feelings about the self, but people with a differentiated self-concept can compartmentalize these elements into separate regions. People who have a fragmented self-concept, however, cannot think about any aspect of the self without encountering a mix of positive and negative thoughts. An unintegrated sense of self is experienced as self-concept uncertainty. When asked to indicate what they are like on trait dimensions (e.g., sociable vs. unsociable, energetic vs. lazy), people lacking integration can provide an answer (moderately sociable, fairly energetic, and so on). However, when they are asked later to indicate how certain they are about their trait ratings, these people tend to express uncertainty. They may express fairly high certainty when rating their global self-esteem, but when asked to reflect on themselves for a period of time, their moment-to-moment selfevaluation tends to be erratic, revealing the unintegrated state of their self-structure. People with a more certain view of themselves, in contrast, are likely to express thoughts about themselves that are relatively stable in evaluation over time. They may express negative self-evaluative comments, but these comments tend to be clustered in time, separated from periods during which they express primarily positive comments about themselves. There are two ways to think about self-concept uncertainty. Perhaps people with an uncertain
sense of themselves simply do not have a clear view of what they are like and entertain a wide range of possibilities, none of which prove particularly convincing or provide a great deal of stability. The other perspective is that uncertainty does not reflect an erratic view of the self, but rather two (or more) fairly coherent self-views that are mutually inconsistent in the self-evaluation they suggest. In both cases, the person can only be certain of the basic actions he or she has done or is inclined to do. The difference centers on what the person does with this low-level information. In the first case, the person does not know how to integrate these elements into a coherent view of the self; in the second case, the person has two (or more) ways of integrating this information, and these conflicting self-views vie for ascendance and promote ambivalence in self-evaluation. Insight into this issue is provided by research employing the implicit association test (IAT), a measurement procedure that exposes unconscious attitudes that go undetected by traditional procedures that ask people to provide explicit reports of their attitudes. A white person might indicate positive attitudes toward members of a minority group, for example, but harbor a stereotyped view of the group that is latent and unexpressed most of the time. The IAT has been employed in recent years to expose implicit self-esteem – people’s latent and unexpressed self-evaluation. When asked to evaluate themselves on self-report measures such as the Rosenberg self-esteem test, most people portray a relatively positive (‘better than average’) self-image. Despite having high explicit self-esteem, however, some people harbor a negative self-view that is out of conscious awareness (and expression) most of the time. A person’s implicit self-esteem is revealed in the IAT procedure by assessing how long it takes him or her to associate ‘self ’ words (I, my, me, and mine) with pleasant words (e.g., kind and wonderful) versus unpleasant words (e.g., filthy and stinky). If the person takes considerably longer to associate the self words with unpleasant than with pleasant words, his or her implicit self-esteem is high. But if the person takes relatively little time to associate the self words with the unpleasant words, he or she is said to have implicit self-esteem that is low.
Self: The Unity of Self, Self-Consistency
Explicit and implicit self-esteem are correlated but not very strongly, so some people can be characterized as having high explicit self-esteem and low implicit self-esteem. There is evidence that this combination of conflicting self-views underlies self-concept uncertainty. People characterized by a discrepancy between explicit and implicit self-esteem tend to show relatively high variability over time in global selfevaluation when they generate stream-of-thought narratives about themselves. Thus, a positive selfevaluative comment is as likely to be followed by a negative self-assessment as by another positive comment. As noted above, such variability in momentary self-evaluation is associated with low self-concept certainty. In contrast, people whose implicit selfesteem is congruent with the level of self-regard expressed on an explicit self-report measure tend to show relatively low moment-to-moment variability in their verbalized self-evaluative statements when asked to reflect on themselves for several minutes. Positive comments tend to be followed by other positive comments, and negative self-evaluative remarks, although less frequent, tend to be followed by other negative self-evaluative remarks.
Maintenance of Self-Concept Coherence Everyday life provides a continuous influx of selfrelevant information, and not all of it is favorable or mutually consistent. Some instances of social feedback and social comparison can provide a direct challenge to a specific self-view (e.g., one’s social charm) or to global self-regard. Yet people usually recover from inconsistent self-relevant information and maintain their self-view in specific areas as well as their overall self-esteem. As noted earlier, self-relevant information does not have a fixed evaluation, but rather is open to interpretation. Behaving in a critical fashion, for example, is often considered negative (e.g., as harmful or insensitive), but it can also be viewed in positive terms (e.g., as constructive criticism or ‘tough love’). The flexibility of self-relevant information, which is instrumental in the formation of a self-concept, is also critical to the defense of a selfconcept once it has formed.
319
Self-Evaluation Maintenance People are motivated to protect their global selfesteem, as well as their specific self-views in different regions of self-structure, from events or new information that would call the validity of their self-assessments into question. Because self-concept is inherently an evaluative structure, its protection is tantamount to maintaining a characteristic selfevaluative assessment. People prefer to think about themselves in positive rather than in negative terms, and most people develop a generally flattering view of their characteristics, seeing themselves as above average on many different dimensions. Hence, the maintenance of a self-concept for most people is synonymous with maintaining a relatively high level of self-esteem. When confronted with negative self-relevant information, then, people are motivated to process the disconfirmatory information in a manner that renders it consistent with the positive tone of the self-view at issue. There are a variety of means by which people can change the implications of self-relevant information for purposes of self-protection. Consider, for example, a person with a positive self-assessment of his or her social skills who is called a crashing bore at a party. The person might experience a temporary setback in his or her self-regard, but is likely to mount a defense to restore his or her self-view in this area. He or she might challenge the critic’s credibility, for example, noting that the critic is simply a negative person or perhaps a bad judge of character. Alternatively, he or she might consider the behavior that prompted the critique to be unrepresentative of the way he or she normally behaves in this context. Yet another line of defense is to question the critic’s motive for the unflattering feedback. The person might conclude that the critic is envious of him or her and is trying to be hurtful rather than informative. Or perhaps the person decides that the critic really likes him or her but fears that the feeling is not reciprocated and so is engaging in a ‘reject before you’re rejected’ strategy to protect his or her own self-esteem. The mind is remarkable at processing information in service of maintaining a point of view or belief concerning the self. Protection of one’s self-concept, even if it means censoring some
320
Self: The Unity of Self, Self-Consistency
information and engaging in outright distortion of other information, goes right to the heart of selfhood. When in self-protection mode, the self can be looked upon as a totalitarian ego, a set of processes whose primary aim is protecting the head of (mental) state. The variety of means available for self-protection has been likened to a self-defense zoo. These means appear to be interchangeable. If discounting the credibility of the source does not work, for example, a person might justify the behavior that was criticized or remind him or herself how unrepresentative the behavior was. An especially well-researched strategy for protecting one’s self-concept is known as the selfserving bias. This simply means that people accept credit for their successes, but blame their failures on external factors or bad luck. Students who perform well on exam, for example, tend to see the grade as a reflection of their intellect, their knowledge of the material, or their preparation. When they do poorly, they see things far differently: the exam was unfair and not a valid measure of their competence. Intuition suggests that the most vigorous defenses of the self would occur when dealing with potential enemies or people with whom people have little in common. If a person is criticized by a stranger at a party, for example, he or she would probably mount a particularly strong defense. Although this is undoubtedly true, there are circumstances in which the greatest threat to selfesteem comes from those who are closest to the person and with whom he or she has a great deal in common. Insight into this state of affairs is provided by Tesser’s model of self-evaluation maintenance (SEM). The SEM model is primarily concerned with social comparison information. Two factors are critical in determining how we respond to such information: the ‘closeness of the comparison person’ and the ‘personal relevance of the behavior’ in question. Normally, people hope that someone who is close to them will excel at the activities in which he or she is engaged. People are proud and delighted when their friends or siblings win a tennis tournament or earn straight A’s in college. This is clearly the case when this person is trying to achieve something that is not directly relevant to one’s own personal ambitions. The problem for
the self arises when the other person excels at something that one also aspires to be good at. Imagine, for example, a person who is a budding tennis star or trying to stand out in college. A close friend or sibling who wins a tennis championship or earns a perfect GPA can make the person uncomfortable because the success represents a direct comparison with your success in that area. According to SEM reasoning, the person might go out of his or her way to find other ways to maintain a positive self-evaluation, or perhaps change his or her pursuits in order to avoid the inevitable comparisons. This scenario can be observed among friends, whose success in a personally relevant area can prove to be highly threatening. Marital partners, too, can experience this sort of difficulty if they have identical career goals. It is unlikely that their respective careers will proceed in lockstep with one another. The first one to receive recognition in the career can pose a serious selfconcept threat to the partner whose time has yet to come. From the perspective of SEM, there are three choices available when someone close to a person experiences success in an area that is personally relevant to him or her. He or she can rethink the importance of that area, distance him or herself from the other person, or sabotage the other person’s performance. The partners to a marriage, for example, can decide to follow different career paths. Alternatively, they could reduce their closeness, which is hardly a desirable solution for a married couple. Finally, the threatened partner could try to hinder the loved one’s success in various ways. There is evidence that these options are exercised. However, the mind’s ability to process information in service of self-esteem protection can render these strategies unnecessary. The selfdefense zoo, alluded to above, enables people to maintain healthy relations with one another when social comparisons prove to be uncomfortable. Brothers and sisters, husbands and wives, and close friends do not have to denigrate one another, sabotage their success, or change their careers when one of them outperforms the other. Instead, they can come up with a clever rationalization or perhaps invoke a self-serving bias to make sense of their successes and failures. If their partners
Self: The Unity of Self, Self-Consistency
are truly close, they will assist in this process and help draft an interpretation that keeps both parties happy. Self-Verification Defending the self is often synonymous with protecting a positive view of one’s self. Negative social feedback and unfavorable social comparisons engage a host of self-defense strategies, which makes sense if one assumes that people are motivated to develop and maintain high self-esteem. Most people do have relatively high self-esteem and feel that they are better than average on almost every dimension. But not everyone has a flattering view of him or herself in every area, nor does everyone have a global view of the self that is highly positive. When such people protect their self-concept, what are they protecting? A positive self-view or the view they actually have of themselves? Imagine a young man with serious doubts about his social charm being told by a young female that he is the most exciting person she has met in years. Would the male feel comfortable with this feedback? If positive self-evaluation were the issue, he might enthusiastically embrace the feedback. If, however, the real issue centered on maintaining a coherent self-concept, he might discount or even reject the feedback as inconsistent with the way he sees himself. The issue can be framed in terms of positive evaluation versus coherence. Is it more important to feel good about oneself or to feel coherent in one’s self-knowledge? Considerable insight into this issue has been generated under the guise of self-verification theory, a model of self-concept developed and researched by William Swann and his colleagues. Swann’s basic answer is that coherence trumps evaluation – but not without a struggle. There is evidence that people choose, like, and retain partners (e.g., roommates and friends) who perceive and evaluate them in a way that matches their own self-view. People prefer romantic partners who see them as they see themselves, even if the self-view at stake has some glaring weaknesses. Having a coherent view of the self, even if it is not all that positive, enables one to predict how one will fare in social settings and thus allows one to decide what to do, what contexts to seek out or avoid, and which people to trust.
321
The tendency to react based on feelings as well as on thinking complicates the picture slightly. Positive feedback obviously feels better than negative feedback, and because feelings occur more quickly and automatically than thoughts, the first reaction of a low self-esteem person to compliments and flattery is acceptance. And if the person is distracted and unable to marshal his or her cognitive resources to mount a defense against the flattery, he or she may continue to prefer hearing things that are positive but inconsistent with his or her self-view. Over time, though, people’s thoughts tend to overtake their feelings. For the low self-esteem person, this means that the bubble eventually bursts and he or she will engage in a variety of actions to reestablish coherence in his or her self-concept – even if this means sabotaging a relationship with someone who considers the person wonderful. Self-Concept Change If people’s efforts were directed solely at maintaining an existing sense of self, they would never display growth or have the potential for benefiting from experience. For that matter, people would never develop a coherent and meaningful selfconcept in the first place if their only concern was protecting what they already knew and believed about themselves in childhood. Clearly, the strategies devoted to self-defense must be overcome in some fashion by forces that prompt a reconsideration of what one is like. Changing a self-concept is not easy. A person can learn from his or her experiences and update his or her self-assessments, but such changes are largely confined to fairly peripheral as opposed to central aspects of the self. If a person thinks of him or herself in terms of intellectual ability but is less invested in his or her athletic prowess, for example, social comparison information is unlikely to have much impact on the former quality but can promote substantial change in the latter. Thus, the person might rationalize receiving a bad grade in a course (e.g., ‘the tests were unfair’) but view a poor showing on the racquetball court as credible information regarding his or her ability. With this in mind, one might think a substantial and hardto-rationalize experience is necessary to induce
322
Self: The Unity of Self, Self-Consistency
change in a central aspect of a person’s selfconcept. To attack someone’s intellectual ability, for example, one must present a strong argument that cannot be easily refuted. Such an approach might work if done with sufficient force and without allowing the person much time to launch a counteroffensive – but then again it might only promote strong resistance and perhaps aggression. A far different strategy follows from the process of self-organization. As discussed earlier, selfunderstanding is characterized by progressively higher levels of integration. This suggests that when a person thinks about the self in fairly specific or detailed terms, he or she is open to new information that could tie together the low-level information in a meaningful fashion. This idea has interesting implications for self-concept change. Rather than focusing attention directly on a person’s integrated self-view, successful agents of influence (e.g., clinical psychologists, friends, or con artists) are likely to experience greater success by inducing the person to think about him or herself in detailed, fragmented terms. In effect, the strategy is to disassemble the person’s integrated structure and then provide clues about how to put the pieces together again. In the routine and familiar contexts of everyday life, self-concept change is rare because people are likely to have an integrated sense of what their behavior represents. Such high-level understanding provides an effective ‘shield’ against new ways of thinking about the self. For self-concept to change, a crucial precondition for emergence must occur – the disassembly of integrated action understanding into more specific ways of thinking about the behavior. Vallacher and Wegner’s action identification theory offers a plausible scenario for the disassembly and reconfiguration process in self-concept change. The theory holds that any action can be identified at different levels, from concrete, mechanistic representations (lower-level identities) to comprehensive representations reflecting goals, values, consequences, and self-evaluative implications (higher-level identities). The act of taking a test, for example, can be identified in low-level terms as ‘answering questions’ or in higher-level terms as ‘showing knowledge,’ ‘earning a grade,’ or ‘demonstrating conscientiousness as a student.’ There is a preference for higher-level identification,
but there are well-documented factors that engender lower-levels of identification. Simply asking a person to recount the details of what he or she has done, for example, can induce a lower level of identification than would otherwise be the case. Interruption of ongoing action, a fairly common feature of everyday life, can also make a person sensitive to the lower-level aspects of his or her behavior. Because of the preference for higher-level understanding, the disassembly of action into its lower-level identities creates the potential for emergence and thus renders the person susceptible to new ways of thinking about the action’s broader meaning. A student induced to identify taking a test as ‘answering questions,’ for example, might be open to a new understanding of what test-taking means. To the extent that an emergent higher-level identity has implications for self-understanding (e.g., traits, goals, and values), the person’s selfconcept may undergo a transformation in one or more relevant domains. The test-taker’s selfconcept as a student, for example, may change from ‘seeking knowledge’ to ‘earning a degree,’ perhaps with corresponding changes in related aspects of his or her self-concept (e.g., from ‘intrinsically motivated’ to ‘motivated by outcomes’).
Coherence and Adaptation Self-concept coherence is considered essential for effective and autonomous functioning. The failure to maintain a coherent self-concept is associated with identity diffusion, promotes uncertainty and ambivalence in social relations, and undermines commitment to long-term goals and persistence in effortful task performance. A stable self-concept is also fundamental for the self-regulation of thought, mood, and action. Because self-concept is defined to a large degree in terms of values, it provides a frame of reference for evaluating courses of action and thus enables the person to control impulses, resist temptation, and behave independently of peer pressure. But effective functioning also requires learning and adaptation to changing circumstances. An inability to update one’s self-understanding can prevent such learning and adaptation. Coherence is an important feature of self-concept, then, but this must be balanced by
Self: The Unity of Self, Self-Consistency
openness to experience that promotes refinement and change in one’s representation of the self. The balance between coherence and flexibility can be difficult to achieve and maintain. An imbalance in either direction – excessive flexibility or excessive coherence – is at the root of important dimensions of individual variation in self-concept. Nonoptimal Flexibility Self-regulation would be impossible without a relatively unified self-concept in the domain for which self-control or decision-making is at issue. Research has established a connection between self-concept uncertainty – a hallmark of low coherence – and a tendency to identify one’s action at a relatively low level of identification. For example, a person with a highly certain sense of self might identify his or her behavior in social interaction as ‘being friendly’ or ‘demonstrating sensitivity,’ whereas someone with a less certain self-concept might identify his or her behavior in the same context as ‘speaking rapidly,’ ‘smiling,’ and ‘listening carefully.’ The higher-level identities have direct relevance to self-concept and provide a clear frame of reference for how to behave in other social contexts. The lower-level identities, in contrast, are elastic in meaning, consistent with widely divergent self-views, and thus provide an equivocal basis for how to behave in other contexts. Speaking rapidly and smiling could indicate friendliness or sensitivity, but they could also support a self-view of ingratiation, submissiveness, or anxiety. In line with the emergence scenario of action identification theory, low-level identification renders the person prone to impulsive action as opposed to maintenance of action plans, and open to social influence in deciding what to do. By the same token, an incoherent self-concept is vulnerable to momentary social feedback, social comparison, and performance setbacks. Everyday life provides a continuous flow of such information and much of this feedback is mutually contradictory. Without a unified sense of one’s enduring characteristics, the erratic nature of incoming self-relevant information would promote volatility in thoughts, feelings, and action tendencies. Research has documented that people who lack self-concept certainty and clarity tend to have unstable self-esteem, react
323
strongly to the self-appraisals provided by other people, experience mood fluctuations, and express highly variable self-evaluation when asked to reflect on themselves in a stream-of-thought manner. Recent clinical research has also identified a subset of people with bipolar depressive disorder who lack stable standards for self-evaluation. These people are at higher risk for hospitalization and suicide ideation than are bipolar depressives who oscillate between well-defined criteria (positive and negative) for self-evaluation. The lack of stable frames of reference undermines effective self-regulation and promotes volatility in mood, cognition, and behavior. Nonoptimal Consistency Several lines of theory and research suggest that a preoccupation with maintaining an internally consistent self-view can undermine adaptive personal and interpersonal functioning. Two scenarios in particular have been identified and investigated. First, a person may encounter self-relevant information that is inconsistent with his or her prevailing self-view. Rather than accommodate the self-view to such information, the person rejects this information and the information source. This reaction is typically manifest as cognitive processes that reinterpret, rationalize, or discount the meaning and relevance of the information. The processes of self-evaluation maintenance and self-verification, described above, represent this response to inconsistent information. These processes are essential to the maintenance of a coherent self-concept, but become problematic and a sign of rigidity and defensiveness when they prevent any new information from being considered on its merits. The rejection of inconsistent self-relevant information can also be manifest behaviorally. Research on self-verification has shown that people will sometimes react to inconsistent social feedback by exaggerating the self-perceived quality in question in an attempt to convince the feedback source that his or her perception is inaccurate. This can result in a counterintuitive situation, in which a person with a negative self-view tries to change the opinion of an observer or a relationship partner who views the person positively. Research on
324
Self: The Unity of Self, Self-Consistency
threatened egotism, meanwhile, shows that the behavioral rejection of inconsistent information can be manifest as antisocial action. When a person is concerned with protecting an insecure self-view, a person who challenges that self-view is considered a threat and may be subject to physical aggression. This tendency is especially likely when the self-view in question is unrealistically positive, suggesting that the combination of high (inflated) self-esteem and low self-concept certainty is a risk factor for violence in interpersonal relations. The second scenario of nonoptimal coherence concerns achievement motivation. People generally seek out performance settings that match their selfperceived capacities or that provide an opportunity to refine or validate these self-views. In some situations, however, the person may have relatively low certainty regarding the self-perceived capacity in question. In such cases, the person is inclined to avoid opportunities that call that capacity into question. If the performance setting cannot be avoided, the person might sabotage his or her performance by creating obstacles that render success unlikely. Prior to a test of intellectual performance, for example, a person might turn down the chance to engage in practice sessions that could facilitate his or her performance. Sabotage might also take the form of using performing-inhibiting drugs. The rationale for such self-handicapping is that the person is more concerned with maintaining a self-view than with engaging in actions that might result in failure and thus call into question the validity of the selfview. The likelihood and manifestation of selfhandicapping shows individual variation and, like individual variation in the tendency toward threatened egotism, is associated with low self-concept certainty.
Summary and Conclusions The development and maintenance of a selfconcept is exceedingly rare in the animal kingdom. A few other species (the great apes and dolphins) reach the cognitive threshold required for selfawareness, but humans alone have the capacity to develop a mental representation that consists of personal characteristics ranging from physical features to personality traits, beliefs, and values.
This human tendency is generally considered to have important evolutionary advantages, but the expression of these advantages requires a selfconcept that is sufficiently unified and coherent to provide a platform for regulating other psychological processes. Attention is focused on one’s own characteristics and processes when one is faced with real or perceived problems centering on self-control and decision-making. The self is rarely activated when other psychological processes are effective and running smoothly in accordance with environmental demands, personal and interpersonal expectations, and social rules. The restricted range of relevance suggests that the basic function of the self is to identify and repair processes that are not performing at a level that is personally adaptive. It is noteworthy in this regard that self-focused attention is associated with performance impairment, critical evaluation by other people, and a set of unique ‘selfconscious emotions’ that typically have negative rather positive valence. The self-regulatory function of explicit selffocused attention has been explored and documented within several research paradigms over the past two decades. A common thread to these accounts is the assumption that people compare their past, present, or possible action to a personal standard or characteristic when evaluating the action. If this comparison process suggests that the action is inconsistent with the regulatory standard, a variety of cognitive, affective, and behavioral mechanisms are engaged to eliminate or reduce the inconsistency. Self-regulation would be impossible without a relatively coherent self-concept in the domain for which self-control or decision making is at issue. To control one’s impulses, resist temptations, delay immediate gratification in favor of a distant reward, take on challenges and risks, and behave in an altruistic manner require a stable and unambiguous frame of reference for deciding the personally appropriate course of action. An internally inconsistent view of the self in the relevant domain would lead to equivocation, ambivalence, indecision, and cognitive distortion in service of the line of least resistance. The self-concept must also retain sufficient flexibility to accommodate the inherent dynamism
Self: The Unity of Self, Self-Consistency
and complexity of everyday life. The same processes that promote the formation of an integrated self-concept are at work when the self-concept undergoes transformation in response to new demands and challenges. In effect, transformation involves the reverse engineering of a self-structure, such that coherent regions are disassembled into their component elements. Because of the press for higher-order integration, the elements comprising the disassembled region are susceptible to influences that suggest a new configuration of self-understanding. With repeated iterations of the assembly and disassembly scenario, the selfstructure represents a dynamic equilibrium that provides both unity and flexibility in people’s self-understanding. See also: Meta-Awareness; Philosophical Accounts of Self-Awareness and Introspection; Self: Body Awareness and Self-Awareness; Self: Personal Identity.
Suggested Readings Baumeister RF (ed.) (1999) The Self in Social Psychology: Key Readings in Social Psychology. Philadelphia, PA: Psychology Press/Taylor & Francis. Baumeister RF, Smart L, and Boden JM (1996) Relation of threatened egotism to violence and aggression: The dark side of high self-esteem. Psychological Review 103: 5–33.
325
Baumeister RF and Twenge JM (2003) The social self. In: Weiner I (series ed.) & Millon T and Lerner MJ (vol. eds.) Handbook of Psychology, Vol. 5: Personality and Social Psychology, pp. 327–352. New York: Wiley. Campbell JD, Trapnell PD, Heine SJ, Katz IM, Lavallee LF, and Lehman DR (1996) Self-concept clarity: Measurement, personality correlates, and cultural boundaries. Journal of Personality & Social Psychology 70: 141–156. Carver CS and Scheier MF (1999) Themes and issues in the self-regulation of behavior. In: Klyer RS Jr. (ed.) Advances in Social Cognition, vol. 12, pp. 1–105. Mahwah, NJ: Erlbaum. Leary MR and Tagney J (eds.) (2003) Handbook of Self and Identity. New York: Guilford Publications. Nowak A, Vallacher RR, Tesser A, and Borkowski W (2000) Society of self: The emergence of collective properties in self-structure. Psychological Review 107: 39–61. Swann WB Jr. (1996) Self Traps: The Elusive Quest for Higher Self-Esteem. New York: Freeman. Tangney JP and Fischer KW (eds.) (1995) The SelfConscious Emotions: The Psychology of Shame, Guilt, Embarrassment, and Pride. New York: Guilford. Tesser A (1988) Toward a self-evaluation maintenance model of social behavior. In: Berkowitz L (ed.) Advances in Experimental Social Psychology, vol. 21, pp. 181–227. San Diego: Academic Press. Tesser A, Felson RB, and Suls JM (eds.) (2000) Psychological Perspectives on Self and Identity. Washington, DC: American Psychological Association. Vallacher RR, Nowak A, Froehlich M, and Rockloff M (2002) The dynamics of self-evaluation. Personality and Social Psychology Review 6: 370–379. Vallacher RR and Wegner DM (1987) What do people think they’re doing? Action identification and human behavior. Psychological Review 94: 1–15.
Biographical Sketch
Robin Vallacher is a professor of psychology at Florida Atlantic University, and research affiliate at the Center for Complex Systems, Warsaw University. He has been a visiting scholar at the University of Bern, Switzerland, and at the Max Planck Institute for Psychological Research in Munich. Dr. Vallacher has published research on a wide variety of topics in social psychology, from individual-level processes such as self-concept, self-regulation, and social judgment, to interpersonal and collective phenomena
Sensory and Immediate Memory N Cowan, University of Missouri, Columbia, MO, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Auditory stream segregation – The tendency for two types of sound that are presented in an intermixed fashion to be perceived as coming from separate sources, or perceptual streams; this is more likely to occur at faster presentation rates. Continuous moment hypothesis – The hypothesis that events are perceived together when they fall into the same sliding window of time; this window is continually updated. Discrete moment hypothesis – The hypothesis that events are perceived together in nonoverlapping bins of time; any events falling into the same bin are perceived as simultaneous. Immediate memory – Temporary information about a stimulus set, such as a word list or object array, that can be retained long enough to allow correct performance on a test shortly afterward, with no intervening interference or distraction. As used here, it includes both sensory memory and temporary memory for concepts. Perceptual memory – A term that can be considered a synonym for sensory memory. See the introduction for possible differences. Perceptual moment – Also sometimes called the psychological moment; the small amount of time in which two successive events seem simultaneous, which happens only when they are within about 200 ms of one another. Psychological present – A period of up to a second or so, during which a single event seems to take place. Sensory memory – Temporarily remembering how certain things look, sound, feel, taste, or smell.
Short-term memory – A term that can be considered a synonym for immediate memory. See the introduction for possible differences. Working memory – A term that can be considered a synonym for immediate memory.
Introduction to Concepts and Terms Sensory memory is temporarily remembering how certain things look, sound, feel, taste, or smell. Immediate memory refers to information about a stimulus set, such as a word list or object array, that can be retained temporarily to allow correct performance on a test shortly afterward, with no intervening interference or distraction. Researchers differ in their definitions but, as used here, immediate memory includes both sensory memory and also the short-lived memory of concepts. The concepts can come either from stimuli or from thoughts. What is distinct about immediate memory is that it is short-lived; it has a certain richness that fades with time or is eliminated by subsequent stimuli that cause distraction or interference. What is left when it is gone is a more impoverished, yet still critically important, long-term memory representation of sensations, events, and concepts from the past. Researchers also vary in the particular terms they use. For some, a synonym for sensory memory is perceptual memory. Other researchers would think of perceptual memory to include identification of the objects that are being perceived (e.g., ‘‘I am seeing a chair’’ or ‘‘I am hearing a bird’’), but would exclude that information from sensory memory. For some, synonyms for immediate memory are short-term memory or working
327
328
Sensory and Immediate Memory
memory. Other researchers would think of shortterm memory as referring only to the information automatically held, whereas working memory or immediate memory would include that information but also information held with the benefit of attention and mnemonic strategies, such as covert verbal rehearsal. We will circumvent these disagreements and stick to the terms sensory and immediate memory. At any moment, you are likely to be aware of some of the information contributing to sensory and immediate memory, but not other information. Imagine that you are at a large gathering of friends in a large room, with refreshments. The vast field of chatter in the room impinges on your auditory system, but you cannot make out all of the conversations and, mostly, your awareness of this chatter recedes into the background. You are holding a conversation with one person, on financing a new automobile, and several other people nearby are talking about a topic that interests you, a new political sex scandal. It is impossible for you to attend fully to both of these things at once. As you start picking up more about the sex scandal, your understanding of what you are being told about automobile financing starts to wane, as you struggle to stay polite. You notice your conversational partner making a distressed facial expression as he sees your increasing boredom with the topic and, to allay his distress, you say something to demonstrate that you are in fact paying attention. You do not remember much of the content of the recent discussion, but you still can recall the last few words that he said, which you repeat in an interested, questioning tone of voice. The general chatter in the room, the snippets from the sex conversation, and the monologue about car finance are all part of your sensory, perceptual, and immediate memory.
Attention and Immediate Memory A very important distinction is between information to which you pay attention and information that you ignore. The ignored information does not disappear immediately. Think of the last few words of the car financing conversation that you could repeat although you were not fully listening.
Similarly, while you are engrossed in a novel, someone may ask you what time it is. You ask them to repeat what they said but, before they do repeat it, you are able to recall what was asked. How did you do that? You pulled it from sensory memory into a more conceptual immediate memory. Donald Broadbent studied this kind of issue in the 1940s and 1950s, helping to establish a new field called cognitive psychology, the study of perception, memory, and thought processes. He and his colleagues in England were interested in some practical issues, such as how a pilot could land a plane safely when the radio instructions arriving through headphones are mixed with instructions being delivered simultaneously to other pilots landing other planes. That problem was eventually solved through advances in radio technology. In the psychology laboratory, however, the problem was studied using a technique called dichotic listening. One message was played to one ear and a different message in a different voice was played to the other ear concurrently. The task was to shadow, or repeat, everything presented to one ear. From time to time the tape was turned off and the subject was asked to repeat any information presented in the other ear, the ignored message. Subjects were stumped. They often could remember a few words from the end of the ignored message, but that was all. This led to a distinction between attended and unattended aspects of immediate memory. So, for example, you take in stimuli from all sensory modalities and some of the sensations outlast their stimuli. You can practically hear the last notes of a symphony for some seconds after they have been played. Following a flash of lightning on a dark night, your memory of the sight of twisting trees lingers on. Usually, though, there is such a wealth of stimulation that you cannot attend to all of it. A few sensations are selected for further processing, memorization, and understanding. They might be the actual words and ideas from a conversation, or a single canvas in an art museum. Broadbent’s theory was essentially that stimuli that are not relevant to one’s current concerns are filtered out of attention and awareness. It is relatively easy to filter out nondesired stimuli on the basis of physical cues, as when one ignores a certain person’s voice or picks only red berries, not green ones. It is more effortful, but still possible,
Sensory and Immediate Memory
to filter out nondesired stimuli on the basis of their meanings, as in a room of chatter when one ignores all conversations that are not related to particular subject, or when one goes into the garage to pick up several tools that are needed for a particular carpentry job while leaving others behind. In the latter type of selection, any kind of mental preoccupation can easily result in a failure of selection. You may then tend to hear a loud but irrelevant conversation, or you may pick up the wrong tools. At one point early in the history of cognitive psychology, it was proposed that people actually take all stimuli into the mind for further processing, and are limited only in the ability to respond to so much at once. That late-filter theory was based partly on a further dichotic listening study by Neville Moray in 1959, in which the subject’s own name was presented in the ignored message. It was found that some people notice their name, suggesting that the message was not actually filtered out. This finding was often highlighted in textbooks for many years, albeit without further support. (Instead, the field seemed to focus on visual stimuli.) Much later studies in the 1990s did confirm that this registration of one’s own name sometimes occurs. A study by Andrew Conway and his colleagues in 2001 discussed Moray’s result and suggested that this type of finding, noticing one’s name in an ignored message, does not actually support the late-filter theory. Another explanation is that Broadbent was essentially correct in suggesting an early attentional filter, but that some people fail to maintain a strong task set so that the information actually attended does not reliably match the information that was supposed to be attended. Conway and colleagues tested subjects on a complex immediate-memory test called operation span, in which arithmetic problems were to be solved along with memorization of a word presented after each problem. All of the words were to be recalled after the last arithmetic problem was solved. The operation span test, like certain other complex working memory span tests, correlates with intelligence rather well, and it correlates with the ability to pay attention. For example, recent research by Michael Kane shows that individuals with low spans experience more involuntary mind-wandering when they are trying to pay
329
attention than do high-span individuals. If dichotic listening really does involve selective filtering and subjects’ attention sometimes wanders, it could wander off of the message that was supposed to be attended and onto the message that was supposed to be ignored, or it could be split between the two messages. That should happen more frequently among low-span individuals. Consistent with this interpretation, Conway and colleagues found that 65% of the subjects in the lowest quartile of operation span noticed their names, whereas only 20% of the subjects in the highest quartile noticed. So Moray’s result does not strongly challenge Broadbent’s filter theory after all. The results for noticing one’s name are consistent with an early filter theory. Johnston and Heinz, in 1978, used a selective listening task to show that people can pay attention either to sensory features, as in the early filter theory, or to conceptual features, as in the latefilter theory. However, it takes much more effort to pay attention according to the late-filter theory. In this experiment, there was a secondary task, reaction time when a light was presented. Reaction time to the light was about the same when the task was carried out alone or along with selective listening based on a sensory feature, voice quality. During selective listening to one of two messages about different topics spoken in the same voice, however, reaction time to the light was much slower. This indicated that it takes effort to pay attention to conceptual features, whereas paying attention to physical features is easy and natural as in the early filter theory.
A Schematic Model of the Information Processing System Using the information that has been presented so far, you can begin to form a concept of how the brain’s information processing system acts. One way to conceive of the system is shown in Figure 1, from Nelson Cowan’s writings. Incoming stimuli all make contact with elements in long-term memory that become activated, and these activated elements of memory include sensory memory. However, not all activated elements enter the focus of one’s attention, just as the entire room full of
330
Sensory and Immediate Memory
Central executive
Long-term memory Activated memory S t i m u l i
Focus of attention
Figure 1 A concept of the immediate-memory system, which includes both sensory and conceptual features. Activated memory and the focus of attention are embedded processes that both contribute to immediate memory. For further support see Cowan’s 1995 and 2005 books.
chatter mentioned above, although it can be heard, cannot all be comprehended. The Central Executive represents mental processes that you use to impose your own will in terms of what information to focus your attention on and what to do with that attended information, as when you decide to concentrate on understanding one speaker while ignoring other conversations. However, other factors help control the focus of your attention, such as abrupt noises, sudden changes in the lighting, and attractive displays on television. These can either help your central executive, as when a speaker uses dramatic changes in voice intonation to get across a point, or can hinder your central executive, as when you try to listen to a monotonous speaker during a thunder storm. The information that is attended receives further processing in any case, resulting in a collection of concepts that accompany the sensory images in one’s conscious mind and mold one’s interpretation of these images. (Both sensations and concepts can count as activated elements from long-term memory.) The tendency for the central executive to be able to control which sensory stimulus channels to attend and which to ignore is roughly equivalent to Broadbent’s filter. In the illustration, paying attention on a sensory or perceptual basis (as in picking only the red berries) is like staying on one stimulus input arrow, whereas paying attention on a
conceptual basis (as in finding which tools one needs) is like hopping from one arrow to another. Not everything in immediate memory can be an activated set of representations from long-term memory, of course. New information is represented by a new set of links between alreadyexisting features of some sort. For example, if you saw a blue apple, it would probably be the first link of the blue feature and the shape associated with apples. You might also receive a new name for it. These new links between a shape, a color, and a combination of phonemes would coexist in the focus of attention and would produce a new long-term memory representation. One reason to use a schematic diagram like this one is that the brain is complicated. Any one of these components included in Figure 1 could correspond to a twisted tangle of tree-like neural structures distributed widely throughout the brain. The schematic model is at a higher level of analysis. By analogy, one can delineate the executive, legislative, and judicial branches of the US government when, in fact, the three branches of government are represented by complex and overlapping webs of people, buildings, and regulations. Nevertheless, there are known neural correlates of these elements of mind. The central executive processes rely a great deal on frontal lobe structures (behind the forehead). The parietal areas, more toward the upper middle of the head, are very much involved in the focus of attention; damage to those areas can result in patients who are unaware of one part of space (unilateral neglect) or are unaware that they have a serious paralysis or other disability (anosognosia). Recent brain research shows that parts of the parietal lobes are especially active when a challenging amount of information is being saved in immediate-memory tasks. The temporal lobe areas (behind the sides of the head) include the hippocampus and other neighboring areas that are especially important in retaining information and forming new long-term memories, and diverse areas represent different kinds of memories that become electrically active when the ideas that are represented are activated by stimulation or thought processes. The characteristics of sensory and immediate memory that make them particularly interesting stem partly from their limits. The discussion of
Sensory and Immediate Memory
what these kinds of memory can do is intricately related to what they cannot do. In the remainder of this article, two types of limits will be discussed. The first is time limits in activated memory and the second is capacity limits in the focus of attention.
Time Limits of Activated Memory It is possible to demonstrate directly that there are limits in the activation of information from longterm memory, and this temporarily activated memory forms the basis of the human information processing scheme shown in Figure 1. These are time-related limits, although the exact limit seems to depend on the nature of the stimulation. Some concepts related to human awareness may be based on this temporarily activated memory. These will be discussed first, and then more direct tests of the time limits of activation will be discussed. There are two concepts related to time limits and awareness that are of special interest, called the perceptual moment (also sometimes called the psychological moment) and the psychological present. The perceptual moment is the small amount of time in which two successive events seem simultaneous, which happens only when they are within about 200 ms of one another. As an example, if you move your finger back and forth as rapidly as you can, it will visually appear as if your finger is in two places at once (as well as being smeared in between these two places). Why does this occur? One way of explaining it is that the neural processes representing each visual viewpoint outlast the actual stimulus and are still ongoing when the next visual viewpoint arrives. This, in fact, is how movies are perceived. Another way of explaining it is to say that there is a visual sensory memory or afterimage of the stimulus that seems like a continuation of that stimulus after it is gone, and that the two afterimages are combined in our awareness into a single stimulus. If there were more time between the stimuli, one afterimage would fade before the next one was formed and the two viewpoints would not appear simultaneous. A great deal of research suggests that this neural explanation and this psychological explanation are fundamentally compatible and that both are apt. The
331
perceptual moment depends on the type of stimuli that are presented but typically ends within a couple hundred milliseconds. In contrast to the fleeting perceptual moment, the psychological present is a longer period of time of a second or so, during which a single event seems to take place. The rapid beats of a drum roll may be grouped together to form a single event, whereas slower beats of a drum may be perceived as separate events. Whereas, for the perceptual moment to be the same for two stimuli, ongoing perceptual processing must overlap, the psychological present has a looser criterion; the second stimulus need only be presented while the first is still vividly recalled. It is as if there are two phases of temporarily activated sensory features in memory: a vivid afterimage that seems as if the stimulus is containing, lasting for several hundred milliseconds, and then a vivid recollection of the sensory events, for a second or so. Like the latter, William James in the late 1800s used to think of primary memory (in this chapter, immediate memory) as the trailing edge of the conscious present. Many studies could be used to demonstrate these concepts. We will focus on a couple of them that seem especially thought-provoking. In 1968, D.A. Allport carried out a study to distinguish between two varieties of the perceptual moment hypothesis. In the discrete moment hypothesis, the stimuli are accumulated into one moment until some fixed amount of time; then subsequent stimuli are accumulated into another moment until a similar fixed amount of time; and so on, just as each sheet of paper in an office could be stamped with a date and multiple sheets that arrived at different times would share the same date stamp. In the continuous moment hypothesis, though, events would be viewed as if through a sliding window (i.e., sliding in time). First the window might integrate events 1, 2, and 3; then the oldest event 1 would fade from the window, to be replaced by a new event 4; and so on. To distinguish between these hypotheses, Allport developed a simple but ingenious test using an oscilloscopic display, as shown in Figure 2. (Actually, the display included 12 lines but the point can be made more simply using four lines as shown here.) The lines would be presented one
332
Sensory and Immediate Memory
1 Presented
1st,
and again
5th
2nd,
2 Presented currently perceived as shadow 3 Presented 3rd 4 Presented 4th etc. 1 2 3 4 1 2 3 4 1 2 3 4 Discrete moment: Shadow = 4, 3, 2, 1….
etc. 1 2 3 4 1
2 3 4 1 2 3 4
Continuous moment: Shadow = 4, 1, 2, 3… Figure 2 Allport’s experiment to distinguish between discrete and continuous varieties of the perceptual moment. The top part of the figure illustrates a particular moment of perception in the situation in which Lines 1–4 were presented one at a time, in order, repeatedly at a fast rate. Line 2 happens to be perceived as a shadow at the moment depicted here, but the perceived shadow location moves as the display progresses. The proposed mechanism is illustrated according to the discrete moment hypothesis in the middle of the figure (with brackets showing which lines four consecutive moments would include), and according to the continuous moment hypothesis at the bottom of the figure (with brackets showing which lines four overlapping, temporally sliding moments would include).
at a time in the repeating order 1, 2, 3, 4, 1, 2, 3, 4. . ., and so on with no break. The subjects were to adjust the rate of line presentation until what was perceived was all but one of the lines at the same time. (If the rate were any faster, all of the lines would be perceived at once or, if the rate were any slower, fewer lines would be perceived at once.) With the rate so adjusted, the line that was not perceived at any one moment was called a shadow and the change over time in which line was seen as a shadow was called shadow movement. Then different predictions about shadow movement could be made for the two types of perceptual moment. For the discrete moment, as shown in the middle of the figure, the assumption was that the succession of lines was parceled out into nonoverlapping moments of a certain duration. That process would result in shadow movement in the opposite direction from the actual movement of lines. If the first
discrete moment were long enough to include Lines 1, 2, and 3, then 4 would be the shadow. The next discrete perceptual moment would open up just in time to capture Line 4, and also the next presentation of Lines 1 and 2, but then it would close before it could include the next presentation of Line 3, which would go in the next window instead. Thus, the shadow has moved from Line 4 to Line 3 and continues in that manner, in a direction opposite from that of the line presentations. In contrast, in the continuous moment hypothesis, shown in the bottom of the figure, the shadow movement would go in the same direction in which the lines were presented because the temporally sliding window always includes the most recently presented three lines and leaves out the least recently presented line. The evidence from the experiment unequivocally supported the continuous moment hypothesis, in which our events are grouped together according to a sliding window of perception. This sliding window can be viewed as sensory memory, in which the least recent line always was the first to fade from view. Albert Bregman has carried out a line of research on a phenomenon that can be seen as demonstrating the power of the psychological present, called auditory stream segregation. Suppose you hear a simple series of two tones in alternation: high–low–high–low in pitch, and so on. How will the series sound? That depends on various factors, but one of these is the time between tones, as shown in Figure 3. When the series is presented slowly, the mind links together adjacent tones into a single, coherent perceptual stream going up and down. When the series is presented more quickly, it is the tones of the same pitch that seem to be grouped together, into a high-pitched stream and a second, low-pitched stream. In fact, it is difficult to hear exactly when the high and low pitches were presented relative to one another; there is no clear perceptual organization across streams. Tones in each stream share the grouping principle of similarity. When the series is considerably slowed down, however, what may be different is that successive tones of the same pitch are no longer within the same psychological present, and so they cannot be grouped together. In terms of memory, it may be that the vividness of one tone has faded by the time
Sensory and Immediate Memory
a similar tone is presented, so that they no longer seem to belong together in time. Time-dependent activated memory does not rest primarily on notions of the perceptual moment and the psychological present, but has repeatedly been examined more directly through a wide variety of experimental procedures. One of these, by George Sperling, was published in 1960. He carried out a number of experiments. A typical one is illustrated in Figure 4. The first panel of the figure shows an array of 12 letters that was presented very briefly, with different letters on every trial. In one kind of trial (not the kind shown in the figure), the task was simply to recall all 12 letters by writing them down. This was called the whole report procedure, and it showed that practiced adults could recall only about 4 of the 12 letters in their correct locations. Sperling realized, though, that he did not know the basis of this limit. It could be that more items than this were held in an activated form of memory but that they disappeared from memory before they could all be written down. What Sperling did to address this problem was to develop the
Stimulus
Perception One stream
High stream
333
partial report procedure, shown in Figure 4. Subjects learned that they should respond in a manner dictated by a tone presented after the array. If the tone was high, the task was to recall the top row; medium, the middle row; or low, the bottom row. Because the tone was not presented until after the array, when the array was presented the subject had to retain all three rows. However, after the tone was presented, the task became simply to recall one row, which did not exceed the number of items that subjects could recall in the whole report situation. The result would indicate how much information about the cued row was held in memory at the time that the cue arrived. Given that the other rows had to be held at the same time, the total amount of information available in memory when the tone cue arrived could be calculated by multiplying by 3 the average number of items recalled in partial report. The outcome is shown in Figure 5. It turned out that the result depended on the amount of time between the array and the tone cue. If the tone was presented very soon after the array, almost all items in the array were still available for recall. However, if the tone cue was delayed for a quarter second (250 ms), it was of almost no value. The number of items available for recall was then no greater than the whole-report limit of about four items. This suggested that all of the items in the array were held in memory at first, but that they faded from memory within about 250 ms. Other experiments showed that the cue had to indicate a physical feature, such as the row of the array to
Figure 3 The phenomenon of auditory stream
segregation, as in Bregman’s work.
?
x
k
I
m
c
q
r
t
b
k
w
s
Variableduration blank interval
?
High tone (= recall top row)
Figure 4 Sperling’s partial report procedure.
?
?
Items available (partial report # times 3)
Low stream 12
4
0
Whole-report limit for comparison
250 ms
0 Variable blank interval
Figure 5 Results of Sperling’s whole and partial report procedures.
334
Sensory and Immediate Memory
recall, and not a more abstract, conceptual feature, such as whether to recall letters or digits from an array that included both types of symbols. This suggests an explanation that is entirely consistent with Donald Broadbent’s explanation of the dichotic listening results. A large amount of sensory information is held for a brief time, but only a small subset of that sensory information can be shifted to the focus of attention before it is lost. It can be selected for attention and further processing on the basis of physical cues (such as voice quality or direction, or written letter color or location) more easily than it can be selected on the basis of conceptual cues (such as the topic of speech, or the category of the written symbols). One possible discrepancy between auditory and visual sensory memory in these studies is that the duration of auditory sensory memory appeared to be several seconds, whereas the visual sensory memory only appeared to last a fraction of a second. In a review that Cowan published in 1988 and discussed further in a 1995 book, an alternative interpretation of the literature on sensory memory was suggested. It appeared that, all information considered, there are two separate phases of sensory memory: a vivid mental afterimage experienced while perceptual processing of the items continued for about 250 ms, and a longer-lasting, vivid sensory recollection for several seconds. There has not been a great deal of discussion of this issue (i.e., whether there are sensory-modality-specific memory durations or, instead, two phases of sensory memory in all modalities), but it remains one on which researchers do not seem to agree. The two phases of sensory memory discussed in the 1995 book, if they do exist, would appear to underlie the perceptual moment and the psychological present, respectively, as we have discussed. These experiments leave open the question of whether nonsensory, conceptual information that is activated also has a limited time period of persistence. This is not an easy issue to examine, because sensory information ordinarily may have to be drawn into the focus of attention before conceptual features can be perceived. Then it becomes possible to rehearse the information by repeating it covertly (mentally and silently) or simply by continuing to attend to it. One can prevent this rehearsal by presenting distracting
tasks, but these can cause interference as well. In 1973, Michael Watkins and his colleagues presented lists of words followed by tones that either could be ignored or, in a different type of trial, had to be identified by button press, and found that memory was lost over time – especially during the first 3 s of a 20-s period – only when the tone information had to be identified. On the other hand, in 2004, Stephan Lewandowsky and his colleagues took a different approach and found that memory for letters was not lost over time when the words had to be recalled either slow or fast, regardless of whether subjects had to recite a word over and over to prevent covert verbal rehearsal. Some details of both studies, and of other studies in this literature, seem to be such that the issue of the loss of conceptual activated memory features over time remains unresolved. Even if a loss of activated memory over time is observed, it has to be asked whether it is the amount of time itself that is important, or the way in which that time is perceived relative to recent events. This is clear in a study of brain function based on electrical signals at the scalp, or event-related potentials, by Istva´n Winker and his colleagues in 2001. They presented series of tones that remained identical for six tones, with the seventh tone sometimes shorter than the others. Even though the subjects are busy reading during this kind of procedure, so that the tones are ignored, subjects’ brains still respond to tone changes if the unusual tone (the deviant) is presented shortly after the other tones (the standards). This mismatch negativity response is said to occur because the brain saves a sensory memory of the standard tones and compares each new tone to that standard representation. Figure 6 shows three different stimulus arrangements and indicates the results. When all tones in the series were a half second apart, all subjects’ brains responded to the change in tone duration. However, when the standard tones were one half second apart and the deviant did not occur until 7 s after the last standard, some subjects’ brains still responded, whereas others’ brains did not. For the subjects whose brains did not respond, one might think that a memory of the standards was lost by 7 s. However, another interpretation is that their brains no longer considered those standard tones to serve as a fair, current
Sensory and Immediate Memory 0.5 0.5 s s
0.5 s
(Beginning of series not shown)
335
Everybody’s brain responds 7.0 s
7.0 s
Only some brains respond 7.0 s
Everybody’s brain responds
Figure 6 The procedure that Winkler and colleagues used to examine why tone memory declines over time.
comparison. After all, the standards in each series formed a tight group in time and the deviant was not part of that group. Verifying that interpretation, all subjects’ brains responded when all tones were 7 s apart, as in the third row of the figure. To summarize this discussion of temporarily activated memory, it can include sensory or conceptual features from long-term memory that have been activated by recent events or thoughts. The conceptual features are much more likely to be activated only after information is drawn into the focus of attention, but then the focus of attention might shift elsewhere and leave behind activated conceptual features that are unattended. For sensory and conceptual activated features, it is clear that there is a time limit to activation but, after all these years of research, it is not yet clear what kind of time limit that is. All of the currently available information might be compatible with the notion that the critical time interval may be derived by the brain relative to the pace of events. At least, that may be the case for the longer form of activated memory that we have linked to the psychological present. It is usually within the range of a few seconds but it is hard to pin down exactly. Within that possibly flexible duration of the psychological present, there appears to be a shorter duration of about 250 ms that may be determined by the time it takes for perceptual processing of a single stimulus to be completed, which therefore may be a more precisely fixed duration. It corresponds to a vivid sensory afterimage in the brain and to the perceptual moment.
Capacity Limits of the Focus of Attention One of the functions of the focus of attention is to overcome time limits. Suppose you want to check the oven in 3 min. Without an external timer, you have no choice but to invest a considerable amount of effort attending to the oven so that your mind does not wander on to other things. The fundamental reason why this is effortful is that there are competing stimuli and competing thoughts, and not all of them can be attended at once. That is, there is a capacity limit to the focus of attention. One dramatic result of the capacity limit is what Dan Simons and others have called inattentional blindness. You can demonstrate it by playing a trick on a friend in the office. Wait until they turn around for a moment and then remove a small item from plain sight on their desk, such as a stapler. If you do this discretely chances are that, when your friend returns attention to your direction a moment later, he or she will not notice the missing object. Provided that the item was not recently attended, despite being well within the perceptual field of view, its disappearance will not attract notice. Simons and his colleagues demonstrated inattentional blindness dramatically by having one experimenter stop a pedestrian to ask directions. In a staged event, two other confederates of the experimenter holding a door in a horizontal position walked between the experimenter and the hapless pedestrian who was asked directions, obscuring their view of each other. Secretly, on the side of the
336
Sensory and Immediate Memory
door across from the pedestrian, the experimenter grabbed the door and was replaced by one of the confederates who had been holding the door. When the two people with the door moved along out of the way, the confederate acted as if he was the original experimenter. In many cases, the pedestrian did not notice that there had been a change in the person to whom he was giving directions before the interruption, and continued to give directions. He had been attending to his directions, and not primarily to the exact features of the person receiving those instructions. Sperling’s partial report procedure shown in Figure 4 could be considered an inattentional blindness procedure because attention is insufficient to encode into the focus of attention all of the items from the array held in visual sensory memory. In 1997, Steve Luck and Ed Vogel simplified that procedure so that it would not require that multiple items be recalled, in order to get the most sensitive measure of what items are in working memory. These might be thought of as the items encoded into the focus of attention. A version of that procedure used recently by Jeffrey Rouder and his colleagues does an excellent job of estimating items conceptually held in working memory. The procedure is illustrated in Figure 7. An array of small, colored squares is briefly presented. (Each pattern in the figure represents a different color.) An interval of about 1 s then elapses, giving the subject plenty of time to encode the array into attention. Then a masking display is presented with a set of multicolored squares in the same location as the colored squares in the previous array. This serves to eliminate any remaining sensory memory. What remains are any conceptual
Probe Post-perceptual mask Delay interval Array Figure 7 A version of the visual working memory procedure by Steven Luck and Edward Vogel.
aspects of working memory that the subject has retained even though the sensory memory features have been overwritten by the mask. The masking display also serves to remind the subject just where the squares were on the screen. This is followed by a single colored square, and the task is to indicate whether this square is the same color as the square in that location within the original array. If it is not, it is a color that was not found in the array. The probe remains until an answer is given. This procedure produced results that conformed very well to a simple mathematical model to calculate capacity. The model included an assumption that on a certain percentage of trials, the subject was not paying attention at some point. That happened on 12% of the trials according to the model. Each subject was said to have a certain capacity k and, when he or she paid attention, k items were retained in working memory. When the array item in the same location as the probe was in working memory, the subject was assumed to know whether there was a color change or not. If the item was not in working memory, either because of inattention or because capacity was exceeded, the subject was assumed to guess. In different trial blocks, different proportions of trials involved color changes to manipulate the guessing rate. The estimate of k stayed fixed across guessing rates and across different numbers of items in the array, as it should according to the mathematical model. The mean k was 3.35 items, consistent with previous estimates of working memory capacity such as those reviewed in Cowan’s 2005 book. Scott Saults and Cowan recently showed that this kind of working memory examined by Rouder and his colleagues is not limited to visual information. They set up four loudspeakers surrounding the subject and played four different spoken digits at the same time, each in a different voice, while also presenting an array of four or eight colored squares on the computer screen. The subject was responsible sometimes for the visual information only, sometimes for the auditory information only, and sometimes for both modalities at once. Provided that a postperceptual mask was presented to eliminate sensory memory in both modalities, the visual and auditory information traded off. In particular, subjects remembered about 3.5 visual items when they did not have to retain auditory items also, and
Sensory and Immediate Memory
they remembered about 3.5 items total when they had to remember the auditory along with the visual. The verbal spoken items displaced some of the nonverbal visual items in working memory, which seems to be evidence that the information being stored in a capacity-limited manner is conceptual and abstract rather than specific to a sensory modality. The memory in these procedures we have been discussing seems less than what was discussed in a famous article by George Miller, who, in 1956, suggested that there is a ‘‘magical number seven plus or minus two,’’ describing how many items people can recall. This estimate is roughly true, but it does not seem to reflect the most basic type of immediate memory. Instead, it seems to reflect a process whereby the items in working memory are grouped together to make larger items. (That grouping process was in fact one of Miller’s main points.) When one reads the telephone number 356–4129, one tries to form a group of three items followed by a group of four. When the groups are formed, each group may take up only one slot in working memory. It is these smaller groups that may reflect the basic capacity limit, with the larger limit occurring as several smaller groups are entered into working memory together. Anders Ericsson and his colleagues have made the point that, with enough practice, at least some individuals manage to raise their memory span from seven items to the astonishing level of eighty or more items, but only for the type of material being practiced (such as digits). They do it by learning to form large chunks based on knowledge they already have, such as memorized athletic records. More precisely, chunks of up to about four items are formed and then these chunks are combined to form even larger, higher-level chunks. The capacity of working memory increases in childhood and then decreases again in old age, as Cowan, Moshe Naveh-Benjamin, and their colleagues have demonstrated recently. Also, as Randall Engle and his colleagues have emphasized, tests of working memory capacity do a good job of discriminating between young adults with more versus less aptitude on complex cognitive tasks and intelligence tests. Yet, we do not understand what it is that distinguishes someone with a high span from someone with a lower span. One theory is that individuals with a high span simply have more slots in working
337
memory than do lower-span individuals. However, an alternative theory is that individuals with a high span are able to pay attention better, and therefore are more likely to fill their working memory capacity with relevant items as opposed to thoughts that are irrelevant to the task that has been assigned. Jim Gold and his colleagues found evidence for the slots theory in distinguishing schizophrenic individuals from normal controls. In one experiment, for example, they presented arrays with colored items of two shapes and usually tested subjects on one of these shapes. Occasionally, they would test subjects on the other shape. In this way they could tell how well individuals were filtering out the usually irrelevant shape (by examining the advantage in performance on the usually tested shape over the other shape) and they could tell how much individuals encoded into working memory (by examining the total of capacity k for the two shapes added together). Surprisingly, the schizophrenic individuals filtered just fine but were below normal on total capacity. In contrast, Vogel and his colleagues have done similar testing using event-related brain potentials and have argued that high- and low-span normal individuals differ primarily in the ability to filter out the irrelevant information. We know of areas in the brain that specialize in holding working memory information (certain areas in the parietal lobes) and areas in the brain that specialize in filtering out irrelevant information (the basal ganglia). It stands to reason, then, that the available slots in working memory and the ability to fill these slots appropriately and efficiently are two important, related, but nonidentical aspects of individual and group differences.
Conclusion Immediate memory, or its close cousin working memory, is critical for human thought and awareness of the environment. It is, as William James said, the trailing edge of the conscious present. The conceptualization of this memory as shown in Figure 1 has proved to be useful in understanding the mind. Activated features of memory, within which sensory memory usually predominates, are collected by our brains, apparently for all stimuli, without the need to pay attention. Even though the
338
Sensory and Immediate Memory
information that is activated remains so only briefly, it provides a rich backdrop from which information can be drawn selectively into the focus of attention for further processing in a timely manner. Conceptual features that come out of this selective processing can be added back into the pool of activated memory. There are some key questions that remain unanswered regarding the principles that produce the limits in activated memory and the focus of attention, but the scientific endeavor is briskly marching toward some tentative answers.
Acknowledgments This work was supported by NIH Grant R01 HD-21338. See also: Visual Experience and Immediate Memory.
Suggested Readings Allport DA (1968) Phenomenal simultaneity and the perceptual moment hypothesis. British Journal of Psychology 59: 395–406. Bregman AS (1990) Auditory Scene Analysis: The Perceptual Organization of Sound. Cambrdige, MA: MIT Press. Broadbent DE (1958) Perception and Communication. New York: Pergamon Press. Cowan N (2005) Working Memory Capacity. Hove, East Sussex, UK: Psychology Press.
Engle RW, Tuholski SW, Laughlin JE, and Conway ARA (1999) Working memory, short-term memory, and general fluid intelligence: A latent-variable approach. Journal of Experimental Psychology: General 128: 309–331. Ericsson KA, Delaney PF, Weaver G, and Mahadevan R (2004) Uncovering the structure of a memorist’s superior ‘‘basic’’ memory capacity. Cognitive Psychology 49: 191–237. Gold JM, Fuller RL, Robinson BM, McMahon RP, Braun EL, and Luck SJ (2006) Intact attentional control of working memory encoding in schizophrenia. Journal of Abnormal Psychology 115: 658–673. Johnston WA and Heinz SP (1978) Flexibility and capacity demands of attention. Journal of Experimental Psychology: General 107: 420–435. Kane MJ, Brown LH, McVay JC, Silvia PJ, Myin-Germeys I, and Kwapil TR (2007) For whom the mind wanders, and when: An experience-sampling study of working memory and executive control in daily life. Psychological Science 18: 614–621. Lewandowsky S, Duncan M, and Brown GDA (2004) Time does not cause forgetting in short-term serial recall. Psychonomic Bulletin & Review 11: 771–790. Luck SJ and Vogel EK (1997) The capacity of visual working memory for features and conjunctions. Nature 390: 279–281. McCollough AW and Vogel EK (2008) Your inner spam filter. Scientific American Mind, June/July: 32–35. Miller GA (1956) The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review 63: 81–97. Simons DJ and Levin DT (1998) Failure to detect changes to people during a real-world interaction. Psychonomic Bulletin & Review 5: 644–649. Winkler I, Schro¨ger E, and Cowan N (2001) The role of largescale memory organization in the mismatch negativity event-related brain potential. Journal of Cognitive Neuroscience 13: 59–71.
Biographical Sketch
Nelson Cowan (PhD, 1980, University of Wisconsin) is a Curators’ Professor at the University of Missouri. He has been interested in the brain and mental processes every since he was in high school, and this interest developed further in college at the University of Michigan. His research, funded by the National Institutes of Health since 1984, focuses on working memory, selective attention,
Sensory and Immediate Memory
339
and the childhood development of these processes. He is the author of Attention and Memory: An Integrated Framework (1995, Oxford University Press) and Working Memory Capacity (2005, Psychology Press). He has served as an associate editor of three journals, Journal of Experimental Psychology: Learning, Memory, and Cognition, Quarterly Journal of Experimental Psychology, and European Journal of Cognitive Psychology. As of September 2008, he is on the governing board of the Psychonomic Society and is the president of the experimental psychology division of the American Psychological Association. Nelson is married and has three grown children; his wife, Jean Ispa, is a professor of human development and family studies.
Sleep: Dreaming Data and Theories K Valli and A Revonsuo, University of Turku, Turku, Finland; University of Sko¨vde, Sko¨vde, Sweden ã 2009 Elsevier Inc. All rights reserved.
Glossary Bizarreness – Dream element that is highly improbable or impossible in waking reality. Content analysis – Method that can be used to convert qualitative data into quantitative so that data can be expressed in numbers, frequencies, proportions, and statistically analyzed. Random activation theories – A set of theories stating that dreaming is a random by-product of neurophysiological events taking place during sleep, and that dream content as such has no function. Recurrent dream – A dream that occurs repetitively and has every time highly similar content or theme. Reflective consciousness – A form of consciousness that allows us to think about, evaluate, and pass judgments about the contents of phenomenal consciousness. Threat simulation theory – An evolutionary psychological theory of dreaming that states that the original function of dreaming was to simulate the threatening events of waking life to rehearse threat recognition and avoidance strategies. Typical dream themes – Themes in dream content that are universal among humans.
Introduction Sleep is a universal biological necessity all humans share. But what about dreaming? One problem in dream research is that researchers do not even concur on the definition of dreaming. In the 1970s, Frederick Snyder suggested that dreaming refers to the subjective conscious experiences we have during sleep, and consists of complex and
organized mental images that show temporal progression or change. To Snyder, dreams were like stories or at least like scenes from a story. Static, simple, and unimodal dream imagery was later distinguished from full-blown dreaming, and referred to as sleep mentation. While these definitions do not state anything about the actual content of dreams, J. Allan Hobson’s definition that dreaming is mental activity during sleep with most of the following features present – hallucinations, delusions, narrative structure, hyperemotionality, and bizarreness – emphasizes the qualitative features of dream content. Antti Revonsuo recently suggested that instead of strict definitions and categorizations, we should consider dream phenomena as a continuum. At the one end of the continuum resides full-blown dreaming, defined as complex, organized, temporally progressing, multimodal contents of consciousness during sleep that amount to a simulation of the perceptual world, and at the other end there is sleep mentation, that is, simple contents of consciousness during sleep that show a low degree of complexity and organization. Another problem faced in dream research is that there is no objective physiological marker for dreaming. Thus, researchers have to rely on subjective reports when they study dreams. Majority of people can remember their dreams, although there is large interindividual variation in dream recall frequency. As Inge Stauch and Barbara Meier have shown, among others, some recall dreams every night, most of us frequently or every now and then, and for others, dream recall is a rare event. Dream recall seems to be highly independent of personality characteristics. Michael Schredl found that the most powerful predictor of recall is a positive attitude toward dreams. Regardless, according to several studies, there are people who state that they have never dreamed. When such people have been studied in a sleep laboratory, and have been woken up from rapid eye
341
342
Sleep: Dreaming Data and Theories
movement (REM) sleep – the most optimal stage of sleep for dreaming to occur – most can recall at least some dreams, but some do not. Also specific brain damage may obliterate dreaming altogether while REM sleep is preserved. Regardless, we do not know if these people dream or not, only that at least they do not recall their dreams. Consequently, dream recall, as well as the accuracy of recall, is a limiting factor in dream research.
Dreaming Data Universality of Dream Themes Numerous laboratory experiments indicate that most people experience dreams every night and several times per night, although not all of these dreams are normally recalled. Not only dreaming as a phenomenon seems to be universal but also the things people around the world dream about share many common themes, independent of the culture the people come from. Typical dream themes are dreams that contain the same general form over and over again across large numbers of people, even though the details of the dream may vary considerably. In questionnaire studies, people have been presented with a list of various dream themes, and asked: ‘‘Have you ever dreamed of. . . ?’’ The most common universal theme is that of being chased or pursued, reported by approximately 80% of dreamers. Being physically attacked or being frozen with fright are also common themes, as is falling from high places, or being on the verge of falling. Being trapped and not able to get out, being lost, or drowning, are also universally dreamed about. Sexual experiences or being naked in public is another well-remembered dream content, and many of us recall dreams about being late, or of school, teachers, or studying. Dreams in which we fly or soar through the air are also remembered by many of us, and in contrast to the previously mentioned themes, these kinds of dreams are usually positive. Typical Dream Content Questionnaire studies offer insights into the most common dream themes shared in different cultures, but a more detailed method for studying
dream content is called content analysis. In content analysis, detailed dream reports are obtained immediately after awakening. Then, the elements present in dream reports are separated and classified, and specific proportions, frequencies, and percentages of different elements calculated. The pioneers of dream content analysis, Calvin S. Hall and Robert Van de Castle, developed in the 1950s and 1960s a detailed content analysis scale still widely used today, that is, the Hall and Van de Castle system. Their content analysis scale was groundbreaking in empirical dream studies, and since then, many different content analysis scales have been developed for various purposes. Sensory modalities in dreams According to studies conducted by, for example, Frederick Snyder, and Inge Strauch and Barbara Meier, all sensory modalities are present in our dreams. All dreams include visual experience, and auditory components are also frequent. Bodily sensations, such as kinaesthesia and touch, are less frequently reported, and smell, taste, and pain experiences are rare but possible. The specific sensory experiences are present in our dreams approximately to the same extent as in our waking reality. The visual qualities of dream experiences were explored in an ingenious experiment by Allan Rechtschaffen and Cheryl Buchignani in the early 1990s. They had a selection of more than 100 photographs in which the visual features (such as chromaticity, saturation, illumination) had been altered. Immediately after awakening the subjects in a laboratory, the participants selected a photograph that most accurately matched the visual quality of their dream. The most often picked photograph was the one that had not been altered in any way, and the next ones chosen presented only slight variations from normal. Thus, the visual qualities of dreams closely resemble the way we perceive the world while we are awake. Most, but not all, dreams are perceived in color. Only approximately 20% of dreams lack color, and are perceived achromatically, in black and white. Seeing in black and white every now and then is actually natural, because in low levels of illumination, we only detect shades of grey, not chromatic colors.
Sleep: Dreaming Data and Theories
The dream self An element present in almost every dream is the dream self. Usually, we experience our dreams from the embodied first-person perspective, in a similar fashion that we experience our waking reality. The dream self most often possesses a body-image much like the one we have while awake, and we are positioned in the center of the dream world, actively partaking the dream events. In this respect, the dream self is not all that different from the waking self. Sometimes, however, we can have a camera-like perspective on dream events and observe the dream and even ourselves from a third person’s point-of-view. What is different between the waking self and the dream self are memory lapses, confabulation, and lack of insight into our own deficient cognition. The dream self often has a limited access to his or her autobiographical memory, suffers from transient amnesia, and is disoriented to time and place. While we may remember some facts concerning our lives correctly when dreaming, often we lose the ability to contemplate whether the events, persons, places, or objects in our dreams are possible. For example, we can meet dead friends and relatives without the realization that they have, in fact, died years ago. We can also create false memories in our dreams, and not be able to reflect on the peculiarity of the dream. For instance, we can confabulate characteristics for our known dream persons, which they do not have in reality, like a different hobby or profession. Sometimes we manage to create ‘friends’ or ‘relatives’ that we do not have in our waking reality, and we have no insight into the fact that these people do not exist in real life. In fact, we are often totally unable to reflect upon the credibility of our own beliefs in our dreams. The lack of full self-awareness can extend to ourselves as well, although this is quite rare. Although the dream self in most dreams appears much the same as during wakefulness, there is a small proportion of dreams in which the self appears in an altered form. The milder variations include cases in which the dreamer is the same person, but appears in strange clothing or is of a different age, say, a small child in the dream. In a more distorted form, we can appear as a completely different person, sometimes of
343
the opposite sex or different race. In very rare occasions, we can even transform into an animal, as in the following dream excerpt. I lived in a small house in the countryside, and I had several dogs. I let the dogs out, to roam free. They run into a nearby field, and then I saw the reason for this: a huge elk was standing in the middle of the field, with magnificent antlers. My dogs wanted to hunt down and kill the elk, and suddenly I was one of the dogs, the leader of the pack. I felt myself running on four legs through the hay, fast and strong. Then I realized that I and the other dogs were no longer dogs, but fierce wolves. Bloodlust was filling my mind, I could almost taste the raw elk meat in my mouth, and drool oozed between my sharp canines. The elk started to run away from us, and we followed, hunting in a coordinated fashion to bring down our prey. (First author’s dream, reported 11 August 2007)
Other basic dream content elements The dream self is not alone or passive in the dream world, but surrounded by other dream characters, actively taking part in events and social interactions. The Hall and Van de Castle content analysis system has revealed in numerous data samples that on average our dreams contain three other animate characters, which are usually human, but sometimes animals or fantastic or fictional characters. Approximately half of the humans in our dreams are persons known to us, and the other half consists of strangers and occupational characters, such as doctors or policemen. The dream self is also an active participant in approximately four out of five dreams, and an uninvolved observer only occasionally. The dream self interacts with other dream characters and various social interactions take place in dreams. Interactions with other characters are more often aggressive than friendly. Almost half of our dreams include an aggressive component, and the dreamer is personally involved in most of these, more often being the victim than the aggressor. Males dream more often of direct physical aggression than females, and females of friendly interactions. Sometimes the positive interactions take on an erotic tone, although more often in men’s than in women’s dreams. Events in which a bad outcome occurs to a dream character independent of anything the character does are called misfortunes. These are,
344
Sleep: Dreaming Data and Theories
by definition, negative events, consisting of mishaps, dangers, accidents, loss of possession, injuries and illnesses, falling, and death. Misfortunes are seven times more frequent in dreams than the opposite type of events, good fortunes. Moreover, the dreamer is personally involved in almost three quarters of misfortunes. In a series of studies investigating negative dream content, conducted by the authors, it has been found that threatening events are frequently experienced in dreams. Approximately two-thirds of all dreams contain one or more event that threaten the well-being of the dream self or his or her significant others. The most common type of threatening events includes aggression, and failures and accidents are also common, whereas disease and illness are seldom encountered. Of the threats we face in our dreams, about half endanger our well-being in a way that would seriously affect future survival opportunities if the threats were to occur in waking life. We also have a full range of emotional experiences in our dreams. Negative emotions are more frequently experienced in dreams than positive emotions. In the classic Hall and Van de Castle study of home-reported dreams, of the over 700 emotions explicitly mentioned in the 1000 dream reports of college students, 80% were negative and only 20% positive. Later, Frederick Snyder, and Inge Strauch and Barbara Meyer have acquired similar results in REM dreams collected in laboratory, indicating that two-thirds of emotions in dreams are negative. They also found that the most commonly reported negative emotions are fear and anger. Furthermore, dream emotions are almost always appropriate to the dreamed situations – a finding compatible with the results indicating that the events in our dreams are often unpleasant. There are, however, a couple of more recent studies that have found a more balanced sample of positive and negative emotions in dreams. Thus, dream emotions deserve further investigation. Reflection Reflective consciousness involves the ability to focus on some particular aspect of the content of consciousness and pass a judgment over it. The traditional view has it that dreams are devoid of reflective consciousness and self-awareness. This
traditional view may now be changing. During dreaming our ability to critically reflect upon the events we witness is diminished, but it is not completely wiped out. Already in the early 1980s, McCarley and Hoffman found that in 16% of 104 REM dreams the dreamer explicitly noticed some bizarre feature of the dream. That is, reflective awareness of something not being quite right was present, at least for a fleeting moment. More recently, David Kahn and J. Allan Hobson divided reflective thinking in dreams into two distinct components. The first component, reflecting on the dream event itself, that is, thinking about the event, is fundamentally deficient, and very different from waking state thinking. We take the dream event for granted, and do not question whether this event is likely or even possible in the waking realm. The second cognitive component, however, thinking and reflecting within the event, within the dream scenario, is similar to our thinking while we are awake. Even though we take the confabulatory dream event itself as real, we can think, pass judgments, make decisions, and guide our behavior the way we do while we are awake. For instance, the dreamer can protect her baby in the dream from dangers in appropriate and reasonable ways, but fail to realize that she does not have a child in waking reality. In sum, while we may not realize that the overall dream event itself is impossible or improbable compared to waking life, we act and think in the situation the same way we would if we were awake.
Bizarreness Our sleeping brain combines dream elements in a novel, creative, odd, sometimes absurd way that is impossible or highly improbable in the waking realm. These dream oddities are referred to as bizarreness. J. Allan Hobson categorized dream bizarreness into three distinct forms: incongruity, discontinuity, and vagueness. Incongruity refers to mismatching dream features appearing together, for example, meeting your dead grandmother in your kitchen. Discontinuity comprises the sudden appearance, disappearance, or transformation of dream elements, and vagueness refers to uncertainty or ambiguity of a dream element, for instance, recognizing the dream character
Sleep: Dreaming Data and Theories
as a familiar person but not being able to specify who exactly the person is. A few content analysis studies on bizarreness have been conducted. In one study by Antti Revonsuo and Christina Salmivalli, the frequency of bizarre elements was compared to the frequency of corresponding nonbizarre, normal-appearing elements. It was found that most dream elements are, in fact, nonbizarre, and not distorted in any manner. When bizarre features were present, they were more often incongruous than discontinuous or vague. A similar study has been conducted with dream characters by Revonsuo and Tarkko, investigating whether dream people are more often intrinsically distorted (internal bizarreness) or appear in a wrong context (contextual bizarreness). About half of the human characters in dreams were found to be bizarre, and two-thirds of these were contextually bizarre while the remaining one-third internally bizarre. Bizarreness is an intriguing phenomenon as it reveals how the unity of consciousness is disrupted in dreams. If we were to figure out what are the underlying mechanisms in how the brain fails to bind together information during dreaming, we might learn about how the unity of consciousness is created by the waking brain. This is called the binding problem: we do not know how the brain mechanisms account for the unified perceptual world we take for granted. Absent dream content Self-awareness and the ability to reflect on the nature of dream events are flawed during dreaming. In dreams, we lack insight, that is, we take for granted the reality of dreams and fail to understand its hallucinatory characteristic. This may be explained by the specific pattern of brain activation during REM sleep: the dorsolateral prefrontal cortex where the neural mechanisms necessary for full self-awareness are believed to reside is very active during wakefulness, but deactivated in REM sleep. There are also some specific dream contents that are almost nonexistent or at least very rare. These include such activities as reading, writing, calculating, and using a computer. The absence of these dream contents seem striking, as they are activities that we engage in daily, sometimes several hours a day. Thus far, there is no all-exclusive
345
explanation as to why elements so frequent in our waking lives do not enter our dreams. Bad Dreams, Nightmares, and Recurrent Dreams Occasional nightmares are commonplace; almost everyone has experienced one or more dreams that contain anxiety or outright fear. A nightmare is a dream which causes a strong unpleasant emotional response in the sleeper, typically fear or horror, and awakens the dreamer, while bad dreams are defined as very disturbing dreams that do not awaken the dreamer. When asked retrospectively how often we experience bad dreams or nightmares, we have the tendency to underestimate the prevalence of these unpleasant dreams. Antonio Zadra and Don Donderi compared retrospective estimates with systematic dream diaries, and found that bad dreams and nightmares are more common than we tend to think. And for some of us, they are more frequent than we would like. Approximately 5% of men and 10% of women report frequent bad dreams and nightmares, that is, one or more per week. Also, children, especially between ages of 3 and 6, frequently experience bad dreams or nightmares. The most common themes of bad dreams and nightmares are those of being pursued or attacked, being stuck in slow motion and unable to move, being naked in public, being late or unprepared for class, or scenes of falling, drowning, or death. Sometimes the dream may be recurrent, that is, the same theme keeps on repeating while the detailed content of the dream changes. Repetitive dreams, in contrast, refer to dreams that are identical in actual content. Research has shown that most recurring and repetitive dreams are described as being unpleasant. Some people seem more prone to nightmares than others. A personality measure proposed by Ernest Hartmann, thin and thick boundaries, correlates with nightmare frequency. Thin boundary people are open, trusting, and vulnerable, whereas thick boundary people do not let anything rock their world. People who rate high on thinness of boundaries report more nightmares than people with thick boundaries. While most nightmares have no apparent cause, psychological stress seems to precipitate nightmares.
346
Sleep: Dreaming Data and Theories
Extreme psychological stress, for example, exposure to a traumatic event, may trigger posttraumatic nightmares. These nightmares often repeat the traumatic event or depict some aspects of it. In a worst-case scenario, posttraumatic dreams may become chronic and last throughout life, but usually posttraumatic nightmares disappear as the person recovers from the trauma. Some repetitive and recurrent dreams can also be traced back to physiological causes, such as high fever or a need to urinate. Differences in Dream Content A wide range of cross-cultural studies on dream content reveals that the content analysis scale developed by Hall and Van de Castle provides similar results in most of the cross-cultural samples of dream reports studied. Thus, the formal features of dreams are alike all around the world, independent of culture. The small variations in dream content from culture to culture seem to relate to unique cultural patterns, and become understandable in terms of what is known about the culture. For example, the dreams of huntergatherer Mehinaku Indians living in the Amazon often include wild animals and hunting themes – elements present in their everyday lives. Male and female dreams are also very much alike. The main differences are that men twice as often as women dream of direct physical aggression. Also, sexual themes are two to three times more common in men’s than in women’s dreams. These gender differences also seem to be crosscultural and reflect the small but universal differences between the sexes. Children’s dreams as opposed to adults’ dreams contain more aggression. Children also dream more about animals. The younger the child, the more often the animals are wild and exotic, and as the child matures, domestic animals start to dominate the dreams. Dream content is powerfully modulated by real-life events that cause high levels of psychological stress. Posttraumatic nightmares are commonly reported by both children and adults who have experienced threatening real-life situations. The content and nature of posttraumatic dreams and nightmares are affected by several variables,
for example, the number of deaths or the extent of loss experienced, the degree to which the event in question has been personally witnessed, the perception of personal threat or threat to significant others, the prominence of specific traumatic imagery, the phase of recovery, the specific meanings attributed to the event, the individual experience, and personality. The higher the degree of personal threat involved, the more probable is the occurrence of posttraumatic nightmares. The effect of trauma has been studied in different populations, for example, veterans of war and children living under military rule. Veterans frequently report posttraumatic dreams depicting actual events experienced during the war. Traumatized children’s dreams are also qualitatively different from nontraumatized children’s dreams. More aggression, more persecution, and more negative emotions are present in traumatized children’s dreams.
Theories of Dreaming To this day, we know quite a lot about the form and content of dreams. But why do we dream in the first place? This question has puzzled both laymen and researchers for ages, and several different types of theories have been suggested to explain dreaming. How do these different theories manage to explain the form and content of dreams?
Psychoanalytic Theories In psychology, the highly influential views of dreaming were presented by Sigmund Freud and Carl Gustav Jung. Freud’s idea was that the manifest content of dreams, dreams the way they appear, is only a reflection of the latent or hidden content of dreams that needs to be disguised in order to protect the sleeper. If we were to become aware of our subconscious wishes, often of sexual or aggressive nature, the anxiety provoked by these wishes would awaken us. Thus, anxiety-provoking subconscious wishes are transformed into the manifest content of dreams, and we can only access the true latent content with dream interpretation techniques.
Sleep: Dreaming Data and Theories
For Jung, our nightly dreams were not trying to hide anything from us. He believed that our dreams were trying to communicate with us, but that the language of dreams, the images and symbols, made dream messages difficult to understand. He treated dream images as symbols, which have meaning in and of themselves. Jung believed the unconscious was purposely trying to speak to us, in order to bring forth a sense of wholeness and added meaning to our lives. Jung divided the unconsciousness into personal and collective. Collective unconsciousness is the part of the psyche that retains and transmits the common psychological inheritance of mankind. Archetypes are symbols that occur in mythology, fairytales, and religions, and take the form of mythological or primordial images. Archetypes reside in and form the collective unconscious, and are shared by all mankind. According to Jung, dreams have a compensatory function; for example, when the dreamer is troubled, he or she will dream of archetypal images, dreams whose aim is to right an imbalance in the psyche of that individual. At the time Freud and Jung formulated their theories, very little was known about the form and content of dreams. They based their ideas on unsystematically collected, small samples of dreams reported by only few individuals, not on the statistical properties of dreams in representative, systematic samples of dream reports. Thus, today Freud’s and Jung’s theories are not considered to be scientifically valid explanations for the form and content of dreams, or why we dream, although these theories are influential in personality psychology. Random Activation Theories The random activation theories (RATs) are a cluster of views that state that dreaming is a by-product of neuronal activity during sleep, that is, nothing but a nonfunctional REM-sleep-related epiphenomenon. The RATs suggest that random neuronal activation during REM sleep erratically activates memories, and the brain then synthesizes or constructs dream content based on these randomly activated memories. The specific activation pattern is determined by the neurophysiological
347
and neurochemical processes occurring during REM sleep. The form and content of dreams is thus seen as a functionless side effect of sleeprelated brain activation, serving no purpose whatsoever on its own. The biggest problem of RATs is that they cannot explain why the form of dreams is so well organized. If memories are randomly activated and then dream content is constructed based on them, should not our dreams consist of chaotic and flash-like mental images and sensations? The RATs do not manage to specify why or how the brain creates such a rich and coherent world analog or world simulation during sleep that we experience as dreams. Even if we accept that random activation could lead to organized dream constructs, we should not be able to pinpoint any over- or underrepresentation in any dream contents that would be similar for all individuals, such as the presence of realistic dream self, the distribution of sensory modalities, or the prevalence of negative emotions in dreams. Continuity Hypothesis The continuity hypothesis (CH) states that there is a continuum between waking experiences and experiences within dreams in a way that dreaming reflects waking life experiences. Several studies actually demonstrate this. Up to 65% of dream elements can be linked to waking experiences, and isolated spatial or temporal features of memories have been found to occur in approximately one-third of dream reports. However, exact episodic memory replays are present only in about 1% of dream reports. It has also been demonstrated that episodic memories are depicted in dreams most likely on the subsequent night, and again about a week later. In contradiction with CH, many such activities that we engage in our daily lives, such as writing, reading, calculating, and using a computer, are almost absent in dreams. In fact, the specific predictions and hypotheses of CH have not been described in a clear manner. The most general form of CH seems to predict that the frequency and nature of any real events will be correlated with the frequency and nature of similar dream events in subsequent dream content. On the basis
348
Sleep: Dreaming Data and Theories
of findings that some elements are almost absent in dreams, whereas others, like negative emotions and events, are overrepresented in dreams compared to waking reality, the CH does not seem to be the most reliable explanation for the form and content of dreams. Furthermore, CH does not claim anything about the possible function of the continuity between dreaming and waking, but, as it has been demonstrated that some waking life experiences are incorporated into dream content, rather states the obvious. What is lacking here is the specification for why some autobiographical memory elements are more likely to be incorporated into dream content than others. Psychological Healing Function of Dreams One influential view within clinical dream theories, supported by many esteemed researchers such as Ernest Hartmann, Rosalind Cartwright, and Milton Kramer, states that dreaming maintains psychological balance. Dreaming is seen as functional because it somehow enhances the individual’s coping capabilities to his or her current concerns, solves current waking life problems, and promotes psychological well-being, and even helps to recover from traumatic experiences. Dreaming helps us to psychologically adjust to any current life stressors that consume psychic energy. It is assumed that the process of dreaming makes connections in the memory networks more broadly than focused waking cognition, and this making of connections is guided by the dominant emotional concern of the dreamer. Negative emotional memory traces in long-term memory mark situations that cause a psychological imbalance, and these traces need to be integrated with other material in the memory networks in order to be neutralized. When the neutralizing process is successful, the emotional concern will disappear. Consequently, the psychological healing theories explain why negative emotions are incorporated into dream content. The clearest examples mentioned in the context of the psychological healing function of dreams are posttraumatic nightmares. Traumatic experience is stored into long-term episodic memory, and it is the dominant emotional concern for the dreamer. Therefore, in posttraumatic nightmares, the emotional
content of the traumatic event is repeated over and over again until the traumatic memories are integrated in the memory networks and the trauma is resolved. If the integration process is not successful, posttraumatic dreams may persist for years after the original trauma. In fact, chronic posttraumatic dreams are interpreted as a failure of dream function. Even though the psychological healing theories are intuitively appealing, some difficulties explaining empirical findings have emerged. To begin with, even though a strong correlation exists between first experiencing a traumatic event and later dreaming about it, there is no evidence that dreaming about the trauma somehow causally contributes to psychological recovery. The currently available empirical evidence on the healing or mental health function of dreams is correlational at best, and recent findings imply that posttraumatic dreams depicting terrifying dream images, even integrating them with other memory traces, do not lead to recovery or psychological healing. Thus, the changes in dream content from nightmarish posttraumatic dreams to normal dreaming may only reflect the recovery process that as such goes on independently of dreaming. Evolutionary Functions of Dreams The most recent types of theories put forth to explain the form, content, and function of dreaming originate from an evolutionary biological or psychological perspective. The closest physiological correlate to dreaming, the REM stage of sleep, is energetically costly, and natural selection does not retain costly characteristics unless they provide advantages. Thus, natural selection would not have favored REM sleep and dreaming unless the benefits overran the costs that went into producing them. The questions asked by the evolutionary psychological dream function theories are as follows: Why did the human brain evolve so as to be equipped with an automatic, involuntary program that generates complex subjective experiences during sleep? Is it conceivable that dreaming about particular contents in the ancestral world was functional in the biological sense – that somehow our ancestors had better chances of surviving because they regularly dreamed during the night,
Sleep: Dreaming Data and Theories
and especially because they regularly dreamed about particular things? When we expose dreaming to an evolutionary analysis, and if we find that dreaming may have increased ancestral survival rates and production of offspring, then we would understand why the human mind is a dreaming mind in the first place. The sentinel function Perhaps the first theory to explicitly present an evolutionary biological approach to sleep and dreaming was presented already in the 1960s by Frederick Snyder. He had noticed that REM sleep is unique only for mammals (although later identified also in birds), and suggested that REM sleep provided early mammals with a selective advantage over the reptiles that dominated the planet when mammals evolved. While long periods of sleep allowed early mammals to minimize metabolic requirements, it also rendered them vulnerable to predators. The selective advantage of REM sleep was to guard the safety of the sleeping animal. First, REM sleep increases the activity level of the brain and prepares the animal for a brief awakening that takes place after each REM period. If during the awakening danger is detected, the activity level of the brain is already high and the animal is prepared for immediate fight or flight. Snyder called this the sentinel function of REM sleep. The sentinel theory as such does not state anything about dreaming, but Snyder also proposed that nonhuman mammals dream during REM sleep. The nature of their dream imagery is modulated by the latest estimate of expected danger at the time of waking up. If the available evidence from the environment predicts approaching danger, the dream will prepare the animal for fight or flight. If no dangers are in sight, then the dreams will preserve continuity of sleep with pleasant dreams. Thus, the evolutionary function of dream content is to be predictive of and preparatory for the situation where the animal finds itself immediately after the currently ongoing REM dream. Snyder’s theory places the function of dreaming to the context of REM sleep evolution. This makes his theory difficult to test, as we know nothing of what ancient mammals dreamed about, not even whether they dreamed at all. This applies to modern mammalian species as well. What we do know
349
from human studies is that external stimuli, such as sounds, smells, or tactile experiences, are detected by the sleeping brain as activity in specific sensory cortices increases when stimuli are presented, but that the stimuli are not easily incorporated into dream content. Thus, if incorporation is infrequent, how would the animal brain be able to predict approaching danger from external environmental cues? The costly signaling theory Patrick McNamara suggests that costly REM features affect dream content, and, through producing hard to fake emotional signals, increase fitness by facilitating courtship displays or cooperative social interactions. The costly signaling theory (CST) builds on the concept that first emerged in the context of sexual selection theory presented by Robert L. Trivers. Sexual selection theory suggests that some traits have evolved because the more resources the organism has to allocate for the production of a particular trait, the more honest is the advertisement of ‘having good genes. When members of the opposite sex can enhance their fitness by selecting mates who possess good genes, they will favor individuals displaying costly, honest, and hard to fake signals. REM sleep involves high metabolic costs for the organism. The costs result from physiological changes taking place during REM, such as high levels of brain activation, increases in cardiac and respiratory rates, paralysis of skeletal muscles, REMs and limb twitches, thermoregulatory lapses, changes in growth hormone release, and penile erections. According to CST, REM features influence dream content and affect the mood states and emotional displays of the individual during following wake periods, regardless of whether specific dream content is remembered or not. Dreams are an emotional burden, and this burden is the greatest after having a highly memorable negative dream that places the dreamer into a ‘handicapped’ position. If the individual is able to display appropriate and functional behavior despite the emotional burden that is carried into waking realm, the emotional signals emitted by the individual during wakefulness are honest and hard to fake, and consequently, she or he is more likely to be favored by other group members, for
350
Sleep: Dreaming Data and Theories
instance, selected as a mate or preferred over others in cooperative alliances. The problem with the CST is that the theory does not specify how exactly behavior is affected by REM sleep and dreaming. The theory does not define what the external behavioral cues are, supposedly modulated by preceding REM sleep and dreaming, which the other group members can monitor and evaluate. Unlike typical costly features that play a role in mate selection (as the peacock’s tail), the costly features of REM sleep and dreaming might not be as such directly observable. However, some directly observable characteristics must be highly and reliably correlated with REM sleep and dreaming so that by preferring these features, the costly features of REM sleep and dreaming would consequently be selected for, albeit indirectly. Furthermore, we should be able to verify that other group members actually find these behavioral cues desirable and that the cues affect favorably the selection of the individual as mate or cooperative party. The psychological problem solving and creativity function Ernest Hartmann, who has suggested a psychological healing function for dreams, has also proposed that the capacity of dreaming to form new connections within the neural networks of the brain might have served two functions useful for our ancestors. Hartmann’s first evolutionary argument concerns the idea that dreaming, as a process making new connections, has a kind of psychological problem solving or psychotherapeutic function, especially in handling traumatic experiences: Dreams after trauma may appear to represent a rare situation and fortunately for many of is this is so. However . . . . one hundred thousand years or so ago, when the human brain was gradually developing to its present form, our lives were considerably more traumatic; the after-effects of trauma may well have been an everyday reality and the resolving of trauma a constant necessity. . . . (Hartmann, 1998: 158).
Hartmann’s second evolutionary proposal is that the making of broader and wider associations during dreaming might have helped in bringing material together in new ways and this would have been useful to our ancestors in their waking lives.
Dreams might have provided new innovations in exploiting resources and solving problems related to everyday life. This particular argument emphasizes the creative and problem-solving nature of dreams. . . .In other words, the functions of dreaming. . . may have been especially important for us at earlier times in our species’ development. . .Only our dreams gave us a chance to do this – to make broader and wider connections, to integrate trauma or other new material, and also to bring material together in new ways that occasionally might have been useful to us in our waking lives. (Hartmann, 1998: 209)
In sum, Hartmann suggests, at least indirectly, that the evolutionary origin of the function(s) of dreaming is in the ancestral environment where life was dangerous, trauma resolution was required on a daily basis, and creative new ideas could have provided valuable selective advantage. Individuals who were able to regain emotional balance and well-being after trauma were better off than those who did not, and therefore the psychotherapeutic function of dreaming was selected for and became a universal feature of the human mind. Similarly, individuals with creative problem-solving dreams were better adapted to their environment and consequently left more offspring than individuals not having extra help from their dreams. We have already dealt with the problems related to the psychological healing theories. In addition, empirical evidence on the creative and problem-solving nature of dreams is mostly negative. New and useful solutions to waking life problems are extremely rarely depicted in dreams. Thus, whatever the function of dreaming is, it does not seem to be the finding of creative, new, and useful solutions to problems faced in waking life. Therefore, these two suggestions for the evolutionary function of dreaming are not supported by the currently available empirical evidence. Simulation functions of dreams Three different simulation function ideas have been suggested: that dreams are similar to play behaviors in mammals, dreams are simulations of social interactions, and that dreams are specialized in the simulation of threatening events. All these views are based on the claim that the dream experience is
Sleep: Dreaming Data and Theories
functionally constructed to resemble waking experiences, and therefore shows clear design features for a world simulation function. Dreaming as play
There are indeed several similarities between dreaming and play behavior, but also a number of dissimilarities. Both seem to be limited, in their clearest forms, to mammals only. Both can simulate reality and rehearse different types of situations and interactions in a safe context. Both may exaggerate, transform, and display enormous variation of behaviors that are originally related to other contexts outside play. Both are energetically costly, biologically programmed behaviors that should therefore be in some way useful. Unfortunately the adaptive functions of play are not entirely clear; thus, it is not possible to explain the functions of dreaming by saying that they are similar to the functions of play behavior. Most likely play has multiple functions, some of which might be closely related to the threat simulation function of dreams. There are forms of mammalian play that seem to be rehearsals of hunting behavior, aggressive encounters, or predator avoidance. Thus, playing and dreaming might have complementary functions in the rehearsal of behaviors: dreaming is perceptually more realistic than play, whereas play is motorically more realistic, involving actual execution of motor programs, muscular movements, and physical exhaustion. Dreaming as social simulation
Another currently popular suggestion is that dreaming simulates human social interactions and rehearses social perception and social skills. Several slightly different versions of the ‘social simulation hypothesis’ have been proposed. The ‘social mapping hypothesis’ suggests that dreaming allows simulation of self, location, and awareness of others, including awareness of their internal mental states. Dreaming is thus suggested to have rehearsed the perceptual and emotional features required in successful social mapping in human evolutionary history, eventually leading to the emergence of self-awareness. Relatedly, David Kahn and J. Allan Hobson emphasize that awareness of what others are thinking and feeling is a robust aspect of human consciousness, and this
351
aspect is maintained during dreaming despite the changes in chemistry and activation patterns of the brain during sleep. Thus, even though Kahn and Hobson do not explicitly express it, they imply that awareness of the minds of others (‘theory of mind’) during dreaming might have contributed to the ability to anticipate the intentions of others while awake. Dreaming about the intentions of others could prepare us for social encounters when awake. This idea leads us to another version of the social simulation hypothesis, suggested by several different researchers. As many of the selection pressures faced by ancestral humans were posed by complex human social life, modeling human relationships, and interpersonal bonds, for example, family politics, attachment, love affairs, and status battles, might have had adaptive value. Interacting with other members of the group was an important selection pressure in the ancestral environment, and simulation of skills such as how to find a mate, build coalitions, and avoid conflict would have been useful. In dreams, it is possible to practice dealing with complex social situations, and because those most adept in their social environment were likely to have the best access to resources in their social group, simulation of social situations would have been selected for. Furthermore, strong family and group cohesion would have enabled organized defenses against predators and other enemies and enhanced survival and health of group members. The social simulation hypothesis is to some extent compatible with what we know about the form and content of dreams. About half of the human characters in our dreams are persons familiar to us, and appearance, behavior exhibited by the character, and feelings evoked by the character in the dreamer are regularly used in the identification of the person. More than 80% of known dream characters evoke some kind of emotional response in the dreamer, most often affection or joy. A significant amount of time in dreams is spent wondering what other dream characters are thinking or planning. Thus, our dreams often represent human characters and give plenty of space for opportunities to practice social interactions. Nevertheless, aggression is a more common type of social interaction than friendliness, while sexual interactions are relatively rare in dreams. Thus, we get
352
Sleep: Dreaming Data and Theories
less practice in forming positive social bonds, such as making friends and allies, than in dealing with negatively toned social interactions. Even less time in dreams is devoted to mate selection and practicing how to form romantic relationships. Although the social simulation hypothesis is consistent with the fact that other people and multiple social interactions are frequently present in our dreams, the hypothesis has some problems as an evolutionary psychological account of the function of dreaming. First, we get a lot of practice in (nonthreatening) social interactions during our waking lives, and this practice does not have high costs. Thus, it remains unclear why it would be advantageous to practice or simulate something like that further in our dreams. Second, there is a lack of studies on the detailed nature of the social interactions in dreams, and the ones conducted reveal the often aggressive nature of social encounters. Do we, in fact, interact with other dream characters in reasonable ways that might be considered useful simulations of or rehearsals for reallife social interactions? To back up an evolutionary hypothesis, a detailed description of the type of dream content and the conditions under which it occurs is required, as well as a cost–benefit analysis that should show why the dream simulation is likely to be useful for us (or was likely to be useful for our ancestors). It remains open whether the social simulation hypothesis will receive support from the more detailed analyses of dream interactions and cost–benefit considerations. Dreaming as threat simulation
The third version of simulation function theories is the threat simulation theory (TST), proposed by Antti Revonsuo. According to him, to study the biological function of dreams we are required to make a systematic, detailed analysis of the content of dreams across a wide range of large data samples: in the ‘normal’ population, in cross-cultural samples, and in various special populations, especially hunter-gatherers, children, frequent nightmare sufferers, and traumatized individuals. If in this analysis some dream content characteristics tend to pop out here and there, again and again, those features probably are traces of the original biological function of dreams.
The TST is based on currently available evidence of the systematically recurring dream content characteristics. As mentioned above, the major statistically significant features of dreams are biased toward representing negative elements. Negative emotions and aggression are prominent dream content characteristics, the universally most often reported dream theme is the dreamer being chased or pursued, and the most frequent themes of recurrent dreams and nightmares consist of the dreamer being chased or attacked. The TST interprets this evidence by suggesting that dream consciousness evolved as an off-line model of the world that is specialized in the simulation of various threatening events encountered in the ancestral environment. In the ancestral environment, a threat simulation system that selected memory traces representing life-threatening experiences from long-term memory and constructed frequent threat simulations based on them could have provided our ancestors with a selective advantage in practicing threat recognition and avoidance skills. During dreaming, threat coping skills could have been maintained and rehearsed without the risks of hazardous consequences that accompany threats in real situations. Due to its beneficial effects in enhancing survival and reproductive success, the threat simulation mechanism was selected for, thus propagating its own existence in the ancestral environment. The predictions of TST have been empirically supported in several studies. Threatening events are frequent in normative dreams, the most frequent type of threat are aggressive encounters (varying in severity from verbal nonphysical aggression to escape and pursuit situations, and to direct physical aggression), the dream self is most often the target of the threatening event, approximately half of the threats pose a severe threat to the dream self, the dream self reacts to the threats in a relevant and adequate manner, and the source of the threat simulation is most often traced back to the personal experiences of the dreamer or to media exposure. Recurrent dreams and nightmares include more severe threat simulations; especially the dangerousness of events is exaggerated in these special dreams. Further, if we are exposed to traumatic events in our lives, our threat simulation system becomes highly activated. The more severe the
Sleep: Dreaming Data and Theories
trauma, the better the dream recall, the more threatening events dreams contain, and the more life threatening and severe the threats in dreams. In sum, the dream content studies on TST have mostly supported the predictions of the theory, although it is too early to draw any definitive conclusions about the accuracy of TST. The typical counterarguments against the theory include the following: (1) perhaps the high amount of threat-related content is only the result of selective memory for emotionally charged content; (2) perhaps the bizarre and disorganized nature of dream content does not allow realistic simulation of reality or real threats; (3) posttraumatic stress disorder (PTSD) and frequent nightmares are dysfunctional and disturb sleep, therefore they cannot be regarded as good or functional for the individual who suffers from them; (4) dreaming simulates so many other things too that surely threat simulation cannot be the (only) function of dreams; and (5) the TST is in principle untestable, and thus cannot be falsified or verified. To begin to answer this critique, even though emotionally charged dream content is easier to recall than mundane content, the same applies for our everyday memories. We tend to forget the ordinary events in our lives, and only remember those that had an emotional impact. When the frequency of threatening events experienced during a specific time period in the waking life was compared with threats simulated in dreams during the same period, it was clear that the dream world contains threats much more frequently than the waking life does. Second, practically all dreams are well-organized simulations of a world including the self, other characters, objects, and a setting or an environment where the dream takes place. Bizarreness disrupts some parts or features of this otherwise coherent organized world, but although dreams include bizarre elements, bizarreness appears to be a relatively small deviation in the otherwise coherently organized dream experience. Third, it is true that many severely traumatized individuals suffer from sleep disturbances due to terrifying nightmares. But there are also reasons to believe that ancestral humans did not suffer from the effects of PTSD to the same extent as some individuals in the present environment.
353
The threats in the ancestral environment were frequent, and the ancestral humans were most likely adapted to higher levels of stress and trauma from early childhood on than most contemporary humans. Moreover, in the ancestral setting the threats were often related to everyday activities and thus predictable. Fourth, although some researchers are willing to accept the idea that dreams are simulations of significant selection pressures, they disagree on what types of events are simulated in dreams. Rehearsal of skills such as adjustment to novelty, social interactions, interpersonal understanding, motor functions, and spatial learning has been suggested. Thus far, however, none of these suggestions have undergone a similar evolutionary cost–benefit analysis as the TST. The suggested alternative evolutionary functions involve little costs if practiced during waking hours, while real threats often result in fatal consequences. Thus, fitness benefits for simulating threats during dreaming are higher than for simulating situations that yield no costs if practiced during wakefulness. Finally, the main concern faced by the TST is its empirical testability. This concern is not unique only to TST, but to all evolutionary psychological theories that infer cognitive mechanisms from the selection pressures operating in the evolutionary environment of the species. We will never be able to acquire data that would tell us what our ancestors dreamed about thousands and thousands of years ago. But the more we learn about ancestral life and threats in that environment, the more specific hypotheses we can draw concerning what types of events should be simulated in dreams, and thus, indirectly test whether contemporary dream content is compatible with ancestral selection pressures. To sum up, the TST takes into account the selection pressures most likely present in the human ancestral environment. It proposes a plausible explanation for how dreaming of negative and threatening events might have provided a slight advantage to our ancestors in maintaining and enhancing their threat recognition and avoidance skills. By referring to a single threat simulation mechanism, it furthermore manages to explain a wide variety of dream content data that already exist in the research literature.
354
Sleep: Dreaming Data and Theories
Summary and Conclusions Dreaming can be defined as a subjective experience occurring during sleep and taking the form of an organized, temporally progressing world simulation. Even though dream content is subjective and highly individual, some dream themes, such as being chased or attacked, falling, drowning, or flying, seem to be universal among humans. Detailed dream content studies also reveal many common underlying elements dreams are composed of. Regardless, we still do not know why people dream or if dreams serve any function. Numerous theories have been proposed to explain why we experience dreams, and some of them possess more explanatory power than others. Nevertheless, none of the theories posed are thus far accepted by all researchers in the field of dream science. See also: Sleep: Implications for Theories of Dreaming and Consciousness.
Suggested Readings Barrett D (ed.) (1996) Trauma and Dreams. Cambridge, MA: Harvard University Press. Barrett D and McNamara P (2007) The New Science of Dreaming. Westport, CN: Praeger.
Domhoff WG (1996) Finding Meaning in Dreams: A Quantitative Approach. New York: Plenum Press. Domhoff WG (2002) The Scientific Study of Dreams: Neural Networks, Cognitive Development, and Content Analysis. Washington, DC: American Psychological Association (APA). Farthing WG (1992) The Psychology of Consciousness. New York: Prentice Hall. Garfield P (2001) The Universal Dream Key: The 12 Most Common Dream Themes Around the World. New York: HarperCollins. Hall CS and Van de Castle R (1966) The Content Analysis of Dreams. New York: Appleton-Century-Crofts. Hartmann E (1998) Dreams and Nightmares: The New Theory on the Origin and Meaning of Dreams. New York: Plenum Press. Hobson JA, Pace-Schott EF, and Stickgold R (2000) Dreaming and the brain: Toward a cognitive neuroscience of conscious states. Behavioral Brain Sciences 23(6): 793–842. Moffit A, Kramer M, and Hoffman R (1993) The Functions of Dreaming. New York: State University Press. Revonsuo A (2000) The reinterpretation of dreams: An evolutionary hypothesis of the function of dreaming. Behavioral and Brain Sciences 23(6): 877–901. Revonsuo A (2006) Inner Presence: Consciousness as a Biological Phenomenon. Cambridge, MA: MIT Press. Schredl M and Hofmann F (2003) Continuity between waking activities and dream activities. Consciousness and Cognition 12: 298–308. Strauch I and Meier B (1996) In Search of Dreams. Results of Experimental Dream Research. New York: SUNY Press. Valli K and Revonsuo A (in press) The threat simulation theory in the light of recent empirical evidence – A review. The American Journal of Psychology.
Biographical Sketch
Katja Valli has a PhD in psychology. She is currently a researcher at the Centre for Cognitive Neuroscience, University of Turku, Finland, and a visiting lecturer at the University of Sko¨vde, Sweden. She is a member of the Consciousness Research Group, led by Professor Antti Revonsuo. Valli has conducted research on dreaming since late 1990s, focusing on the biological function of dreaming. Her other areas of expertise include evolutionary psychology, sleep laboratory research, and sleep-related altered states of consciousness.
Sleep: Dreaming Data and Theories
355
Antti Revonsuo is a professor of psychology at the University of Turku, Finland, and a visiting professor of cognitive neuroscience at the University of Sko¨vde, Sweden. He is an associate editor of the journal Consciousness and Cognition, the director of the Consciousness Research Group at the Centre for Cognitive Neuroscience at the University of Turku, and a founding member of the faculty of an undergraduate degree program on consciousness studies at the University of Sko¨vde, Sweden. He has done research on consciousness since the early 1990s, focusing mostly on dreaming, hypnosis, visual consciousness and its neural correlates, and theories of consciousness.
Sleep: Implications for Theories of Dreaming and Consciousness T T Dang-Vu, University of Lie`ge, Lie`ge, Belgium; Centre Hospitalier Universitaire de Lie`ge Sart Tilman, Lie`ge, Belgium M Schabus, University of Lie`ge, Lie`ge, Belgium; Division of Physiological Psychology, University of Salzburg, Austria V Cologan, University of Lie`ge, Lie`ge, Belgium P Maquet, University of Lie`ge, Lie`ge, Belgium; Centre Hospitalier Universitaire de Lie`ge Sart Tilman, Lie`ge, Belgium ã 2009 Elsevier Inc. All rights reserved.
Glossary Coma – Behavioral state of unresponsiveness in which the individual lies with eyes closed, cannot be aroused and has no awareness of self and surroundings. K complexes – Brain electrical activity present during non-REM sleep, mostly stage 2, and appearing on electroencephalographic recordings as large amplitude and biphasic slow waves lasting more than 0.5 s. Locked-in syndrome – Syndrome induced by the disruption of corticospinal and corticobulbar pathways and characterized by sustained eye opening, awareness of self and surroundings, aphonia or hypophonia, and quadriplegia or quadriparesis; the patient usually remains able to communicate through vertical/lateral eye movements or blinking of the upper eyelid. Non-REM sleep – Stage of sleep, subdivided in four stages (1–4), and characterized by a slowing of the electroencephalographic activity, along with the appearance of spindles, K complexes (prominent in stage 2) and slow waves (prominent in stages 3–4), and a progressive decrease in muscle tone and eye movements. REM sleep – Stage of sleep characterized by low amplitude, high frequency electroencephalographic rhythms, rapid eye movements, complete muscular atonia, and intense dreaming mentation.
Sleep spindles – Brain electrical activity present during non-REM sleep, mostly stage 2, and appearing on electroencephalographic recordings as waxing-and-waning waves at a frequency of 12–15 Hz and lasting more than 0.5 s. Vegetative state – Behavioral state of unresponsiveness in which the individual lies with eyes usually opened, is awake but has no awareness of self and surroundings.
Introduction Sleep, dreaming, and consciousness are intimately connected processes that have fascinated humans since the dawn of time. Their conception however has dramatically evolved through the ages, and mostly in the last century. Until the discovery of rapid eye movement (REM) sleep, sleep was widely considered as a passive state of brain inactivity, resulting from sensory input reduction. This idea was further reinforced when researchers in the first half of the twentieth century demonstrated that sensory collaterals discharged into the brainstem reticular formation during wakefulness. The concept of the ascending reticular activating system was born: the active brain state of wakefulness is initiated in structures located in the brainstem core, receiving afferent input from sensory collaterals and projecting to distributed cortical areas of cerebral hemispheres. Conversely, lesions in this brainstem ascending reticular activating system produced a 357
358
Sleep: Implications for Theories of Dreaming and Consciousness
state of immobility with the presence of slow waves in the electroencephalogram (EEG), similar to what is found in natural sleep. The conclusion was that sleep resulted from the lack of brain activation underlying wakefulness. This passive theory of sleep was challenged when Nathaniel Kleitman and Eugene Aserinsky evidenced in the 1950s the recurring appearance during sleep of periods of rapid eye motility, associated with autonomic changes (irregular respiration, heart rate variability) and a high probability of vivid dream reports at awakening from this stage of sleep, which was since then called REM sleep or paradoxical sleep. Michel Jouvet later insisted on the concomitant active suppression of spinal motor activity. On the EEG recordings, REM sleep was characterized by low-amplitude, highfrequency brain wave patterns, resembling active wakefulness. Sleep could no more be viewed as a passive state of brain inactivity. Since the last decade, functional brain imaging studies have confirmed the persistence of regionally segregated brain activity increases during REM sleep, and more recently during non-REM sleep, in humans. In line with sleep as a dynamic and active process, there are now numerous experimental data supporting a beneficial role of both REM and nonREM sleep in the off-line mnemonic consolidation of learned information and skills. The concepts of dreaming and sleep evolved in parallel, as sleep is a necessary concomitant of dreaming. In the early Antiquity, dreams were viewed as messages from gods and prophecies. Aristotle challenged this common belief by emphasizing that dreams may be endogenously generated. For centuries, as long as sleep was considered as a homogeneous passive state, dreams were then regarded as transient interruptions from this quiescent state, potentially triggered by external or internal stimulation. Popular interest in dreaming was significantly enhanced with the emergence of the psychoanalytic approach of dream interpretation. Indeed, in the beginning of the twentieth century, Sigmund Freud formulated that dreams may be the expression of satisfaction of repressed desires or the ‘royal road to the unconscious.’ Biological aspects of dreaming mainly developed since the discovery of REM sleep. It was observed that this stage of sleep was associated with more frequent and vivid dream reports. Descriptive features of dreams were correlated
with findings on REM sleep physiology. In particular, functional neuroimaging studies more recently showed that the distribution of regional brain activity during REM sleep might also account for some hallmarks of dreaming content. The generation of dreams was thus supposed to be restricted to REM sleep. This concept also changed since dreaming also seems to occur during non-REM sleep, although reports from this sleep stage are more thought-like and less hallucinatory than REM sleep dreaming reports. Sleep and consciousness are intimately related, sleep being a reversible state during which consciousness spontaneously fades. Consciousness during sleep is certainly modified as compared to wakefulness. Awareness of the environment is significantly reduced during non-REM sleep, possibly in relation with altered functional cortical connectivity. In REM sleep, consciousness is modified and filled with self-generated oneiric mental representations. Despite this modification of consciousness, sensory stimuli continue to be processed during sleep. After a brief overview of normal sleep neurophysiology, we will discuss in this chapter some aspects of dreaming and consciousness related to sleep. Especially, biological aspects of dreaming, in the light of REM sleep physiology, will be reviewed. The question of consciousness during sleep will be approached, including the topics of sensory experience and brain connectivity during sleep. Sleep characteristics in patients suffering from altered states of consciousness will also be discussed. This chapter is not intended to be an exhaustive article about sleep, dreaming, and consciousness, which obviously cannot be fully covered in a single article. We hope to provide the reader with key information and broad concepts for the understanding of this constantly evolving area.
Sleep: Definitions and Mechanisms On a behavioral perspective, sleep may be defined as a normal and reversible state of perceptual disengagement from and unresponsiveness to the environment. It is also characterized by a diminished motor output and a typical position, usually recumbent with eyes closed. Sleep is further characterized by a homeostatic regulation.
Sleep: Implications for Theories of Dreaming and Consciousness
Stages
Sleep deprivation (or extension of the waking period) induces a compensatory increase in the duration and intensity of the next sleep. The recording of neurophysiological parameters allows to distinguish between two main stages of sleep in Homeotherms: REM sleep and non-REM sleep (stages 1–4). In humans, these sleep stages periodically recur in an ordered succession throughout the night, usually for a total of four to six cycles in a single night (Figure 1(a)). Each cycle begins with non-REM sleep, progressing from light
359
(stages 1–2) to deep (stages 3–4) and back to light non-REM sleep, then followed by REM sleep. Duration of deep non-REM sleep decreases and duration of REM sleep increases in the course of the successive cycles of a normal sleep night. Non-REM Sleep In humans, non-REM sleep is divided into light and deep phases according to specific EEG patterns (Figure 1(b)). Light non-REM sleep, mainly
M W R 1 2 3 4
0
1
2
3
(a)
4 Time (h)
5
6
7
8
EEG EOG W
N1
N2
N3
N4
R
EMG EEG EOG EMG (b) 1
N
2 –28 mm
3 –16 mm
–8 mm
Rd
Ra
Value
3 2
4
–4 mm
5
8 mm
6
24 mm
6 5 4 3 2 1 0
Value
4
5 4 3 2 1
Value
6
6 5 4 3 2 1 0
1 0
(c) Figure 1 Organization of normal human sleep. (a) Hypnogram of a normal sleep in a young healthy adult. This figure illustrates the dynamics of sleep–wake cycles in the course of a normal night of sleep. The black bars represent the time spent in a specific stage of sleep (or wakefulness). The time of the night is indicated on the horizontal axis. Sleep stages are indicated on the vertical axis. Note the progressive decrease in deep non-REM sleep (stages 3–4) duration and the increase in REM sleep periods across the night. M: movement time; W: waking; R: REM sleep; 1–4: non-REM sleep stages 1–4. Courtesy of Gilberte Tinguely. (b) Polygraphic patterns of sleep stages. The six panels show each a 20 s epoch of wakefulness (W), non-REM sleep (stages 1–4: N1–4) and REM sleep (R), respectively, with EEG (upper trace), EOG (middle trace) and EMG (lower trace). Scale bars on the left ¼ 75 mV. Courtesy of Gilberte Tinguely. (c) Functional brain imaging patterns of sleep stages. In comparison to wakefulness, brain areas significantly less active during non-REM sleep (N), more active during REM sleep (Ra), or less active during REM sleep (Rd). Reproduced from Maquet P, Degueldre C, Delfiore G, Aerts J, Peters JM, Luxen A, and Franck G (1997) Functional neuroanatomy of human slow wave sleep. J Neurosci 17(8): 2807–2812, with permission from Society for Neuroscience; Maquet P, Peters J, Aerts J, Delfiore G, Degueldre C, Luxen A, and Franck G (1996) Functional neuroanatomy of human rapid-eye-movement sleep and dreaming. Nature 383(6596): 163–166, with permission from Nature Publishing Group.
360
Sleep: Implications for Theories of Dreaming and Consciousness
stage 2, is characterized by the occurrence of sleep spindles and K complexes. Sleep spindles are waxing-and-waning oscillations at a frequency of about 12–15 Hz and a duration of more than 500 ms. K complexes are large amplitude and biphasic slow waves lasting more than 500 ms. During deep non-REM sleep, also called slow wave sleep, large amplitude slow waves become more abundant and constitute 20% to 50% (stage 3) or more than 50% (stage 4) of the EEG signal. Mechanisms of non-REM sleep generation involve complex interactions between the thalamus (reticular thalamic nucleus and thalamocortical neurons) and the cerebral cortex. At sleep onset, the thalamic neurons receive less activating input from the brainstem, become hyperpolarized, and therefore change their firing mode from tonic to phasic. During light non-REM sleep, GABAergic neurons of the reticular thalamic nucleus, due to their intrinsic membrane properties, burst in the spindle frequency range and entrain thalamocortical neurons in spindle oscillation. During deep non-REM sleep, thalamic neurons get more hyperpolarized and a clocklike delta rhythm (1–4 Hz) emerges from thalamocortical neurons. Thalamic oscillations are conveyed to the cortex, which in turn modulates their occurrence and grouping by a cortically generated slow rhythm (out-group) members. In summary, these studies provide evidence that perceptual processes may be automatically biased toward in-group members and that these early attentional differences may have implications for downstream memory and PNS activity.
Evaluation The brain is engaged in an ongoing and continually updating process of evaluating people in the social environment. Evaluations reflect a combination of positive and negative associations toward a person that can be used to guide behavior. Social neuroscience offers the potential for isolating component processes involved in evaluation and exploring how processing goals, motivations and social context affect these component processes. For example, the amygdala, a small structure in the temporal lobe, has been linked to an array of social and affective processes, including learning emotional information
382
A Social Neuroscience Approach to Intergroup Perception and Evaluation
and directing attention to important stimuli. There is also evidence that the amygdala even engages in affective processing when stimuli are presented outside of conscious awareness. The first fMRI investigation of intergroup processing examined amygdala activity while Black and White individuals saw both Black and White faces. This study found that the perception of racial out-group (>in-group) members is associated with the greatest amygdala activity. Because the amygdala plays an important role in affective processing and people generally prefer in-group members, this study is often taken as evidence for neural-mediated negativity toward racial out-group members. Several recent experiments have found that amygdala activity to racial out-groups is correlated with implicit measures of racial bias. Implicit measures are designed to measure relatively automatic aspects of evaluation that operate without deliberate thought and sometimes without conscious awareness. Evaluations based on more automatic processing may differ from evaluations that include more controlled processing. For example, most North Americans reveal negative biases toward Blacks, the elderly, or foreigners on implicit measures; yet they report egalitarian attitudes on questionnaires that allow more controlled evaluations of the same groups. Importantly, implicit measures can be a stronger predictor of discrimination than self-report measures in certain circumstances. These studies find that White individuals with a strong implicit preference for White (>Black) people tend to have stronger amygdala activity to Black (>White) faces during fMRI. These data provide an important link between the amygdala, a region known to respond to nonconscious presentations of stimuli, and measures of racial bias that do not rely on explicit report or necessarily conscious awareness. Indeed, the relationship between implicit measures of racial bias and amygdala activity to racial out-group members is even stronger when the faces are presented outside conscious awareness. Although more traditional explicit measures of racial bias that rely on self-report are not correlated with amygdala activity, they are correlated with ERP waves occurring approximately 500 ms after the presentation of faces. These data raise the interesting possibility that responses on explicit measures are linked to
downstream information processing, while implicit responses are linked to early, and perhaps nonconscious, information processing in the amygdala. In addition to valence (positivity/negativity), different groups may evoke specific emotions such as joy, disgust, anger, or fear, depending on the context and the stereotypes associated with the group. The emotions evoked during social perception are likely to impact subsequent information processing and ultimately behavior. Studies have examined brain activity while individuals watch various groups known to be associated with different emotions (envy, pride, pity, disgust). These studies have found that members of groups that typically trigger feelings of disgust (e.g., drug users, homeless people) are associated with less activity in the ventral medial prefrontal cortex (PFC) compared to other groups. The medial PFC is a region of the brain associated with mentalizing about others (i.e., inferring thoughts and intentions) and self-referential processing. The researchers interpreted this decrease in ventral medial PFC as initial evidence of dehumanization – the process by which individuals perceive or treat other people as less than human. These provocative conclusions extend intergroup evaluations beyond mere valence and underscore the importance of emotions on more reflective information processing.
Effects of Social Categorization on Evaluation There is an extensive evidence that intergroup social categorization and evaluation are closely linked. Perhaps the most compelling demonstration of this link has been classic research on minimal groups. Several famous studies have shown that the simple act of categorizing people into groups is sufficient to generate evaluative preferences for ingroup (>out-group) members – even when there is no competition or animosity between the groups. Moreover, anything that heightens the importance or salience of a given social category can have important downstream effects on evaluation. Several studies have explored the effects of intergroup social categorization on amygdala activity. One way to vary the salience of race as a social category is to vary skin tone (e.g., making White
A Social Neuroscience Approach to Intergroup Perception and Evaluation
faces darker and Black faces lighter). Interestingly, White individuals have increased amygdala activity to racial out-group members regardless of skin tone. However, these individuals also have increased amygdala activity to racial in-group members with dark (>light) skin tone. This pattern of results suggests that the salience of a social category (e.g., skin tone) can alter evaluative processing. Social categorization can also be altered by manipulating explicit goals to categorize individuals according to different dimensions. One study manipulated these goals by having Black and White participants categorize Black and White faces using explicit category labels (‘African-American’ vs. ‘Caucasian’) or perceptual information (e.g., matching one Black face to another Black face). When individuals categorized according to perceptual information, they had greater amygdala activity to Black than White people. However, there was no difference in amygdala activity when individuals categorized according to linguistic labels. This research offers evidence that explicit linguistic social categorization can lead people to engage in controlled processes that inhibit evaluative processing in the amygdala. Individuals can be perceived in terms of their category membership or individual characteristics and focusing on either aspect can modulate the brain regions engaged in evaluation, especially when attitudes toward the category and individual differ. Thinking about novel Black individuals in terms of their racial category membership should activate the amygdala, because this social category has potent evaluative connotations. However, thinking about them as individuals should attenuate this effect because novel individuals should be relatively neutral. As predicted, categorical judgments are associated with more amygdala activity to racial out-group Black (>White) people, presumably because categorical thinking activated evaluative information associated with the racial out-group. In contrast, individual judgments were associated with more amygdala activity to in-group White (>Black) people. These effects illustrate the flexibility of evaluative component processes as a function of current processing goals and the nature of social categorization. This approach has also been applied to the dehumanization of groups that evoke disgust. Recall that drug users evoke disgust and are associated with less
383
medial PFC activity relative to other groups. However, making individual (>categorical) judgments about drug users increases medial PFC activity. The authors argue that individuation may increase mentalizing about this group and thus increase the extent to which they are processed as human entities. These studies raise the possibility that certain groups may be perceived as less than human unless more in-depth processing goals are instigated. Social psychological theory also specifies that individuals vary in the extent to which they identify with or categorize themselves as members of social groups. These self-categorizations have important implications for the way in which individuals process members of other groups. Indeed, when individuals mentalize about the opinions and preferences of a person with the same political affiliation as themselves, a region of ventral medial PFC is active – an area of the brain engaged when people think about themselves. However, when they mentalize about a person with a different political affiliation, a region of dorsal medial PFC is active. Moreover, an implicit measure of self-categorization measured the extent to which individuals associated themselves with a political party (Republican or Democrat). Individuals with strong party associations had the strongest activity in the ventral medial PFC to a person who shared their affiliation and the weakest activity in the dorsal medial PFC to a person with the opposite party affiliation. The authors concluded that mentalizing about one’s self and people with a similar political affiliation (i.e., ingroup members) engages the same set of neural processes whereas dissimilar people engage a different set of processes, and the strength of processing varies as a function of the strength of self-categorization. This study suggests that self-categorization may play an important role in the manner in which people not only categorize and evaluate, but in the way they think about the mental states of others.
Effects of Motivation on Evaluation Although intergroup social categorization and evaluation have important implications for intergroup relations, much human behavior is ultimately modulated by values, goals, and motivations. The ability of social neuroscience to break down aspects
384
A Social Neuroscience Approach to Intergroup Perception and Evaluation
of intergroup relations into numerous component processes provides a lens for examining the complex interactions between social categorization, evaluation, and motivation. Emotion regulation and motivated processing have received extensive attention by affective neuroscientists and provide a framework for interpreting motivational and regulatory processing in the context of intergroup relations. This section will highlight the motivational flexibility of components involved in intergroup processing. The study of motivation and controlled processing in intergroup relations has focused extensively on motivations to respond without prejudice toward racial and stigmatized minority groups. People may have internal or external motivations for modulating their evaluations of certain groups. Internal motives include personal values and beliefs that one should be egalitarian and avoid acting or thinking in a prejudiced fashion. External motives include sensitivity to social normative or contextual constraints that discourage the expression or application of prejudice. These motivations appear to interact such that people who report the strongest internal motivations and the lowest external motivations are the least likely to display prejudice in any context. Several studies have explored the relationship between neural activity and these motivations to respond without prejudice. Initial research on this topic used startle eyeblink responses as a physiological index of amygdala activity. Startle eyeblink is measured by giving people a noxious blast of white noise and measuring the magnitude of their blinking reflex. Eyeblink magnitude increases in the presence of emotionally evocative stimuli and is linked to amygdala activation and is therefore used to index the degree of affective response. While individuals high in internal motivation report less racial prejudice, only individuals who are also low in external motivation reveal less racial bias on a startle eyeblink measure. These data offer initial evidence that affective central nervous system activity to racial out-group members is sensitive to different reported motivations. Subsequent research has explored the specific neural generators involved in implementing controlled processing during interracial perception. These studies have provided evidence that processes associated with control are engaged at very
early stages of intergroup processing. The anterior cingulate cortex (ACC), a region of the brain involved in the detection of potential conflict, may index initial controlled processing. It turns out that the ERN (error-related negativity), an ERP component generated by the ACC, is related to controlled processing during the ‘shooter task,’ a common measure of automatic stereotypes toward Blacks. Inspired by a horrific incident in New York city in which police shot an innocent Black man while he attempted to pull out his wallet, this task has individuals categorize rapidly presented images of tools or guns following the presentation of Black or White male facial primes. Incorrect responses that reflect racial stereotypes (i.e., tools that were classified as guns following a Black facial prime) are associated with larger ERNs. Moreover, individuals with the largest ERNs to incorrect, but stereotype-consistent responses show the highest levels of controlled processing across the shooter task, consistent with the proposed role of the ACC in signaling the need for additional controlled processing. The ERN can be further decomposed into different subcomponents that track initial dorsal ACC and subsequent rostral ACC activity. While the dorsal ACC is linked to the control of racial bias on the shooter task across conditions, the rostral ACC component is only associated with control in the shooter task among participants with a high external motivation when they are in a public situation. It appears that more rapid, automatic aspects of control are relatively insensitive to the interface between context and personal motivations, whereas later aspects of controlled processing are sensitive to context and motivation. This research helps distinguish multiple component processes that may be involved in enacting motivational agendas to control racial bias and provides additional evidence that later component processes are associated with responses on explicit measures. Although certain ERP waveforms can be localized to neural generators, fMRI offers superior spatial resolution for identifying the specific brain regions engaged during motivated processing. A number of studies find a negative correlation between amygdala activity to Black (>White) individuals and activity in the lateral PFC, suggesting that some degree of controlled processing modulates biased amygdala activity. Indeed, when faces
A Social Neuroscience Approach to Intergroup Perception and Evaluation
are presented outside of conscious awareness, White individuals have greater amygdala activity to Black (>White) faces, and this difference is highly correlated with implicit racial bias. However, when the same faces are presented long enough for conscious processing (c. 500 ms), individuals show greater activity to Black (>White) faces in the lateral PFC and ACC and less amygdala activity. Moreover, activity in these frontal regions, which are associated with cognitive control and emotion regulation, is correlated with the decrease in activity in the amygdala to Black (>White) faces under conscious processing. These data suggest that more controlled component processes operate on more automatic evaluative processes, perhaps in the service of motivations to control prejudice, illustrating the complex, interactive aspects of controlled intergroup processing. A number of social psychological studies illustrate the effects of interracial interactions on controlled processing. White individuals with pro-White racial bias on implicit measures show impaired cognitive ability following an interracial interaction, suggesting biased individuals need to engage in greater levels of controlled processing to successfully navigate interracial interactions. The extra regulatory effort required during interracial interactions leads to subsequent impairments in controlled processing. Interestingly, lateral PFC activity during the perception of Black (>White) faces mediates the relationship between implicit measures of racial bias and impaired cognitive control following an interracial interaction. This research illustrates the value of social neuroscience, showing the relationships between social phenomenon (interracial interactions and racial biases), cognitive processes (control) and neuroscience (lateral PFC), and using neuroscience to help isolate the cognitive operations that mediate realworld social behavior. There are also groups (e.g., the obese) about whom people feel no motivation to control their evaluations. Although these groups have received little attention in social neuroscience, there is evidence that these normatively stigmatized individuals activate regions associated with affective (amygdala and insula) and controlled processing (ACC and lateral PFC). Moreover, amygdala and lateral PFC activity are positively correlated
385
during the perception and evaluation of normatively stigmatized individuals. Although these data have been interpreted as evidence of inhibition, it seems plausible that controlled processes may also be recruited in certain circumstances to enhance (negative) evaluations toward certain groups. Research on intergroup relations has been heavily informed by basic research on attitudes and evaluations, including classic research on dissonance. Dissonance is an uncomfortable subjective state induced by conflicting or ambivalent feelings and individuals are often motivated to reduce dissonance by resolving the underlying conflict or ambivalence. A study used EEG to examine the possibility that egalitarian individuals forced to consider working with an angry (vs. happy) Black partner would be likely to experience conflict between their egalitarian motivations and the prospect of working with an angry partner. Indeed, egalitarian individuals had the greatest contingent negative variation (CNV) waves, an ERP component believed to index controlled processing, and took the longest to decide whether or not they would work with the potential partner under these conditions. In contrast, racially biased individuals are quick to reject angry, racial out-group members and have small CNV waves while contemplating their decision. Presumably, these individuals are disposed to reject Black partners and may feel especially justified when the potential Black partner is angry, making the decision to reject these racial out-group members quick and easy. Similarly, ambivalence between implicit and explicit measures of racial bias is associated with ventral lateral PFC activity to Black (>White) faces, providing additional evidence of the recruitment of controlled processing during intergroup ambivalence.
Complexity Although most laboratory research on intergroup relations focuses on a single social category (e.g., age, sex, race), social reality is complex and real world situations are more likely to be characterized by interactions with people who simultaneously belong to multiple social categories. One of the goals of intergroup research is to understand
386
A Social Neuroscience Approach to Intergroup Perception and Evaluation
how multiple categories are processed. One possibility is that certain categories (e.g., gender) are more important and therefore privileged in processing. A second possibility is that different categories are processing relatively independently and subsequently integrated to create some sort of weighted average during evaluative processing. A third possibility is that individuals use subtypes or superordinate categories. Social neuroscience provides a useful framework for addressing these possibilities by assessing the component processes and timecourse associated with different category information of complex individuals. The analysis of ERP waves can dissociate aspects of processing multiple social categories. When individuals see people who vary in race and gender, early processing (c. 100 ms) is more sensitive to racial outgroup (>in-group) members, while subsequent processing (c. 150 ms) is sensitive to differences in gender. Despite the rapid speed with which both of these categories are processed and the brief gap between them (c. 50 ms), the high temporal resolution of ERP offers evidence that race and gender are processed as separate social categories during very early stages of perception, perhaps before the visual information from the stimuli reaches conscious awareness. ERP studies illustrate that intergroup processing unfolds rapidly and that different social category information can be processed independently within milliseconds.
Behavior Ultimately, the study of intergroup relations is concerned with predicting and reducing discriminatory behavior. Several studies have linked neural and physiological processes to discrimination. One provocative study explored the relationship between cultural stereotypes and neural processes on the decision to shoot Black and White people in a videogame environment. Unsurprisingly, people who believed that Blacks were associated with criminality were more likely to shoot them if they were unarmed, making the false assumption that they were carrying a weapon even if it was an innocuous object (e.g., cell phone). Moreover, when individuals made the decision to shoot, they had greater ERP waves to Black (>White) faces
during early processing (c. 200 ms). These data suggest that individuals rapidly, and perhaps unconsciously, increase their attention to Black individuals when making the decision to shoot armed or unarmed individuals. Moreover, this ERP wave magnitude mediates the relationship between racial stereotypes and racially biased decisions on the shooting task (e.g., shooting unarmed Black individuals). Although these data are correlational, they suggest that the knowledge of cultural stereotypes may lead to increased attention to Black people and rapid acts of discrimination consistent with these stereotypes. Other research has linked responses in the PNS to discrimination. Whereas facial EMG is implicit in the sense that it can be measured without self-report, it has a slower timecourse than aspects of central nervous system processing indexed by ERP waves and may reflect the output of processing within the central nervous system. It is, therefore, unsurprising that facial EMG is linked to self-reported racial bias and overt acts of discrimination, such as hiring racial minorities. Indeed, facial EMG (specifically cheek activity, associated with smiling) toward racial minorities correlates negatively with decisions to discriminate against racial minorities during hiring decisions. In other words, people with the least positive evaluations of racial minorities according to their cheek activity (i.e., less smiling) are the most likely to exclude Black people from jobs. Interestingly, cheek activity was a stronger predictor of hiring decisions than a more traditional implicit measure of racial bias. One potential explanation for this relation is that facial EMG may reflect more downstream processing than traditional implicit measures, providing a closer temporal link to hiring decisions. Facial EMG also predicts racially biased decision making better than explicit measures, although facial EMG does converge with explicit measures of racial bias when individuals are not motivated to alter their biases toward racial minorities.
Conclusion The study of intergroup relations from a social neuroscience perspective remains relatively young.
A Social Neuroscience Approach to Intergroup Perception and Evaluation
Nevertheless, the last decade of research has revealed several important insights into the complexity of intergroup perception and evaluation. This research has provided exciting evidence of the automaticity of intergroup perception and evaluation, and the complex interactions between the component processes that guide behavior. These studies highlight the speed at which individuals distinguish different groups, their ability to do so without conscious awareness, and their ability to alter these initial processes according to their motivations or goals. Improvements in technology and convergence across methods will add precision and novel insights about an evaluative system that governs intergroup relations. In addition, the insights gleaned from social neuroscience will eventually lead to novel predictions for traditional behavioral investigations and ultimately interventions aimed to improve intergroup relations. See also: Implicit Social Cognition; Phenomenology of Consciousness; Social Foundations of Consciousness.
387
Suggested Readings Amodio DM and Lieberman MD (in press) Pictures in our heads: Contributions of fMRI to the study of prejudice and stereotyping. In: Nelson T (ed.) Handbook of Prejudice, Stereotyping, and Discrimination. Earlbaum Press. Crandall CS and Eshleman A (2003) A justificationsuppression model of the expression and experience of prejudice. Psychological Bulletin 129: 414–446. Cunningham WA and Van Bavel JJ (in press) A neural analysis of intergroup perception and evaluation. In: Berntson GG and Cacioppo JT (eds.) Handbook of Neuroscience for the Behavioral Sciences. Wiley. Eberhardt JL (2005) Imaging race. American Psychologist 60: 181–190. Ochsner KN and Lieberman MD (2001) The emergence of social cognitive neuroscience. American Psychologist 56: 717–734. Phelps EA and Thomas LA (2003) Race, behavior and the brain: The role of neuroimaging in understanding complex human behaviors. Political Psychology 24: 747–758. Stangor C (2000) Stereotypes and Prejudice. Philadelphia: Taylor & Francis Group. Stanley D, Phelps EA, and Banaji MR (2008) The neural basis of implicit attitudes. Current Directions in Psychological Science 17(2): 164–170.
Biographical Sketch
Jay J Van Bavel is presently a SSHRC postdoctoral fellow at The Ohio State University. His research explores the intergroup perception and evaluation using theory and methods from social psychology and cognitive neuroscience.
388
A Social Neuroscience Approach to Intergroup Perception and Evaluation
William A Cunningham is currently an assistant professor of psychology at the Ohio State University. Wil’s research takes a social cognitive neuroscience approach to understand the cognitive and motivational processes underlying affective and evaluative responses.
Subjectivity; First- and Third-Person Methodologies J-M Roy, University of Lyon, Lyon, France ã 2009 Elsevier Inc. All rights reserved.
Glossary Cognitive revolution – A new approach to the scientific study of cognition directed against behaviorism that took place in the mid-twentieth century. Explanatory gap argument – An argument to the effect that conscious phenomenal properties are subjective properties that cannot be explained on the basis of physical and biological properties. First-person investigation – An investigation of subjective properties as such and by the subject for whom these properties are subjective properties. Phenomenal consciousness – The consciousness of the phenomenal properties generated by our sensory experience of reality. Subjective property – A property that characterizes what an entity is from the point of view of a subject, versus what it is in itself. The notion is best illustrated by phenomenal properties of sensory experience.
The Cognitive Revolution and the Return of the Subjective Is a scientific knowledge of subjective properties possible? And if so, how? Does it require any substantial modification in the standard conception of science, especially from a methodological point of view? These questions are as ancient as the project of a science of the mental, since the mental was traditionally seen as the home of subjectivity. However, they only recently resurfaced as central problems in the foundations of contemporary theory of cognition, although the rehabilitation of the idea of a science of the mind is at the core of the anti-behaviorist revolt of the mid-1950s
known as the Cognitive Revolution from which this theory is born. The reason is that the mental is commonly considered as subjective inasmuch as it is conscious, and that the Cognitive Revolution did not go as far as rehabilitating the study of consciousness in its first step, operating what has been described as a surprising return of the mind without a return of consciousness. It is therefore only when the newly established cognitive science started, in the late 1980s, to realize that leaving out the conscious dimension of the mental was an error that it had to confront again the issues surrounding the notion of a scientific knowledge of subjective properties. This delayed return of consciousness led in fact to a serious crisis in the foundations of cognitive science, centered around the argument that its most basic assumptions are challenged by the subjective dimension of a central kind of consciousness, commonly designated as phenomenal consciousness. Different analyses of the resistance supposedly offered by the subjectivity of phenomenal consciousness have been proposed: to some, subjectivity is beyond the reach of scientific knowledge, while to others it is only beyond the reach of a naturalist kind of scientific explanation, and to others still only incompatible with specific forms of naturalist explanation, such as a reductionist or a functionalist one. It is essentially in the context of these discussions that a debate arose about the relevance of introducing again first-person methods of investigation into the study of cognition, in addition to the third-person ones only acknowledged as legitimate since the advent of behaviorism. This debate is focused upon the idea that a first-person study of the conscious and subjective dimension of the mental could be instrumental in overcoming the specific problem that this dimension raises for the program of cognitive science in the case of phenomenal consciousness. It echoes directly an independently motivated current of reflection on the nature and value of first-person knowledge
389
390
Subjectivity; First- and Third-Person Methodologies
that goes back to the early days of analytic philosophy, and is illustrated in particular by Roderick Chisolm, David Armstrong, Sydney Shoemaker, or Crispin Wright, although it has remained rather disconnected from this current so far. Interestingly, this debate also significantly contributed to the involvement of the tradition of continental philosophy in the analysis of the foundations of contemporary cognitive science, in particular, the phenomenological current initiated by Edmund Husserl and rooted in the concept of descriptive psychology of the school of Franz Brentano, because of its central concern for establishing a rigorous science of experience through a disciplined first-person investigation. The emergence of such a debate about the relevance of a complementary first-person methodology is of high significance for the evolution of contemporary cognitive science, since it corresponds to a third major step in the progressive rehabilitation of the prebehaviorist approach to the study of cognition. Indeed, after successively reinstating the scientific legitimacy of the notions of mental property and consciousness, contemporary cognitive science is now led to reevaluate some insights of the introspective experimental psychology founded by Wilhem Wundt, and developed later by the school of systematic introspection of his disciples Oswald Ku¨lpe and Edward Titchener, as well as by the Gestalt school. This renewed interest in first-person knowledge is not without precedent, as verbal reports in particular already stirred lively methodological discussions about introspection in the late 1970s. It nevertheless is an innovative development because of its larger proportions, its explicit connection with the problem of consciousness and the new ideas that it brings forth. Whether cognitive science will really take a first-person turn is still unclear. It is fair to say that the proposals remain for the most part tentative and programmatic, that they have had so far limited impact on many segments of the cognitive community, and that they are also confronted with strong criticism. The controversy, well reflected in a number of collective publications, is nevertheless important due to the historical significance and the theoretical depth of the issues involved, as well as its high stakes for the future of cognitive science.
Phenomenal Consciousness and the Notion of Subjective Property If phenomenal consciousness is the main source of the revival of the interrogation on the possibility of a science of the subjective, and thereby of a firstperson methodology, it is indispensable to first clarify what it means and in what sense it can be considered as subjective. We have in fact a pretty good intuitive understanding of what the expression of phenomenal consciousness refers to: it first and foremost designates, although not exclusively, our everyday perceptive experience. Let us imagine an experimental situation where a consciousness investigator asks me to sit alone in a totally dark room and to open my eyes when a beep is emitted as light is established. Let us further imagine that when I do open my eyes, I discover the spectacle of a pool of blue water lying at my feet. What the term ‘phenomenal consciousness’ designates in such a case is nothing other than the very fact of apprehending the visual appearance or phenomenon of a pool of blue water, as well as of knowing that I do apprehend it. What philosopher T. Nagel referred to as ‘what it is like’ to perceive the pool of water. On the basis of this example, the reputedly difficult notion of phenomenal consciousness can be analyzed schematically along the following lines: 1. Phenomenal consciousness is the consciousness of phenomena. The specificity of phenomenal consciousness first and foremost derives from its content, from what it is a consciousness of, as opposed to the fact of being conscious itself. This does not preclude that this fact might also offer specific features, different from the ones it exhibits, when other types of content are involved. Ned Block, for instance, famously advocates the view that phenomenal consciousness is a specific mode of being conscious, to be primarily distinguished from another one, labeled ‘access consciousness’. The difference between them is seen by Block as a difference between experiencing and accessing something. Neither his analysis nor his terminology will nevertheless be followed here. 2. The consciousness of phenomena corresponds to a specific kind of experience of phenomena. Indeed,
Subjectivity; First- and Third-Person Methodologies
the word ‘experience’ seems more appropriate for designating the fundamental relation that a subject entertains with phenomena, whether this relation is conscious or not. Bertrand Russell proposed for instance in his early epistemology that a visual experience such as the one involved in the above example is a two-term relation called acquaintance between a subject and an appearance of a specific kind that he also conceived of as a sense datum. He moreover suggested that experience should be considered as a cognitive kind of relation. As a matter of fact, when I open my eyes in the laboratory room I suddenly learn something (right or wrong) about the phenomenal features of my physical environment and I become able to answer such questions from the experimenter as: what are you surrounded by? According to this perspective, experience has therefore no essential link with consciousness and it divides into a conscious and a nonconscious types. Phenomenal consciousness is therefore to be analyzed as an experience of phenomena of the conscious kind. Although the prevailing opinion is that experiential or qualitative mental states cannot be nonconscious, such a view is defended by some influential consciousness theorists such as David Rosenthal. It is arguably supported by a series of neuropsychological discoveries that have firmly established the existence of a ‘blind’ stimulation in many sensory modalities. A possible reading of the spectacular phenomenon of blindsight, for instance, where it is demonstrated that the visual system of a partially blind subject actually processes some inputs located in the blind part of his visual field, is to say that the subject visually experiences something in that part but is totally unaware of it, as opposed to saying that he is informed of something but deprived of any visual experience. 3. Consciousness is an immediate reflective knowledge. If one accepts the distinction between conscious and unconscious experiences, the fundamental difference between them is clearly that a conscious experience, as correctly indicated by etymology, is one that its subject knows about. When I have the conscious visual experience of a pool of water, I apprehend a determinate visual phenomena and I know that I do. The whole
391
difficulty is to specify the nature of the knowledge that consciousness consists in. A traditional view is that it is an immediate as well as a reflective (and possibly higher order) one. 4. A Phenomenon is subjective inasmuch as it is a being for. Through consciousness so conceived, what I know about what I visually experience when I open my eyes in the laboratory room and see a pool of water can be characterized as a certain state of affairs including a liquid element, endowed with a certain degree of transparency, and a certain hue of blueness. However, such a characterization can be understood in two fundamentally different ways: as what the state of affairs is in itself, independently of my conscious experience of it, or as what the state of affairs is as a result of this conscious experience: in other words, as what it is for the experiencing and conscious subject. This traditional philosophical distinction between a being in itself and a ‘being for’ is the core of the opposition between the objective and the subjective. Indeed, being subjective can be fundamentally defined as being what there is for a subject, or from the point of view of a subject. The idea expressed by the word ‘for’ is particularly delicate to unfold. It includes at a minimum the notion of relativity to a subject: what is subjective is what an entity is relatively to a subject, as opposed to what it is independently of any relation with a subject. But it is in addition, relative to a subject endowed with a capacity for knowledge in the most general sense of this term. The subjective can thus take various specific forms depending on the kind of knowledge capacity involved, such as experience. Accordingly, the subjective content of a conscious experience of a subject S is what an entity is relative to this subject S and specifically to both his experience and his consciousness of it, considered as two forms of immediate knowledge. For instance, the subjective content of my conscious visual experience of the pool of water is what the pool of water is relative to me, as well as to my visual experiencing and to my consciousness of this visual experiencing. In such a perspective, one should thus distinguish between the subjective content of an experience and the subjective content of a conscious experience, that is to say between what the pool of water is for me in virtue
392
Subjectivity; First- and Third-Person Methodologies
of experiencing it visually and what it is for me in virtue of being conscious of what it is for me through visual experience. The distinction is important because these two aspects of subjective content do not necessarily coincide. I might for instance be conscious of less than what there is in the content of my experience, or modify the content of my experience by becoming conscious of it. This possible discrepancy should not be confused with the one that might exist between the subjective – what there is for a subject – and the objective – what there is in itself. It is a discrepancy within the subjective. The temptation to consider that what is subjective is not only what there is for a subject, but also what is contained in a subject should be resisted for a number of reasons. The question of the ontological status of phenomena is one of the most difficult in philosophy. If a color is objectively speaking a wavelength, where else than in the mind should the phenomenal property such as blueness be located? It has been argued however, for instance by Husserl, that the phenomenal property of blueness should not be confused with the sensation of blue, and that only the second can be said to be in the mind, although the first one is not part of physical reality either. Moreover, even if phenomenal properties are ontologically internal elements of the mind, there is a sense of being in the mind that does not apply to all of them. Indeed, as emphasized by Brentano, the phenomenon of blueness refers to something external to the mind, while the phenomenal property of feeling sad refers to something internal to it, even though in his opinion both of them, as phenomenal properties, are ontologically speaking elements of the mind. Finally, whatever is in the mind is not necessarily a being for. As a matter of fact, a leading assumption of contemporary cognitive science is that the essential properties of mental processes are not accessible through consciousness, and not even experienced unconsciously: most of the mental properties that its hypotheses postulate are about what the mind is in itself and not as it manifests itself to the subject, even though they can reasonably be located in the mind (at least in part, since this assumption is challenged by the externalist approach to cognition).
Consequently, the essential meaning of the subjective is that of a property for, or as apprehended by, a subject, especially through experience, as opposed to a property independent of any such apprehension, as it is in itself. And such a subjective property so understood can be considered as external to the subject, in the sense of not being ontologically in it, but also at times in the sense of referring to something outside of it. It is in this fundamental sense of subjectivity that the content of experience, and of conscious experience in particular, is subjective.
The Possibility of a Science of Subjective Properties And it is clearly as such that its reintroduction into the domain of cognitive science elicited a new wave of epistemological worries. The most important of these worries is that a scientific knowledge of properties for a subject is a contradiction in terms. The general argument supporting it can be formulated in the following way. Science is by definition a search for objective properties: what the chemist is for example interested in is not what the pool of water looks like to him, but what it is in itself. Subjective properties, however, lack by definition objective determinations. And even if they don’t, because a subjective property is not necessarily a purely subjective one, in capturing those objective determinations a scientific investigation would not capture the subjective as such, and thus would not properly speaking be a science of subjectivity. It can however be retorted that this argument is sophistic because it ignores the relativity of the subjective/objective distinction, and that there is a perfectly legitimate sense in which one can claim to acquire an objective knowledge of the subjective as such. The consciousness theorist who puts me in the laboratory room wants to study how the pool of water looks to me when the light goes on, and not the triggered information processing between my retina and the V1 area of my brain, of which I remain perfectly unaware. However, his interest is in how things really look to me, and not in how it looks to him that things are looking to
Subjectivity; First- and Third-Person Methodologies
me. In this sense, it is unquestionably legitimate to claim that he wants to capture my subjectivity as it is in itself and not as it is for him. Consequently, the subjective must be acknowledged an objective dimension qua subjective; and this objective dimension is what a being for a subject is in itself, as opposed to what it is for a third party. There is a second important, although less fundamental, sense in which one can legitimately claim to investigate an objective dimension of the subjective as such. The knowledge that the scientist seeks to obtain when experimenting on my visual consciousness is a knowledge that goes beyond my singular case and applies to human subjectivity in general. It is an objective one in the sense of holding of more than one subject, and is therefore based on the assumption that some features of my subjectivity are shared by all other human subjectivities. Disputing the validity of this assumption supposes that subjective properties present a unique degree of singularity in the world, so that there are no significant commonalities between my phenomenality of the water in the pool and that of the rest of human kind. Although often asserted in a melodramatic tone of voice, this suggestion is however seldom supported by real arguments. Moreover, it is rather counterintuitive, since conscious experience seems to be on the contrary astonishingly similar in content across physiological and cultural conditions. It is therefore far from obvious that the degree of singularity that, like any other scientific knowledge, a science of subjective properties must ignore is significantly higher than what it is in the study of standard objective properties. In view of these counterobjections speaking in favor of the theoretical possibility of a science of the subjective, a second less radical argument can nevertheless be formulated against its actual feasibility: the idea is that we do not, as a matter of fact, have the means to reach the objective dimension of the subjective as subjective. The consciousness scientist does not have for instance the actual means to grasp how the pool of water really looks like to me and what makes it a typical case of human visual apprehension of a pool of water. This second kind of skepticism is usually nourished by considerations about the lack of a reliable access to subjective properties. Scientific knowledge needs data at
393
its basis, and no adequate source of observation is supposed to provide such data when it comes to investigating the subjective. The difficulty is vividly illustrated by the case of animal subjectivity: assuming that a bat has an experience of the world around it, if not a conscious one, how would the pool of water look to it? Quite different from what it looks to me, given that the bat has no equivalent of my visual system. However, the only way to penetrate this specific phenomenality would be to have the capacity to adopt the point of view of a bat, that is to say to be, or at least to simulate being, a bat. But human scientists do not have such capacity. A natural reply is that this argument at best demonstrates that a science of the subjective cannot be universal and requires a certain degree of homogeneity between the subjectivity of the knowing subject and that of the investigated subject. However, this counterobjection can clearly be circumvented by combining the argument with the previous idea that conscious experience is fundamentally singular. The claim thus becomes that the proximity between the investigator and the investigated will never be sufficient because subjectivity is too essentially individual. There is however one obvious way out of this last reformulation of the second skeptical argument. For if it seems reasonable to say that we cannot ask the bat to communicate what the world looks like to it, it seems unreasonable to say that we cannot ask a human being to do so. As a matter of fact we do every day, and for the longest time, self-characterization of conscious experience was the very foundation of psychology defined as the science of internal experience. The sceptic must therefore add an argument to the effect that the self-characterization of experience, even when possible, is not a reliable source of knowledge about it. And he does, but usually capitalizing on a long history of suspicion towards what has traditionally been called introspection, rather than by contributing new reasons to share this suspicion. Even if such skepticism is in fact well grounded, it is nevertheless still necessary, in order to rule out the feasibility of a science of the subjective, to show that subjective properties cannot be treated as merely theoretical ones, that is to say as properties hypothetically inferred on the basis of
394
Subjectivity; First- and Third-Person Methodologies
objective data, such as behavioral or neurobiological ones. As a matter of fact, this has been the standard treatment of those mental properties that are considered as objective ones, like propositional attitudes, since the return of the mental brought by the demise of behaviorism. Why could not this acceptable treatment for the objective dimension of the mind be extended to its subjective dimension? If indeed the scientist can hypothesize on purely objective grounds the nature of the information processing going on in my head when I see the pool of water, what reasons are there to think that he cannot proceed similarly when it comes to subjective properties like what the pool of water looks like to me? A third argument of importance to be found in the literature against the possibility of a science of the subjective is the well known explanatory gap argument. It also incarnates a weaker form of skepticism than the first argument, since it only denounces the possibility of obtaining a scientific knowledge of subjective properties of an explanatory kind, and moreover a naturalist kind of explanation, in the sense of an explanation by natural properties. The explanatory gap argument comes under a variety of forms, the most popular being due to Thomas Nagel, David Armstrong, Ned Block, Saul Kripke, and Jospeh Levine. All of them are subjectivity based arguments in the sense that they all consider that subjective properties offer specific resistance to a naturalist explanation. Some of them, however, are more directly based than others on what makes these properties subjective ones. The best known in this respect was offered by Nagel. It purports to show that the standard procedure of reduction used to establish that water is H2O, or that lightening is an electrical discharge, cannot be used to show that phenomenal properties are themselves nothing else, for example, than certain neural discharges. According to Nagel, reduction is indeed a process of substituting more objective properties for subjective ones. Reducing water to H2O means to show that what is characterized as a transparent, bluish liquid body can also be characterized as a mass of H2O molecules, thereby also showing that the first characterization is a subject dependent one. However, in Nagel’s opinion, such a strategy presupposes a duality of subjective and objective determinations that is absent in the case of subjectivity.
The property that has the subjective determination of looking transparent to me cannot, for instance, be the same that has the determination of being a firing cortical area, because if it has that determination, it is not a subjective property. In other words, mental subjectivity is purely subjective and therefore cannot meet the requirements of standard scientific reduction understood as a move towards objectivity. This version of the explanatory gap argument is clearly controversial, as are in fact all the others. A detailed discussion of them is beyond the scope of this general presentation of the issues raised by the subjectivity of consciousness, but the problem they pose certainly is the most important one that a science of subjectivity must face today, and it is for that reason the most debated one.
The Need for First-Person Knowledge It is primarily in the context of this debate that the question of first-person methods has in fact resurfaced. The line between first- and third-person methods is more difficult to draw than it looks. But it is intuitively clear: a study of my visual consciousness of the pool of water calls upon a first-person methodology to the extent that it grants a role to my own characterization of the content of my visual consciousness, as opposed to a study that only relies on data accessible to the investigator. In short, a first-person investigation can be defined in first approximation as an investigation of conscious subjective properties carried out by the subject of these properties. The move from the explanatory gap problem to the recourse to first-person methods is easily understandable. A theory of consciousness is a theory of subjective properties to the extent that it is a theory of the content of consciousness. It follows from the definition of a subjective property that only the subject of such a property can have it as well as the immediate knowledge that makes it a datum. A theory of consciousness therefore needs a first-person characterization of the content of consciousness to the extent that it needs data about it. Now, a natural reaction to the explanatory gap problem is to suppose that conscious phenomenal properties resist explanation in naturalist terms
Subjectivity; First- and Third-Person Methodologies
because they are insufficiently or incorrectly characterized, and that they are because they are specified in an indirect way due to the methodological constraints that contemporary cognitive science inherited from behaviorism. In other words, the supposition backing up the current calls in favor of the introduction of first-person methods is that consciousness must be rehabilitated as a source of information about its own content in order to overcome the explanatory gap problem, and that methods for capturing its testimony in a reliable way can be devised. It is important to underline that the use of firstperson methods can also be, and has been, motivated independently of any specific concern for a theory of consciousness and the explanatory gap problem that it allegedly faces. From this different perspective, the characterization by the subject of a mental process of what he knows about this process through consciousness is a useful source of information about what this process is in itself, and not only about what it looks like to its subject. In other words, about the objective properties of the mental, and not only about its subjective properties. As a matter of fact, a long tradition in the philosophy of psychology went as far as claiming that consciousness is immune to error, and that the mental manifests itself in consciousness as it really is. But even from a less confident perspective in the epistemological virtues of consciousness, it is reasonable to think that what we know about mental processes through consciousness is a useful source of information about them, unless it can be established that their conscious manifestations are so distorting that nothing can be safely inferred from these manifestations about their true nature. It is to this line of reasoning that most contemporary attempts to revise the exclusion of first-person knowledge belong that took place prior to its re-examination in the specific context of the problem of consciousness. An essential characteristic of the rehabilitations of first-person methods undertaken in this latter context is that they give to first-person knowledge a purely descriptive role: what the subject is asked to provide is a detailed and precise characterization of his content of consciousness taking great care not to introduce determinations that do not belong to it. There is in fact no intrinsic connection between the concept of first-person knowledge and that of
395
descriptive knowledge: a first-person knowledge can in principle be explanatory and a third-person one descriptive, in the general sense in which a description is a characterization of what something is. In the case of the content of consciousness, however what differentiates a first-person from a third-person description is that the first-person one is based on an immediate knowledge of what is to be described. A first-person characterization is therefore also descriptive in the sense of being somehow intuition based, like a perception judgment. In the effort to remain faithful to what is apprehended by consciousness, the difficulty is firstly to avoid misapprehending the scope of consciousness and to avoid presenting as elements of the content of the consciousness that accompany a certain mental event, elements that only come to consciousness at a later stage. Change blindness experiments show for example that we do not notice significant elements of our visual scene, as important as a gorilla walking through a group of human ball players. However, when attention is called to them, we suddenly discover the presence of the missed element, and then become conscious of it. In such a case, however, we have arguably extended the scope of our visual consciousness, and a faithful description of what we are initially visually conscious of should not mention the gorilla. The second difficulty is to avoid confusing explanatory determinations, in the sense of determinations that can account for what we are conscious of but are not actual elements of it, with descriptive ones. An instance is when we present as an element of the content of our perception of a movement, the source of that movement, that we have in fact not perceived, because it was hidden to us, and that was only later inferred from what we really perceived. If self-descriptions must therefore avoid the danger of incorporating extraneous elements, the basic idea of first-person methods is nevertheless that a process of clarification is needed in order to establish what we are conscious of. In that sense, a distinction must be introduced between what we are immediately conscious of and what we become conscious of through the process of description. But this difference remains a purely qualitative one: we become more conscious of what we are conscious of, but not conscious of more.
396
Subjectivity; First- and Third-Person Methodologies
The preliminary definition of first-person knowledge offered above was suggested by comparing the case when I characterize the content of what I am aware of when I see the pool of water and the case when the experimenter himself characterizes it. Indeed, the only difference between the two characterizations that might account for their respective qualifications as a first- and a third-person knowledge lies in the identity of the subject who performs the characterization. Is this definition sufficiently general however to capture the essence of the distinction? It is tempting to define first-person knowledge in a more general way, as the knowledge of any kind of property that can be obtained from a unique viewpoint, and accordingly by a single subject, and third-person knowledge as a knowledge accessible from any viewpoint or to any subject. First-person knowledge of subjective properties would thus only represent a specific case of first-person knowledge so defined. Such a definition is nevertheless clearly too broad, since the knowledge that can be obtained from the point of view of God should then qualify as first-person knowledge, although intuitively it does not. First-person knowledge must consequently be a knowledge of subjective properties. A definition in terms of unique viewpoint is acceptable only under the condition that this unique viewpoint be specified as that of the subject for whom the properties under investigation are properties. But such a formulation does include in fact a reference to subjective properties. It is clear, in particular, that it would not be enough to define first-person knowledge as a knowledge obtained from the point of view of the subject of the properties under study. For an investigation of the anatomical properties of my toes by myself does not intuitively qualify as an example of first-person knowledge. The above definition nevertheless needs at least one refinement on pain of making any observational knowledge a first-person one, and transforming any scientific knowledge into a mixture of first- and third-person ones. When the chemist looks at a sample of water under his microscope, he is having a conscious visual experience and his task in part is to characterize what he sees. His observation based knowledge cannot however qualify as a first-person descriptive knowledge. It
can only to the extent that the chemist’s characterization of what he sees is focused on the subjective dimension of his subjective properties, in the spirit of protocol statements of early logical positivism. However, ordinary observation statements in science are meant, on the contrary, to extract objective determinations, usually of a quantitative kind such as metric ones. First-person knowledge seems therefore to be better defined at its most general level as any knowledge of subjective properties as such by their own subject. And a first-person methodology is any procedure resulting in such knowledge. It is a descriptive first-person methodology when the knowledge obtained is descriptive. An interesting result of this definition is to reveal that first-person descriptive knowledge is in fact present in standard procedures of psychology and cognitive neuroscience usually seen as merely third-person ones. Verbal reports, to mention a central case, are verbally couched descriptions by a conscious subject of his content of consciousness as such. This is the reason why psychologists Kenneth Ericsson and Herbert Simon accused contemporary cognitive science of being schizophrenic about introspection, violating in its effective procedures its pronouncements of methodological continuity with behaviorism as regards the exclusion of first-person knowledge. Even pressing a button to express whether I saw a pool of water or a sand dune, is a way of describing my content of consciousness, and therefore a method for obtaining first-person descriptive knowledge. Button pressing is in fact an interesting case, showing how thin the line is between the two kinds of knowledge and how intricate they are in actual research. For button pressing can also be understood as a purely objective method, providing behavioral data from which an hypothesis can be inferred about the conscious subjective properties of the subject under investigation, instead of an expression of the description, carried out by the subject, of these properties. As a third-person method, button pressing is treated by the investigator as an indication of the content of consciousness of the subject; as a first-person method, it is treated as a testimony of the subject about the content of his consciousness. Accordingly, any procedure that counts as a testimony-collecting one is a firstperson descriptive method, because its goal is to
Subjectivity; First- and Third-Person Methodologies
obtain a description by the investigated subject of his conscious subjective properties as such. If the standard methodology of cognitive science is already a mixed bag of first- and thirdperson methods, the return of the first person cannot be understood as the abandonment of a purely objective approach to the study of subjectivity, where subjective properties are simply inferred from objective indicators such as behavioral or neurobiological manifestations by a third party investigator, who gives no role to what their subject can express about them. It must rather be seen as a call for the development of a more lucid, systematic and refined kind of first-person knowledge. And for the development of new means of self descriptions in particular, since there is no reason in principle to limit the descriptive tools of firstperson knowledge to ordinary language: new logical and mathematical formalisms have for instance been suggestively employed to that effect.
The Reappearance of Scepticism about First-Person Knowledge With the return of the first-person perspective, the reasons that motivated its rejection since the early days of nineteenth century positivism also made a come back. All of them more or less concur in the idea that self-characterization of the content of consciousness is structurally unreliable, and therefore unable to qualify as a scientific form of knowledge. This first-person knowledge skepticism can even be considered as still prevailing, and it is obviously crucial that any attempt to rehabilitate first-person knowledge answers the arguments, old and new, supposed to speak in its favor. Leaving details of argumentation aside, one can distinguish four main types of reasons for claiming that firstperson knowledge is unreliable, that lead to four types of skepticism with a different threatening power on the project of reintegrating such knowledge into contemporary cognitive science. The analysis of the content of an experience like my visual experience of the pool of water must first be further refined, by distinguishing what is manifested to me as the result of the stimulation of my sensory organs and the interpretation given to that manifestation. This distinction corresponds
397
to the traditional one between sensation and perception, but it can be theoretically elaborated in a variety of ways. In particular, several levels of interpretation can be set apart, from a basic one internal to the visual process itself, to higher degree ones resulting from the application of concepts and beliefs to sensory output. The need for such a preliminary distinction is evidenced by such cases as illusions, where what there is for a subject does not correspond to what should normally be there for him given the nature of its sensory stimulation, or also the case of bistable percepts, where a constant stimulus gives rise to two alternate perceptual scenes, like in the famous duck/rabbit example, where the eye switches back and forth between the apprehension of one single drawing as representing a rabbit and as representing a duck. The distinction can be generalized to any form of being for and fixed terminologically as the distinction between what there is for a subject and what there is for a subject as interpreted by that subject. A first form of skepticism can now be formulated as the thesis that consciousness is structurally impotent to reveal what the mental is in itself. In the case of visual conscious experience, for instance, it means firstly that, even at the mental level, what the process of perceiving really is, is so remote from what it is from the point of view of consciousness that it renders the accurate description of the content of its conscious apprehension a futile exercise scientifically. And similarly for the content itself of visual experience: what I know through consciousness about how the water looks visually to me is supposed to be so different from what it really looks visually to me that there is no way to get the truth about the latter from an investigation of the former. It should be underlined that this form of first-person skepticism runs against a venerable tradition of belief in the infallibility of consciousness considered as an immediate knowledge of the mental. According to this traditional view, if we have good reasons to doubt that the pool of water is in itself as it looks visually and consciously to me, we have no good reasons to doubt that we really perceive it as consciousness says we do. A standard argument to that effect is that the immediate knowledge of the mental provided by consciousness is free of the inconsistencies that have traditionally motivated the denial
398
Subjectivity; First- and Third-Person Methodologies
of the epistemological value of the immediate knowledge of the external world. More recently a different kind of argument was developed in favor of the truthfulness of the specific consciousness of agency. According to this influential argument derived from Wittgenstein, if I consciously apprehend an action as being authored by me, it must be mine, because there is no process of identification of myself involved in this apprehension and only the presence of such a process would open the possibility of a mistake in the testimony of consciousness. Given this entrenched belief in the infallibility of consciousness, this first type of skepticism thus assumes a challenger position and must show not only that blind confidence in the deliveries of consciousness is ungrounded, but that these deliveries call in fact for particular suspicion. A second form of first-person skepticism is more specifically directed at the content of experience and capitalizes on the preliminary distinction between what is manifested to a subject and the way the subject interprets that manifestation. The criticism is that consciousness can only give us access to what there is for us as we interpret it, and is thus unable to detect systematic discrepancies that as a matter of fact do occur between the manifestations and their interpretations. First-person investigation should therefore be rejected as too limited as well as gullible. Two spectacular phenomena lend support to this type of rejection: perceptual filling in and the already mentioned one of change blindness. It is now well established that a central aspect of the visual scene is in fact not supported by any sensory visual input, because the macula is deprived of photoreceptors: the visual system nevertheless somehow fills the void and the content of visual experience is free of any blind spot. In a way we therefore see less than what consciousness tells us we see. In other words still, there is less in our visual field than what there should be according to the testimony of consciousness about it. Change blindness, being a phenomenon of ignorance of what is visually manifested to us, can be taken on the contrary to show that there is at times more in what is manifested visually to us than in the interpretation we give to it, and to which consciousness gives us access. Finally, given that an illusion is an unwarranted interpretation of visual stimuli, our incapacity to detect
many perceptual illusions can be used to make the same point about the limitation and candidness of consciousness regarding the content of perceptual experience. These first two forms of skepticism raise a difficulty only for a return to first-person knowledge conceived as a means to reach the objective nature of mental properties. For what they claim is that we cannot trust the value of the testimony of consciousness in this regard, either because it is a deeply distorting one or because it is overtly limited and consequently naive. However, they are innocuous if first-person knowledge is specifically put to the service of a theory of the content of consciousness. For what such a theory is after is not what the mental is in itself, but what we know about it through consciousness, what it is for a conscious subject. And from this point of view, the adequacy of the view of the mental delivered by consciousness to its real nature is simply irrelevant. The fact that the testimony of consciousness might indeed be drastically distorting leaves its possibility and interest intact, because the object of the inquiry is precisely in the distortions themselves. Similarly, the fact that consciousness seems to ignore the distinction that should be made between what is manifested to us in experience and the interpretation we give to it is irrelevant to whoever wants to capture how experiential content is apprehended consciously. The opposite view arguably rests on a controversial concept of the content of experience that assimilates it to something immediately apprehended only in the sense of something given or intuited, or again passively received, thereby limiting it to what is manifested, according to the above terminology. However, the immediacy of experience must clearly be – and in fact has traditionally been – understood in a broader and more temporal way, as how things are firstly apprehended by a subject: for instance how the pool of water is apprehended in the operation of perception before any additional cognitive operation is performed on it. This understanding of the immediacy of experience is perfectly compatible with the idea that experience is not only a process of passive reception, but also of construction or projective interpretation. Accordingly, a description of what is experienced should not be concerned with the fact that it goes
Subjectivity; First- and Third-Person Methodologies
beyond what is manifested: in the case of change blindness previously mentioned, for instance, faithfully capturing its characteristics should not mention any gorilla since this is how things are for us immediately in the appropriate sense of immediacy. And to the extent that such characteristics are retrievable through consciousness and therefore amenable to a first-person description, this description cannot be charged with undue limitation or naivety. Consequently, first-person skepticism can only threaten a first-person knowledge elaborated in the context of a theory of consciousness if it shows that what a subject takes his content of consciousness to be is not what it really is as content of consciousness; for instance, that what I claim to be conscious of when I visually experience a pool of water is not what I am really conscious of. In other words, firstperson skepticism can only undermine the use of first-person knowledge for the purpose of a theory of consciousness if it can demonstrate that it fails to capture the very content of consciousness itself. And this is a more challenging task than demonstrating that it fails to capture the true nature of what this content refers to. Indeed, reasons to doubt about the first thing are harder to come by than reasons to doubt the second one. It might be that the mechanisms of consciousness distort the content of visual experience, that what we are conscious of experiencing visually does not correspond to what we do experience visually. But is there any reason at all to doubt that when we claim to be conscious of a certain visual appearance, like something appearing as a pool of water, we are mistaken about what we are conscious of itself ? There are in fact two main ways to claim that we might be. One consists in locating the difficulty in consciousness itself and arguing that consciousness understood as an immediate knowledge is an illusion: what we claim to be an immediate sort of access to the mental is in fact an elaborate theoretical construction. Such is for instance the claim of the new anti-introspectionist view of Roger Nisbett and Timothy Wilson, who suggest that our conscious access to the mental is much more limited than we think, and that we in fact often theorize about what we believe to apprehend in an immediate fashion. Daniel Dennett’s well known rejection of first-person knowledge as a
399
fantasy is also essentially based on these sorts of considerations. Alternatively, the blame can be put on the operation of self-characterization, contending that, although consciousness as a form of immediate knowledge does not suffer from any structural weakness, we are structurally unable to characterize what it delivers in a reliably descriptive way, to the point of making the idea of a first-person descriptive knowledge worthless. As a matter of fact, the oldest criticisms of first-person knowledge belong to this category. One of them going back to Auguste Comte and John Stuart Mill is that observation of the content of consciousness is impossible, or that what we can at best observe is a mental event retained in, and modified by, memory. Another type of argument that can be related to Ludwig Wittgenstein’s private language argument or Eugene Fink’s criticism of Husserlian phenomenology is that a linguistic instrument apt to the needed descriptive task is not available and cannot be devised.
The Testimony of Consciousness In order to explain the functionalist view of mental properties, the philosopher David Lewis compared a mentalist hypothesis to a detective story in which each one of a number of suspected protagonists is solely characterized in terms of the role he played in a murder, leaving his identity unknown. Similarly, one could say that the consciousness theorist is like a detective interested in what the witnesses of a murder saw. What he is after is not an explanation of the process that resulted in the dead body, but an explanation of how it looked to these witnesses. And in order to elaborate such an explanation, he must firstly establish how it did look to them. If he infers their subjective view of the murder process from such clues as their position on the murder scene, the movements they might have made, etc., he builds his hypothesis on purely objective methods. But the moment he collects their testimonies and treats them as real witnesses, he gives room to first-person knowledge in his investigation. These testimonies can take the form of verbal descriptions, but also of drawings or imitations. And this is a reasonable move indeed, unless the
400
Subjectivity; First- and Third-Person Methodologies
investigator is so nai¨ve as to think that what they saw necessarily corresponds to what happened. The only real problem for him lies in the reliability of these testimonies. Except for the fact that they might be intentionally insincere, the witnesses might be too vague, unable to express themselves with sufficient care, or reporting things they thought they saw but did not actually see. But unless he thinks he has good reasons to believe that these possible defects are systematic and insuperable by appropriate methods for collecting testimonies, he has none to renounce this approach. The key issue in the debate about first-person knowledge in cognitive science is whether there are indeed such reasons. See also: Phenomenology of Consciousness; Philosophical Accounts of Self-Awareness and Introspection; Self: Body Awareness and Self-Awareness; Self: Personal Identity; Self: The Unity of Self, Self-Consistency.
Suggested Readings Armstrong D (1963) Is introspective knowledge incorrigible? Philosophical Review LXXII(4). Chisholm RM (1981) The First Person: An Essay on Reference and Intentionality. Brighton, Sussex: Harvester Press.
Dennett D (1991) Consciousness Explained. Boston: Little Brown. Ericsson KA and Simon H (1980) Verbal reports as data. Psychological Review 87(3). Jack A and Roepstorff A (2004) Trusting the subject: Part 1. Journal of Consciousness Studies 11(6–8). Lutz A (ed.) (2004) Special Issue: Naturalizing Phenomenology. Phenomenology and the Cognitive Sciences. . Lyons W (1986) The Disappearance of Introspection. Cambridge, MA: MIT Press. Nisbett R and Wilson T (1977) Telling more than we can know: Verbal reports on mental processes. Psychological Review 84(3). Petitot J, Varela F, Pachoud B, and Roy J-M (eds.) (1999) Naturalizing Phenomenology: Issues in Contemporary Phenomenology and Cognitive Science. Stanford: Stanford University Press. Shoemaker S (1996) The First-Person Perspective: And Other Essays. Cambridge, England; New York: Cambridge University Press. Van Gulick R (1993) Understanding the phenomenal mind: Are we just Armadillos? In: Davies M and Humphreys G (eds.) Consciousness. Oxford: Blackwell. Varela F and Shear J (eds.) (1999) The View from Within: First-Person Approaches to the Study of Consciousness. Paris: Seuil. Velmans M (ed.) (2000) Investigating Phenomenal Consciousness. Amsterdam: John Benjamins. Wright C (1998) Self-Knowledge: The Wittgensteinian Legacy. In: Wright BSC and Macdonald C (eds.) Knowing Our Own Minds. Oxford: Oxford University Press. Zahavi D (ed.) (2004) Special Issue: The Return of Subjectivity Phenomenology and the Cognitive Sciences 3(3).
Biographical Sketch
Jean-Michel Roy is teaching philosophy of cognitive science at the Ecole Normale Supe´rieure Lettres et Sciences Humaines of the French University of Lyon. He is the author of several studies on the epistemological issues raised by the notion of phenomenal consciousness and a coeditor of the 1999 book Naturalizing Phenomenology: Issues in Contemporary Phenomenology and Cognitive Science, Stanford University Press.
Theory of Mind (Neural Basis) R Saxe, Massachusetts Institute of Technology, Cambridge, MA, USA ã 2009 Elsevier Inc. All rights reserved.
Introduction The externally observable components of human actions carry only a tiny fraction of the information that matters. Human observers are vastly more interested in perceiving or inferring the mental states – the beliefs, desires and intentions – that lie behind the observable shell. If a person checks her watch, is she uncertain about the time, late for an appointment, or bored with the conversation? If a person shoots his friend on a hunting trip, did he intend revenge or just mistake his friend for a partridge? The mechanism people use to infer and reason about another person’s states of mind is called a ‘Theory of Mind’ (ToM). One of the most striking discoveries of recent human cognitive neuroscience is that there is a group of brain regions in human cortex that selectively and specifically underlie this mechanism. The importance of mental states for behavior prediction is especially clear when the person doing the behavior is misinformed: that is, when the actor has a false belief. False beliefs therefore figure heavily in the study of ToM. The original ‘False Belief task’ was developed for use with preschool aged children (Figure 1). In the basic design, a child watches while a puppet places an object in location A. The puppet leaves the scene and the object is transferred to location B. The puppet returns and the child is asked to predict where the puppet will look for the object. Threeyear-olds say the puppet will look in location B, where the object actually is; older children say the puppet will look in location A, where the puppet last saw the object. Following the tradition in developmental psychology, many of the early neuroimaging investigations of ToM required subjects to attribute false beliefs to people in stories or cartoons. Most of these studies have used functional magnetic resonance imaging (fMRI) to measure changes in cortical blood oxygenation. Across these studies,
a very reliable set of brain regions was implicated in the ‘false belief ’ condition of each study, including right and left temporoparietal junction (TPJ), medial parietal cortex (including posterior cingulate and precuneus), and medial prefrontal cortex (MPFC). The same brain regions have also been identified by converging methods, including both lesion and transcranial magnetic stimulation (TMS) studies of ToM. This group of brain regions is sometimes collectively called the ‘ToM network.’ A critical question that emerged from this work is whether the ToM network is a distinct and domain-specific neural system for thinking about the mind. The alternative is that some or all of the regions in the ToM network are actually recruited for some other aspect of solving a false belief task. There is more to solving a false belief task than a concept of belief, and there is more to a concept of belief than passing the false belief task. In particular, attributing a false belief to another person depends heavily on two cognitive capacities that are not specific to ToM: language and inhibitory control. Plausibly, activation in the ToM network could reflect a combination of language processing and inhibitory control. This article thus first reviews evidence for the relationship between ToM, language, and inhibitory control in the brain, and then turns to more detailed investigations of the neural basis of ToM.
Theory of Mind and Language: fMRI and Lesion Studies Mental states cannot be observed directly; beliefs and desires are invisible, abstract entities. Nor is there any simple correlation between mental states and behavior. For example, there is no observable behavior that is reliably diagnostic of whether a person currently believes that today is Tuesday. One invaluable way to learn about the elusive
401
402
Theory of Mind (Neural Basis)
Figure 1 Brain regions showing higher metabolism during false belief than false photograph stories. Clockwise from bottom left: axial, coronal and saggital slices, and a rendering of the lateral surface. Clearly visible are left and right TPJ, and the medial parietal (precuneus) regions.
contents of the mind is therefore to listen to how other people talk about the mind. During development, linguistic ability is correlated with false belief task performance. In a striking example, deaf children of hearing parents (that is, whose parents are nonnative signers) are selectively delayed in passing the false belief task. These children have similar difficulty even on non-verbal tests of false belief understanding, suggesting that the delay does not reflect the language demands of the tasks themselves. Moreover, even after accounting for the child’s own skill with sign language, the child’s performance on the false belief task is independently predicted by the mother’s proficiency with sign language, and specifically her use of mental state signs. Deaf children of deaf parents (native signers), by contrast, are not delayed. Clearly, linguistic exposure influences ToM development, but the mechanism underlying this influence remains controversial.
Given these results, it is possible that the brain regions implicated in ToM are instead involved in some aspect of language processing. One way to test this hypothesis is to present stimuli that imply the protagonists’ false belief, but do not include any explicit statements of beliefs (and therefore do not include sentence complements). One study used verbal stories that merely described a sequence of actions by a character. No mental states were explicitly mentioned, but the sequence of actions could most readily be explained in light of the protagonist’s belief. The control stories were sequences of mechanical or physical events that similarly required subjects to infer an unstated physical causal process. As predicted, all of the regions in the ‘ToM network’ – the right and left TPJ, medial parietal and MPFC – showed a stronger response during the implied-belief stories than during the implied-physical stories. Other studies have used nonverbal single-frame or animated
Theory of Mind (Neural Basis)
cartoons that required subjects to attribute beliefs to the characters, but used no language at all. In all cases, the same brain regions were recruited when subjects attributed beliefs to the characters, independent of modality. No brain regions in the ‘ToM network’ appear to be associated with the linguistic demands of the task, at least in adults. These results raise the possibility that while language is necessary for ToM during development, once concepts of beliefs and desires are formulated, a mature ToM could function even in the absence of language. The critical test is therefore to investigate the consequences of late acquired aphasia (especially loss of grammatical skills) on an already mature ToM. In a series of studies, Rosemary Varley and Michael Siegal found that aphasic patients could be relatively spared in ToM. Ian Apperly and colleagues provided strong new evidence for the same hypothesis. PH, a young man who had a left hemisphere stroke, was tested on a battery of language and ToM tests. Although PH was severely impaired on tests of syntax, including specifically the syntax of embedded clauses, he showed no impairments on non-verbal tests of ToM, including first- and second-order (what X thinks that Y thinks) false belief tasks. Taken together, these studies provide clear evidence that adults with a mature ToM can formulate thoughts about other people’s thoughts, even in the face of severe grammatical impairments, and thus by adulthood, the neural bases of ToM and language are largely distinct.
Theory of Mind and Executive Control: fMRI and Lesion Studies The other major cognitive resource that is implicated in passing a false belief task is inhibitory control. In the classic false belief task, the participant must be able to juggle two competing representations (the actual world and the world represented in the other person’s head) and to inhibit responding based on the true location of the object. In development, performance on false belief tasks is correlated with the child’s overall ability to select among competing responses. In a protocol closely matched to the false belief
403
task, a photograph is taken (or a map is drawn) of the object in location A, the object is moved to location B, and children are asked where the object will be in the photograph. No mental state understanding is required but 3-year-olds fail the task. In fact, typically developing children find the false photograph and false map task harder than the false belief task; by contrast, children with autism solve the false photograph and false map tasks, but fail the false belief. So are brain regions in the ‘ToM network’ implicated specifically in reasoning about other people’s mental states, or in the inhibitory control necessary to pass a false belief task? If these brain regions are involved specifically in reasoning about beliefs, they should not show a high response during other stories about false representations, like photographs or maps. Consistent with this prediction, Rebecca Saxe and Nancy Kanwisher initially reported that all of the brain regions in the ToM network are recruited significantly more when participants read about beliefs than about physical representations. In a careful subsequent study, Josef Perner and colleagues distinguished between stories describing photographs versus maps. Maps are more conceptually similar to beliefs because photographs are designed to represent the past, whereas both maps and beliefs are supposed to represent current reality. Using this tighter comparison, Perner et. al, found that only the right TPJ showed the profile of a region selectively involved in ToM: a significantly higher response to stories about beliefs than about photographs or maps, and no difference between photographs or maps and a fourth control condition involving no representations at all. Taken together, these results suggest that at least the right TPJ is specifically recruited for attribution of beliefs, and not for the inhibitory demands of the task. There is a challenge to this view, though. Lesion and imaging studies implicate a region of right TPJ in a different kind of task: changing the direction of attention in response to unexpected stimuli. Could this be the same brain region as the right TPJ implicated in studies of ToM? If so, it seems unlikely that it’s a specific mechanism for thinking about mental states. However, two recent studies have found that the regions for belief attribution and exogenous attention are neighboring but
404
Theory of Mind (Neural Basis)
distinct. Both studies revealed that the attention region is located approximately 10 mm superior to the region involved in ToM, close enough to be confusable, but far enough to probably be reflect a distinct group of neurons. So the current evidence suggests that regions in ToM network are involved in reasoning about minds, not in the inhibitory control necessary for performing false belief tasks. Importantly, there are other regions in the brain that are associated with inhibitory control, including intraparietal sulcus and frontal eye-fields. Reasoning about beliefs does provoke robust activity in these brain regions as well – but no more than is provoked by false photograph stories. That is, domain-general mechanisms for inhibitory control, response selection, and so on are recruited by both false belief and false photograph stories, and there are additional brain regions that are recruited only for beliefs. These results suggest that belief reasoning depends on both domain-general and domainspecific mechanisms, while reasoning about false photographs depends only on the domain-general components. A similar division between specific and general contributions to false belief task is suggested by a double dissociation observed in focal lesion patients. Samson and colleagues developed a series of elegant nonverbal false belief tasks. In one set of tasks (reality known), an object was moved without a character’s knowledge, but the participants themselves always knew the true location of the object. Passing these tasks required both the ability to represent the character’s belief, and the ability to inhibit the participant’s own knowledge (high inhibition). By contrast, in a second group of tasks (reality unknown), an object was moved without the character’s knowledge, but the participant also did not know the true location of the object and therefore did not have to inhibit any alternative location (low inhibition). WBA, a patient with left lateral frontal damage, was selectively impaired on the reality-known (i.e., high inhibition) false belief tasks. WBA also failed tests of inhibitory control that did not involve ToM, but performed normally on the realityunknown (i.e., low inhibition) false belief tasks. By contrast, a group of patients with left TPJ damage failed both high and low inhibition false
belief tasks, but did not fail general tests of inhibitory control outside of the domain of ToM. In all, fMRI and lesion evidence converge on three conclusions about the relationship between inhibitory control and ToM in the adult brain. (1) Successful performance on many kinds of false belief task depends on both domain-general inhibitory control and on domain-specific mechanisms for representing mental states. (2) These contributions are supported by distinct neural mechanisms, which can be dissociated. (3) The domain-specific component of ToM is supported, at least in part, by the TPJ. One important open question concerns the relative role of the right and left TPJ in ToM. The fMRI studies have pointed to right TPJ, while the left TPJ has been the focus of the lesion studies. Future studies should use the non-verbal tasks developed by Apperly and colleagues with patients who have lesions to right TPJ.
Theory of Mind and Person Perception: fMRI and Lesion Studies The research described above suggests that the ‘ToM network’ reflects a domain-specific mechanism, or mechanisms, for reasoning about other people and other minds. Of course, identifying the brain regions involved in ToM is just the first step. The real challenge is to understand ‘how’ these brain regions perform the relevant computations. A first step may be to understand the distinct contributions of each of these brain regions. Initial studies suggest that regions of MPFC are implicated in distinct components of social cognition, and the right TPJ region is the most selectively recruited for ToM. Some evidence for a division of labor between the TPJ and MPFC came from a study using two new kinds of stories, highlighting different aspects of reasoning about another person: (1) ‘Appearance’ stories described representing socially relevant information about a person that is visible from the outside, like the person’s clothing and hair color; and (2) ‘Thoughts’ stories described the contents of another person’s thoughts or beliefs, specifically tapping the later-developing component of ToM. A region involved in any general aspect of social cognition, including detecting the presence of a
Theory of Mind (Neural Basis)
person in the story, or tracking information about people, would show a high response in both story conditions, since all included socially relevant information about a protagonist. The MPFC showed exactly this predicted response: an equally high response to all story conditions that describe people. By contrast, a region involved specifically in attributing and reasoning about mental states like beliefs and desires should show a high response only in the ‘Thoughts’ condition. Consistent with this prediction, the right and left TPJ showed a high response when subjects read about thoughts or beliefs, but a low response (no different from resting baseline) when participants read about false photographs or a person’s physical appearance. Generally, the fMRI literature suggests a division in the neural system involved in making social judgments about others, with one component (the RTPJ) specifically recruited for the attribution of mental states, while a second component (the MPFC) is involved more generally in the consideration of the other person. If so, there should be tasks that do involve reasoning about another person, but do not involve attribution of specific mental states, and that therefore recruit the MPFC but not the RTPJ. There are, and they do. One example is the attribution of personality traits. Personality traits may be conceived as a heuristic alternative to ToM, allowing observers to predict a person’s ‘typical’ behavior, without considering that person’s specific mental states. In a series of studies, participants read single words describing human personality traits like ‘daring’ or ‘shy,’ and judged either (1) whether each word is semantically positive or negative, or (2) whether that trait applies to a specific person – either the self or a well-known other. As predicted, the MPFC, but not right or left TPJ, is recruited significantly more for traitattribution than for semantic judgment. In all, one critical topic of future investigation will be to differentiate the specific role of right and left TPJ, PC, and MPFC in ToM. In the meantime, though, the pattern response of brain regions in the ToM network can already be used to test cognitive hypotheses about the nature of ToM. For example, one important question arising from cognitive and developmental psychology is whether ToM function is automatic, or controlled.
405
Theory of Mind: Automatic or Controlled When do people formulate thoughts about thoughts? One possibility is that belief attribution is automatic, and occurs spontaneously whenever the situation ‘affords’ it. This model predicts that ToM is almost always engaged. Any observation of a human action affords the attribution of many desires and an infinite number of true beliefs. At the other end of the spectrum, mental state attribution may be applied only sparingly, when the situation specifically demands it. There is some recent behavioral evidence consistent with this latter option. In one study, participants watched another person act based on a false belief. In one condition, that person’s actions and beliefs were specifically relevant to the participant’s own task at that moment; in the other condition, the person’s actions and beliefs were not specifically relevant. Then, the participants were occasionally and unpredictably probed, to see whether they were automatically holding onto a representation of the person’s false belief. The results (using reaction times) suggested the opposite: participants only kept track of the other person’s false beliefs when those beliefs were specifically relevant to the participant’s own task. These two theories about ToM also make different predictions for the pattern of activation in the ToM network. If belief attribution is automatic, then whenever participants watch someone else act based on a belief, there will be activation in the ToM network. By contrast, if belief attribution is controlled and task dependent, then activation in the ToM network will depend on whether the observed person’s belief is specifically relevant to the participant’s current task. Converging with the behavioral studies, some recent fMRI results support the controlled view of ToM. Sommer and colleagues had participants in the scanner watch characters who had either a true or a false belief, and then make a prediction about their next action. The false belief trials thus required the participants to use the observer’s false beliefs to predict the observer’s behavior (e.g., looking in the wrong place). By contrast, the true belief trials do not require belief attributions at all; participants simply have to respond based on
406
Theory of Mind (Neural Basis)
the true location of the object. The response in the ToM network brain regions was higher on false belief trials, when belief attribution is required, than on true belief trials, when belief attribution is optional for task performance. There is another possible interpretation of Sommer and colleagues’ results, though. The authors themselves concluded that the ToM network regions (and especially the right TPJ) were recruited specifically for ‘false’ beliefs. However, in other previous studies, the response in the ToM network (and especially the right TPJ) has been equivalent for true and false belief conditions. More generally, one would expect ToM to be used for both true and false belief attributions: in our daily lives, we have to expect other people to have mostly true beliefs, or their actions would become completely unpredictable. So what explains the conflicting results? The difference between those previous studies, and the task used by Sommer and colleagues, is that the other studies used explicit verbal statements of the beliefs (e.g., ‘Sarah believes that her shoes are under the bed’). When the belief is stated explicitly in the story, then participants have no choice about whether to think about thoughts while reading the story: the decision is out of their hands. Stories that explicitly state beliefs thus cannot be used to test the automaticity of ToM. On the other hand, those stories can be used to distinguish between the two interpretations of Sommer and colleagues’ results. Since explicitly stated true and false beliefs elicit equal responses in the ToM network, it seems likely that the difference observed by Sommer and colleagues was due, not to truth and falseness, but to the optional versus required belief attributions. If so, then those results are also evidence that thinking about other people’s thoughts does not occur automatically (Tables 1 and 2). More broadly, predicting which tasks or stimuli will elicit ToM – and thus elicit activation in the ToM network – is a critical challenge for cognitive neuroscience. Research on mental state attribution outside of the false belief task is thus beset by a chicken-and-egg problem. In order to know that a brain region is truly the neural substrate of ToM, we must establish that it is recruited for all and only the tasks that involve ToM – including going beyond the false belief task. However, exactly when people do think about thoughts, outside the
Table 1 Verbal stimuli eliciting a robust response in the right TPJ 1. False Beliefs. Anne made lasagna in the blue dish. After Anne left, Ian came home and ate the lasagna. Then he filled the blue dish with spaghetti and replaced it in the fridge. Anne thinks the blue dish contains (choose one) spaghetti/ lasagna. 2. Beliefs – not false. Kate knew her colleague was very punctual and always came to work by 9. One day, he still wasn’t in at 9:30. Since he was never late, Kate assumed he was home sick. 3. False beliefs – no mental state verbs. A boy is making a papier mache project for his art class. He spends hours ripping newspaper into even strips. Then he goes out to buy flour. His mother comes home and throws all the newspaper strips away. 4. Mental states that are unexpected, given other information about the protagonist. (Background: Your friend Carla lives in San Francisco. She has a top position at a large computer company there. She has been working at the same corporation for over 10 years.) Carla has always told you that she wants a husband who would expect her to stay at home as a housewife, instead of having her own career. 5. Mental states described in the second person. You went to London for a weekend trip and you would like to visit some museums and different parks around London. In the morning, when you leave the hotel, the sky is blue and the sun is shining. So you do not expect it to start raining. However, walking around in a big park later, the sky becomes gray and it starts to rain heavily. You forgot your umbrella. 6. Morally relevant beliefs. (. . .) Because the substance is in a container marked ‘toxic,’ Grace thinks that it is toxic. 7. Induced false beliefs. The path to the castle leads via the lake, but the children tell the tourists, ‘‘The way to the castle goes through the woods’’. The tourists now think the castle is?
false belief task, has not yet been established by independent methods. In some cases, the best evidence that a task does differentially invoke ToM may be the level of activation in the ToM network (a ‘reverse’ inference)! Fortunately, this problem is temporary. Different theories of ToM and different theories of the function of the ToM network can be tested simultaneously. The correct resolution will be determined by the consistency of the data that emerge from these tests, and the richness of the theoretical progress those data support. One key dimension of task design is whether we think of ToM as predominantly (1) the process of initially forming a representation of someone’s thought, or (2) the process of using that representation to answer questions and make predictions.
Theory of Mind (Neural Basis) Verbal stimuli not eliciting a robust response in the right TPJ
Table 2
1. False physical representations. This map shows the ground floor plan. A photocopy was sent to the architect yesterday. The map initially had a flaw: the kitchen door was missing. It was added to the map this morning. The architect’s photocopy (choose one) includes/doesn’t include the kitchen door. 2. Physical inferences. The night was warm and dry. There had not been a cloud anywhere for days. The moisture was certainly not from rain. And yet, in the early morning, the long grasses were dripping with cool water. 3. Descriptions of visible facts about people. Harry looks just like a math professor. He wears dark old cardigans with holes in the elbows, corduroy trousers and brown loafers over green argyle socks. The shoes Harry wears are (choose one) brown/green. 3b. Visible facts that convey social information. Joe was a heavy-set man, with a gut that fell over his belt. He was balding and combed his blonde hair over the top of his head. His face was pleasant, with large brown eyes. 4. Social background information about a person. Your friend Carla lives in San Francisco. She has a top position at a large computer company there. She has been working at the same corporation for over 10 years. 5. Bodily sensations – subjective but not representational states. Marcus had been sick for 3 days. He had felt weak and had a high fever. On the fourth day his fever broke, and he woke up feeling cool and alert. 6. Morally relevant facts. (. . .) The white powder by the coffee is not sugar, but a toxic substance left behind by a scientist. 7. False signs. The sign to the monastery points to the path through the woods. While playing the children make the sign point to the golf course. According to the sign, the monastery is now?
This question has important implications for studies of ToM that use methods with high temporal resolution, like event-related potentials (ERPs) from an electroencephalogram (EEG).
ERP Studies of Theory of Mind The studies described above, using fMRI, lesions and TMS, help to establish ‘where’ ToM processing happens in the brain. In order to eventually understand ‘how’ this processing happens, another group of methods must be brought into the picture: methods with high temporal resolution, that can help reveal ‘when’ the relevant processing happens. High temporal resolution has two benefits. First, knowing when a certain kind of processing occurs can be useful in its own right. For
407
example, the speed of a process can constrain the relative contributions of bottom-up (faster) and top-down (slower) input to that process. Second, once we know both where and when to look for a specific process, then we can begin to measure the actual firing of neurons that is responsible for that process. At the moment, recording for individual neurons involved in ToM would be extremely hard because we don’t know when, during the task, ToM is likely to be happening. As a first approach to these challenges, Mark Sabbagh, David Liu, Henry Wellman and colleagues have conducted a series of studies of ToM using event-related EEG potentials. ERPs have very high temporal resolution (it is possible to distinguish between signals that are separated by 10 ms) but are not very good for spatial resolution (establishing where those signals come from). Critically, ERP analyses require that there is a specific and predictable start time for the events of interest. In this case, the event of interest is thinking about another person’s thoughts. When would that start? As hinted above, the answer depends on whether we think of ToM as the process of initially forming a representation of someone’s thought, or as the process of using that representation to answer questions and make predictions. The timing of initially forming a representation is very hard to predict. In a typical false belief task, the participant might begin to consider the character’s mental states when the character turns his back and/or leaves the room, when the object is removed from its original location or placed into the novel location, or when the character returns to the room. Sabbagh and colleagues avoided this uncertainty by focusing on the process involved in answering the critical ToM question. In their version of the false belief (or false photograph) task, there are always two objects in the scene: A and B. While the character is outside of the room (or after the photograph is taken), object A is moved to a new location, and object B remains in its original location. Participants are then asked ‘‘According to [the character/ the photograph], where is the [Object A/Object B]?’’ Retrieval of the character’s specific belief about the target object must therefore begin at the onset of the final word of the question – the first moment at which participants know which object is the target.
408
Theory of Mind (Neural Basis)
Using this paradigm, Sabbagh and colleagues compared the brain activation when participants answered questions about false beliefs versus false photographs. They did find a specific difference: a left-lateralized frontal late slow wave was larger on false belief relative to false photograph trials. However, this ERP pattern did not come from any of the brain regions in the ‘ToM network.’ In fact, none of the brain regions in the ToM network were identified by the ERP method. How should we explain the discrepancy between ERP and other methods? One possibility is that the ERP study focused on the processes for answering questions about ToM, whereas the fMRI, lesion and TMS studies may have been mostly tapping the process for forming a representation of someone’s belief, in the first place. In Sabbagh and colleagues’ task, since participants are asked about the content of a belief or photograph on every trial, they may preemptively formulate a representation of all the relevant beliefs or photographs as the trial is evolving, long before the critical question. Consistent with this idea, the ToM network revealed by fMRI appears to be involved mostly in the initial inference of beliefs. Although fMRI lacks the temporal resolution of ERPs, false belief stories are typically presented for 10–20 s, followed by a question presented for 4–6 s. The response in the TPJ, MPFC and medial parietal regions occurs almost exclusively in the first time period, while the participants are reading the story and forming their initial inferences and representations. These results suggest a distinction between the processes supporting initially inferring someone’s belief, and then using that belief to answer questions and make predictions. If this is right, then it will be important to devise new tasks and methods for studying the elusive inference process with high temporal resolution.
sensory perception of the local environment, motor planning and control. Understanding someone else’s action may seem by contrast like a highly abstract – if not semimiraculous – achievement. Recently, though, many researchers have proposed that this abstract higher cognitive function could have concrete sensorimotor foundations. That is, an observer might understand someone else’s action using the same cognitive and neural mechanisms that she uses to plan her own. The idea is sometimes called the ‘motor theory of social cognition.’ One advantage of the motor theory of social cognition is its parsimony. Action prediction and understanding could be achieved with the same cognitive and neural mechanisms that the observer already uses for her own action planning and execution; she does not need a whole extra system for ToM. As described above though, recent neuroscientific evidence undermines this view. There are brain regions specifically implicated in attributing mental states, and these brain regions are not part of the observer’s own motor system. The ToM network is completely distinct, anatomically, from the brain regions implicated in action execution or action perception. Many neuroimaging studies have focused on the overlapping activation during action perception and action execution, of ventral premotor cortex, inferior frontal gyrus and right inferior parietal cortex. By contrast, the regions implicated in ToM have no known role in motor planning or action execution. Although the two groups of regions are clearly anatomically segregated, their functional properties have not yet been investigated within a common task. Future work should investigate tasks in which both ToM and the mirror system are involved, to allow for a direct functional dissociation between these mechanisms.
Conclusions Theory of Mind and Mirror Neurons In the domain of ‘understanding human actions,’ there is an intuitive distinction between one’s own actions, and actions executed by others. The neural mechanisms necessary for executing one’s own goal directed action are fairly concrete, including
In all, recent studies are beginning to shed light on the brain regions involved when human adults reason about one another’s minds – that is, in ToM. One surprising result of these early studies is that a specific group of cortical regions is reliably implicated in ToM, the so-called ‘ToM network,’ including the right and left TPJ, MPFC and PC.
Theory of Mind (Neural Basis)
The brain regions involved in ToM are incredibly robust. These regions can be identified in 90% of individual subjects, after just 20 min of scan time; the same regions have been reported by labs in different continents, using different procedures and different stimuli. The same group of regions have been identified as relevant for ToM by lesion and TMS studies. Similarly, reliable patterns of activation are routinely observed for perceptual mechanisms, like primary sensory and motor cortices, but rarely for dimensions of cognition as abstract and complex as ToM. The ‘ToM’ regions thus offer a rare window through the brain to the mind. Nevertheless, critically important questions remain open: What are the specific and distinct roles of the brain regions that make up the ToM network? How does the ToM network interact with other brain systems underlying language, inhibitory control, and action perception (e.g., the mirror system)? How do these brain regions develop? Are these brain mechanisms universal to all human beings? Are they uniquely human? Lots of important research remains to be done. See also: Development: Consciousness From Birth to Adulthood; Social Foundations of Consciousness.
Suggested Readings Apperly IA, Samson D, et al. (2006) Intact first- and second-order false belief reasoning in a patient with severely impaired grammar. Social Neuroscience 1(3–4): 334–348.
409
Bloom P and German TP (2000) Two reasons to abandon the false belief task as a test of theory of mind. Cognition 77(1): B25–B31. Jacob P and Jeannerod M (2005) The motor theory of social cognition: A critique. Trends in Cognitive Science 9(1): 21–25. Leslie AM and Thaiss L (1992) Domain specificity in conceptual development: Neuropsychological evidence from autism. Cognition 43(3): 225–251. Perner J, Aichorn M, et al. (2006) Thinking of mental and other representations: The roles of left and right temporo-parietal junction. Social Neuroscience 1(3–4). Rizzolatti G, Fogassi L, et al. (2001) Neurophysiological mechanisms underlying the understanding and imitation of action. Nature Reviews Neuroscience 2(9): 661–670. Sabbagh MA and Taylor M (2000) Neural correlates of theory-of-mind reasoning: An eventrelated potential study. Psychological Science 11(1): 46–50. Samson D, Apperly IA, et al. (2004) Left temporoparietal junction is necessary for representing someone else’s belief. Nature Neuroscience 7(5): 499–500. Saxe R and Kanwisher N (2003) People thinking about thinking people. The role of the temporo-parietal junction in ‘‘theory of mind’’ Neuroimage 19(4): 1835–1842. Saxe R and Powell LJ (2006) It’s the thought that counts: specific brain regions for one component of theory of mind. Psychological Science 17(8): 692–699. Sommer M, Dohnel K, et al. (2007) Neural correlates of true and false belief reasoning. Neuroimage 35(3): 1378–1384. Varley R and Siegal M (2000) Evidence for cognition without grammar from causal reasoning and ‘theory of mind’ in an agrammatic aphasic patient. Current Biology 10(12): 723–726. Wellman HM, Cross D, et al. (2001) Meta-analysis of theory-of-mind development: The truth about false belief. Child Development 72(3): 655–684.
Biographical Sketch
Unconscious Cognition J F Kihlstrom, University of California, Berkeley, CA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Automaticity – The label given to cognitive and other mental operations that are executed outside phenomenal awareness and voluntary control. Automatic processes contrast with controlled processes in terms of four canonical features: they are inevitably evoked by the presence of a particular stimulus; incorrigibly executed, in a ballistic fashion, once evoked; effortless, meaning that they consume little or no cognitive capacity; and efficient, in that they create little or no interference with other cognitive activities. Conscious shyness – The resistance of psychology and other cognitive sciences to making consciousness a focus of inquiry. As described by Owen Flanagan, conscious shyness has four facets: ‘positivistic reserve,’ or a preference for publicly verifiable behavioral data over introspective self-reports; ‘piecemeal approach,’ the assumption that an understanding of consciousness will emerge from more narrowly defined studies of perception, memory, and the like; ‘conscious inessentialism,’ the view that consciousness is not necessary for certain cognitive functions, such as perceiving and remembering; and ‘epiphenomenalism,’ the view that consciousness plays no causal role in mental life. Dissociation – An experimental outcome in which one independent variable, such as brain damage, has different (e.g., opposite) effects on two or more dependent variables. Many studies of unconscious cognition indicate that some variable (such as brain damage or the direction of attention) dissociates, or has different effects on, explicit and implicit expressions of perception, memory, and the like.
Inattentional blindness – Failure to consciously perceive an object in the visual field because one’s attention is directed elsewhere. In various forms of ‘attentional blindness,’ such as repetition blindness, the attentional blink, and change blindness, the subject fails to perceive an object (or a change in an object) despite the fact that attention is properly directed to the appropriate portion of the visual field. Learning – A relatively permanent change in knowledge which occurs as a result of experience. ‘Explicit learning’ covers cases where the subject is consciously aware of this new knowledge. ‘Implicit learning’ refers to any effect of acquired knowledge affects experience, thought, or action in the absence of conscious awareness of what has been learned. Limen – In classical psychophysics, the ‘threshold,’ or minimum amount of stimulus energy needed for an observer to consciously detect the presence of, or change in, a stimulus. In ‘subliminal perception’ the stimulus is too weak to be consciously perceived, or has been degraded by virtue of brief presentation, a masking stimulus, or the deployment of attention elsewhere. Memory – The mental representation of some past event. ‘Explicit memory’ entails conscious recollection of some past event, as in recall or recognition. ‘Implicit memory’ refers to any effect on experience, thought, or action of a past event, in the absence of conscious recollection of that event. Perception – The mental representation of an event in the current stimulus environment, or that of the very recent past. ‘Explicit perception’ is indicated by the observer’s conscious awareness of the size, shape,
411
412
Unconscious Cognition
distance, motion, and identity of some stimulus. ‘Implicit perception’ refers to any effect on experience, thought, or action of such an event, in the absence of conscious perception of that event. Preattentive processes – Perceptual and cognitive operations that occur before attention has been paid to a stimulus. Priming – The facilitation (or impairment) of processing of a stimulus (the target) by prior presentation of another stimulus (the prime). In ‘repetition priming,’ the prime physically resembles the target; in ‘semantic priming,’ the prime and target are related in terms of meaning. Process-dissociation procedure – A method, invented by Jacoby, for quantifying the contributions of automatic and controlled processes to task performance by pitting the one against the other. In the ‘Inclusion’ condition, automatic and controlled processes are allowed to work in the same direction; in the ‘Exclusion’ condition, automatic processes are placed in opposition to controlled processes. Thought – As a verb, thinking refers to various cognitive operations, such as reasoning, problem-solving, judgment, and decision-making. As a noun, thought refers to any mental representation, such as the idea or an image of a thing, which is itself neither a percept (a mental representation of a stimulus in the current environment) nor a memory (a mental representation of a past stimulus). Unconscious inference – As described by Helmholtz, the inferences or thought processes that, automatically and unconsciously, generate conscious perception from stimulus information. For example, perceivers unconsciously apply the size–distance rule to take information about size into account when judging the distance of an object, and vice-versa, to create the ‘moon illusion’ and other perceptual effects.
Unconscious Cognition As psychology developed as an empirical science, theory and research focused on those mental states that were accessible to consciousness, as represented by classical psychophysics (with its interest in sensory thresholds), the structuralists’ method of introspection, and William James’ classic essay, in the Principles of Psychology, on the stream of consciousness. Following the onslaught of radical behaviorism, interest in consciousness declined precipitously in the years after World War I. Woodworth’s work on the span of apprehension was an exception, and consciousness was the subtext of the Gestalt approach to perception. But serious scientific interest in consciousness had to await the cognitive revolution, with its interest in attention, short-term memory, and mental imagery. The widespread current interest in consciousness is illustrated by the very existence of encyclopedias like the present one. Even in the nineteenth century, however, psychologists recognized that the mental structures and processes underlying experience, thought, and action were not completely encompassed within the span of conscious awareness. Herbart’s ‘limen’ made room for Leibniz’s ‘petites perceptions,’ and of course Helmholtz argued that conscious perception was based on unconscious inferences. Perhaps the most forceful advocate of nonconscious mental life was William James, who (again in the Principles) held that mental states could be unconscious in at least two different senses. First, a mental event can be deliberately excluded from attention or consciousness: ‘‘We can neglect to attend to that which we nevertheless feel’’ (p. 201). These unattended, unconscious thoughts and feelings are themselves mental states. Second, and more important, James drew on clinical observations of cases of hysteria and multiple personality to argue for a division of consciousness into primary and secondary (and, for that matter, tertiary and more) consciousnesses (sic), only one of which is accessible to phenomenal awareness at any point in time. To avoid possible oxymoron in the negation of consciousness, James preferred to speak of ‘co-conscious’ or ‘subconscious’ mental states,
Unconscious Cognition
rather than ‘unconscious’ ones. There, was, in fact, considerable interest in unconscious mental life in the early twentieth century – even setting aside the work of Sigmund Freud. But, as with consciousness, the entire topic was banished from serious discourse by behaviorism. History repeated itself in the latter half of the twentieth century, as the ‘consciousness revolution’ which followed on the cognitive revolution spun off an interest in unconscious mental life as well. This happened slowly. For a long time, if the unconscious was discussed at all, it was construed merely as a repository for unattended percepts and forgotten memories. But in addition to this wastebasket view of the unconscious, the classic multistore model of information processing made room for the unconscious by identifying unconscious mental life with early, ‘preattentive’ mental processes that occur prior to the formation of a mental representation of an event in primary memory. However, by implicitly identifying consciousness with ‘higher’ mental processes, the classic multistore model left little or no room for a more interesting view of the ‘psychological unconscious’ – complex mental structures and processes, and full-fledged mental states, that influence experience, thought, and action, but which are nevertheless inaccessible to phenomenal awareness, insusceptible to voluntary control, or both. The rediscovery of the unconscious by modern scientific psychology began with comparisons between automatic and effortful mental processes and between explicit and implicit memory. Since then, it has continued with the extension of the explicit–implicit distinction into the domains of perception, learning, and thought. Taken together, this literature describes the ‘cognitive unconscious.’
Attention, Automaticity, and Unconscious Processing The earliest information-processing theories of attention, such as those proposed in the 1950s by Broadbent and others, were based, to one degree or another, on the metaphor of the filter. Information which made it past the filter was available for ‘higher’ information-processing activities, such as semantic analysis, while information which did not
413
make it past the filter was not. This same attentional filter was also the threshold which had to be crossed for information to be represented in phenomenal awareness. The filter theories of attention, in turn, raised questions about how permeable the attentional filter was, and how much information processing could occur preattentively. Was preattentive – preconscious – processing limited to elementary physical features of the stimulus, or could it extend to the meaning of the stimulus as well? In part to solve these problems, the notion of an attentional filter was replaced by the notion of attentional capacity. Capacity theories, such as the one proposed by Kahneman in the early 1970s, began with the assumption that human information-processing capacity is limited, and proposed that the ability to perform one or more task(s) depended both on the resources available and the resources required by the task(s) themselves. Whereas the filter models conceived of information processing as serial in nature, the capacity models implied that several tasks could be carried out simultaneously, so long as their attentional requirements did not exceed available resources. The capacity view, in turn, led in the mid-1970s to a distinction between two types of cognitive processes, ‘controlled’ and ‘automatic.’ Controlled processes are conscious, deliberate, and consume cognitive capacity – they are what most people mean by cognition. By contrast, automatic processes are more involuntary. That is, they are ‘inevitably evoked’ by the presentation of specific stimulus inputs, regardless of any intention on the part of the subject. And once evoked by an effective environmental stimulus, they are ‘incorrigibly executed,’ in a ballistic fashion. Automatic processes are ‘effortless,’ in that they consume little or no attentional capacity. And they are ‘efficient,’ in that they do not interfere with other ongoing mental activities. Perhaps because they are fast, or perhaps because they do not consume cognitive capacity, automatic processes are unconscious in the strict sense that they are inaccessible to phenomenal awareness under any circumstances, and can be known only by inference. Automatic processes are exemplified by the Stroop color-word task, in which subjects must name the color of ink in which words are printed: Subjects show a great deal of interference when the
414
Unconscious Cognition
word names a color that is different from that in which the word is printed. Apparently, subjects simply cannot help reading the words. Helmholtz’s unconscious inferences may be viewed, in retrospect, as an early foreshadowing of the concept of automaticity. More recently, linguists such as Chomsky have argued that the universal grammar that underlies our ability to use language operates unconsciously and automatically. Over subsequent decades, the concept of automaticity has evolved further. For example, some theorists proposed that automatic processes have properties other than the four canonical ones outlined above. Others have suggested that the features represent continuous dimensions, so that processes can be more or less automatic in nature. There has been some question as to whether all the canonical features must be present to identify a process as automatic: the four canonical features might comprise a kind of prototype of automaticity, rather than being singly necessary and jointly sufficient to define a process as automatic. Even the automaticity of the Stroop effect has been cast into doubt. Challenges to capacity theory, from which the earliest ideas about automaticity emerged, have led to alternative theoretical conceptualizations in terms of memory rather than attention. Nevertheless, the concept of automaticity has gained a firm foothold in the literature of cognitive psychology, and investigators have sought to develop methods, such as Jacoby’s ‘process-dissociation procedure,’ to distinguish between the automatic and controlled contributions to task performance.
Implicit Memory While automatic processes may be considered to be unconscious, the mental contents on which they operate, and which they in turn generate, are ordinarily thought to be available to conscious awareness – just as Helmholtz’s unconscious inferences took consciously accessible stimulus information and operated unconsciously to generate the conscious perception of distance and size. Thus, we generally assume that people notice and can describe the salient features of an object or event, even if they cannot articulate the way in which those features have been integrated to form
certain judgments made about it, or cannot control their response to it. The further possibility, that cognitive processes can operate on mental states – percepts, memories, and the like – that are not themselves accessible to conscious awareness, was first raised in modern psychology by observations of priming in neurological patients with the ‘amnesic syndrome’ resulting from bilateral damage to the medial temporal lobe, including the hippocampus. These patients could not remember words that they had just studied, but nevertheless showed normal levels of priming on tasks such as word-fragment completion and stem completion. On the basis of results such as these, Schacter drew a distinction between explicit memory, which involves the conscious recollection of some past event, and implicit memory, which is revealed by any change in task performance that is attributable to that event. Explicit memory, in this view, is identified with the typical sorts of memory tasks, such as recall or recognition. Following Schacter, we may define implicit memory formally as the effect of a past event on the subject’s ongoing experience, thought, and action, in the absence of, or independent of, conscious recollection of that event. Implicit memory is, in these terms, unconscious memory. Spared priming has also been observed in various other forms of amnesia, including the anterograde and retrograde amnesia associated with electroconvulsive therapy for depression; the anterograde amnesia produced by general anesthesia administered to surgical patients, as well as that associated with conscious sedation in outpatient surgery; memory disorders observed in dementia, including Alzheimer’s disease, as well as those encountered in normal aging; hypnotic and posthypnotic amnesia following appropriate suggestions to hypnotizable subjects; and the ‘functional’ or ‘psychogenic’ amnesias encountered in genuine cases of dissociative disorder, including dissociative amnesia, dissociative fugue, and the interpersonality amnesia of dissociative identity disorder (also known as multiple personality disorder). In each of these cases, the memory disorder primarily impairs explicit memory and spares implicit memory, which is either wholly or relatively intact. It is in this sense that implicit memory can be observed in the absence of explicit memory.
Unconscious Cognition
However, implicit memory can be observed in individuals with normal memory functions as well. For example, normal subjects show significant savings in relearning for items that they can neither recall nor recognize. In principle, such phenomena as proactive and retroactive inhibition might also be expressions of implicit memory, if the interference occurred in the absence of conscious recollection for the interfering list. For the most part, however, documented dissociations between explicit and implicit memory in individuals with normal memory function take the form of experimental dissociations, in which some variable is shown to affect explicit but not implicit memory, or vice-versa. For example, depth of processing has a substantial effect on explicit memory, but relatively little impact on many priming effects. Even when explicit memory is not impaired, implicit memory may be said to be independent of explicit memory, in that priming does not depend on whether the prime is consciously remembered. The nature of implicit memory is the subject of continuing theoretical controversy. The earliest view was that priming reflected the activation, through perception and attention, of stimulusrelevant knowledge already stored in memory. More recently, under the influence of the doctrine of modularity in cognitive neuroscience, explicit and implicit memory have been attributed to the operation of separate and independent (thus dissociable) memory systems in the brain – or, alternatively, to brain systems dedicated to perception, action, and conceptual analysis that have the capacity to learn. Other theories hold that explicit and implicit memory are the products of a single memory system, but differ in terms of the processes that are deployed at the time of encoding and retrieval. For example, many priming tasks are held to be perceptually driven, while recall and recognition is conceptually driven. Alternatively, priming may occur automatically, while recall and recognition require conscious effort in retrieval. Along these lines, some theorists have redefined implicit memory as involuntary memory, thus making conscious control rather than conscious awareness, the defining feature of the phenomenon. The various theories have proved hard to test against each other. By and large, multiple-memory systems theories have emerged from studies of
415
brain-damaged patients, while single-systems theories have been developed in the context of data collected from neurologically intact subjects. Activation theories seem to imply that priming cannot occur for novel materials, for which subjects have no pre-existing mental representations to activate. Priming can occur for novel materials, in apparent contradiction to the activation view; but priming is greatest for novel materials that are composed of familiar elements, lending support to the activation view after all. Despite this theoretical controversy, the essential concept of implicit memory, including its dissociation from explicit memory, is now widely accepted within cognitive psychology. The theoretical debate is likely to persist until investigators in the field recognize that most subsequent research on implicit memory, whether in neurologically impaired patients or intact subjects, has been dominated by studies of repetition priming effects of the sort observed in lexical decision, perceptual identification, stem-completion, and fragment-completion tasks. However, there is more to priming than repetition priming. Semantic and conceptual priming effects can also be observed in amnesia, as on free association and category generation tasks. These priming effects require some analysis of the meaning of the stimulus, going beyond mere physical structure. Understanding the nature of implicit memory, when it is spared and when it is impaired, and the relations between implicit and explicit memory, requires research on these other expressions of unconscious memory.
Implicit Learning Closely related to implicit memory is the concept of ‘implicit learning’ – a term introduced to psychological discourse by Reber’s pioneering experiments on artificial-grammar learning. In these experiments, subjects first studied a set of letter strings which had been generated by an artificial grammar. Later, they were asked to examine a new set of letter strings, and identify those that conformed to the grammar. They were able to do so at better than chance levels, suggesting that they had induced the grammatical rule from the study set.
416
Unconscious Cognition
Nevertheless, Reber reported that the subjects were not able to describe the grammar itself. In a paradigm somewhat similar to artificial grammar learning, subjects have learned to identify instances of novel concepts, such as patterns of dots that vary around a prototype, without being able to describe the defining or characteristic features of the concepts themselves. They can also detect the covariation between two features, such as hair length and personality, even though they cannot identify the basis for their predictions. They can learn the sequence in which certain stimuli will occur, without being able to specify the sequence itself. And they can learn to control the output of a complex system by manipulating an input variable, without being able to specify the relationship between the two. In each case, subjects appear to have acquired new knowledge through experience – the very definition of learning – but were apparently unable to gain conscious access to this knowledge. Thus, it appears that explicit learning can be dissociated from implicit learning. Following the model of implicit memory, implicit learning may be defined as a relatively permanent change in knowledge, resulting from experience, in the absence of conscious awareness of what has been learned. Demonstrations that amnesic patients retain the ability to learn, even though they do not remember the learning experiences themselves, have led some theorists to construe implicit learning as a variant on implicit memory. However, there is an important distinction between the two concepts: implicit memory is a feature of episodic knowledge, in which subjects lack conscious memory for a specific event in their lives. By contrast, in implicit learning subjects lack conscious access to certain pieces of semantic or procedural knowledge acquired through a learning experience. Cases where subjects are aware of knowledge acquired through a learning experience, but cannot consciously recollect the learning experience itself, are better classified as ‘source amnesia,’ a variant on implicit memory. Implicit learning is also distinguished from merely ‘incidental’ learning, in which new knowledge is acquired in the absence of instructions or intention to learn, but the subject retains conscious access to that knowledge. Incidental learning is unintended, just as involuntary memory is unintended; but implicit
learning reflects unconscious knowledge, just as implicit memory reflects unconscious memory. Whether implicit learning can be dissociated from explicit learning remains an open question. The fact that subjects cannot articulate the Markov process by which grammatical strings were generated does not mean that they are unaware of what they have learned. Above-chance classification performance could well result from partial knowledge which is consciously accessible. The same argument holds true for other paradigms in which implicit learning has been ostensibly demonstrated. The best that can be said, for now, is that the subjects in artificial grammar and sequence learning experiments often experience themselves as behaving randomly, without an awareness of what they are doing. However, this assumption rests on relatively informal evidence. The careful matching of explicit and implicit tasks that is characteristic of studies on implicit memory has not generally been imported into the study of implicit learning. A major agenda item in the study of implicit learning is to carry out more detailed analyses of subjects’ experiences in implicit learning situations, to make sure that they are really unconscious of what they evidently know and put into action. It is sometimes claimed that implicit learning is more powerful than conscious knowledge acquisition – that, for example, subjects can learn more complicated things, and more quickly, implicitly than they can explicitly. There may be cases in which this is true – for example, the learning of one’s native language(s). But for the most part, such claims are rarely supported by adequately designed experiments. Most studies of implicit memory are little more than demonstration experiments, showing that something can be learned without any instruction or intention to do so, and with little conscious representation of what has been learned. But by and large, they lack the careful control conditions that would enable the investigators to pit conscious and unconscious learning against each other.
Implicit Perception Just as there are palpable effects on experience, thought, and action of past events that cannot be
Unconscious Cognition
consciously remembered, there also appear to be similar effects of events in the current stimulus environment, which cannot be consciously perceived. By analogy with memory, explicit perception may be identified with the subject’s conscious perception of some object in the current environment, or the environment of the very recent past, as reflected in his or her ability to report the presence, location, form, identity, and/or activity of that object. By the same token, implicit perception entails any change in the person’s experience, thought, or action which is attributable to such an event, in the absence of (or independent of) conscious perception of that event. Implicit perception is exemplified by what has been called ‘subliminal’ perception, involving stimuli that are in some sense too weak to cross the threshold (limen) of conscious perception. In the earliest demonstrations of subliminal perception, the stimuli in question were of very low intensity – ‘subliminal’ in the very strict sense of the term. Later, stimuli of higher, supraliminal levels of intensity were rendered ‘subliminal’ by virtue of extremely brief presentations, or by the addition of a ‘masking’ stimulus which for all practical purposes rendered the stimuli invisible. Subliminal perception quickly entered everyday vocabulary, as in the controversy over subliminal advertising and the marketing of subliminal selfhelp tapes. Almost from the beginning of this line of research, a variety of methodological critiques sought to undercut claims for subliminal perception. However, beginning with the now-classic studies of Marcel, followed by the work Merikle, Greenwald, and their associates, an increasing body of literature has demonstrated subliminal perception, broadly defined, in a manner that convincingly addresses these criticisms. The adjective ‘implicit’ is preferable to ‘subliminal,’ because there are many cases of unconscious perception in which the stimulus is not subliminal in the technical sense of the term. The most dramatic example is that of ‘blindsight’ in patients with lesions in striate cortex. Implicit perception can also be observed in the syndromes of visual neglect resulting from temporoparietal damage, and in discriminative physiological responses to familiar and unfamiliar faces in cases of prosopagnosia. Priming and related effects have also been
417
observed in the ‘functional’ disorders of perception, such as cases of ‘hysterical’ blindness and deafness (now relabeled ‘conversion disorder’), as well as in normal subjects given hypnotic suggestions for blindness and deafness. Also in normals, implicit perception has also been observed in subjects whose attention has been deflected from the stimulus, so that it is processed outside conscious awareness, as in parafoveal vision and dichotic listening. Priming has also been observed in normal failures of conscious perception such as ‘inattentional blindness,’ as well as some forms of ‘attentional’ blindness such as ‘repetition blindness,’ and the ‘attentional blink’ (conscious perception also fails in the phenomenon of ‘change blindness,’ but to date there have been no demonstrations of priming in this phenomenon). As with implicit learning, implicit perception effects are sometimes discussed under the rubric of implicit memory. Indeed, the prime in most implicit perception studies does occur in the past, but it is the past of only a few moments ago, with no intervening distraction – what James called the ‘specious present,’ and well within the span of primary memory. Implicit perception can be distinguished from implicit memory as follows: In implicit memory, the subject was perceptually aware of the event at the time it occurred; but the memory of that event has been lost to conscious recollection. In implicit perception, the subjects were unaware of the event at the time it occurred; conscious recollection may be vitiated by the failure of conscious perception, but it is the perception itself, not memory, that is unconscious. Preserved priming in general anesthesia might be construed as an instance of implicit memory, because the test of memory occurs some time after the primes were presented; on the other hand, because the patients were not aware of the primes at the time they were presented, the same phenomenon should also count as an instance of implicit perception. With subliminal and other forms of implicit perception now demonstrated to the methodological satisfaction of (almost) everyone, most of the remaining controversy concerns the extent of its influence. Dramatic claims for the power of subliminal influence, popularized by Vance Packard’s expose ‘The Hidden Persuaders’ and revived in
418
Unconscious Cognition
the American presidential election of 2000 by a Republican television advertisement that appeared to present the word ‘rats’ subliminally when discussing the Democrats, appear to be grossly exaggerated. In the past, debate focused on whether subliminal perception was limited to the analysis of the physical characteristics of the subliminal stimulus, as in repetition priming, or could extend to meaning analyses, leading to semantic priming. The work of Marcel, Merikle, and Greenwald come down squarely in favor of semantic priming, but even here subliminal perception appears to be analytically limited.
Implicit Thought Thinking and consciousness are so closely related that the notion of unconscious thought seems to be a contradiction in terms – certainly James thought so. In line with Helmholtz’s concept of unconscious inferences, it is now generally accepted that many of the thought processes that intervene between perception and action are automatic, and it is in this sense that we may say that at least some thinking can occur outside of awareness. However, there is some evidence that thoughts themselves, and not just the process of generating them, can be unconscious. In one example, Bowers and his associates found that subjects could distinguish between soluble and insoluble puzzles, without knowing what the solution to the soluble puzzle was. Subsequent research has showed priming effects on lexical decision from the solutions to word problems, even when subjects were unaware of the solution itself. It has also been shown that subjects can display changes in skin-conductance response when making objectively risky choices, even when they cannot subjectively distinguish between those that are risky and those that are safe – much as prosopagnosic patients show differential physiological responses to objectively familiar faces that they cannot recognize. In some sense, the ability of Bowers’ subjects to discriminate between soluble and insoluble puzzles resembles the ability of Reber’s subjects to discriminate between grammatical and ungrammatical letter strings – with the exception that they have had no opportunity to acquire the
distinction through learning. Similarly, Shames’ semantic priming effect resembles those observed in implicit perception and memory – except that the prime is neither a perceptual representation of a current stimulus nor a representation available in memory of a stimulus presented in the past. The subjects in these experiments appear to experience a ‘feeling of knowing’ analogous to that observed in metamemory tasks. Their choices are clearly being guided by something that is neither a percept (because the solution is not currently being presented to them) nor a memory (because the solution has not been presented in the past). By analogy to implicit perception and memory, we may define ‘implicit thought’ as a mental representation – an idea or an image, for example – that influences ongoing experience, thought, and action in the absence of conscious awareness of that representation. Implicit thought may underlie the intuition stage of insight problem solving: the subject has not yet consciously arrived at the solution, as a conscious insight, but knows that a solution is possible, the solution itself is unconscious. In this analysis, the incubation stage may be thought of as the process by which the transformation from unconscious influence to conscious access takes place. Siegler and Stern may have observed this transformation in children who were learning to solve a particular kind of problem, who showed by their reaction times that they had achieved an insight into the solution of the problem, even though their verbal reports indicated that they were not yet consciously aware of their insight. Although intuition is commonly viewed as a source of error in human judgment, it may also be an adaptive component of the creative process – gut feelings that motivate the problem-solver to keep working. In fact, Dijksterhuis and his colleagues have argued that a reliance on intuition, or unconscious thinking, may be especially valuable in situations involving complex decision-making.
The Unconscious beyond Cognition Intentional mental states go beyond cognitive states of perceiving, remembering, knowing, and believing: they also include affective states, such as
Unconscious Cognition
feelings of pleasure and pain, and motivational states, such as the desire to approach or avoid. If cognitive states can be unconscious, perhaps affective and conative states can be unconscious as well. Certainly, conscious feelings and desires can be generated automatically, in response to the appearance of effective stimuli in the environment. In addition, conscious feelings and desires can emerge as expressions of implicit perception, memory, and thought – as in the subliminal mere exposure effect, by which subjects come to prefer stimuli to which they have been exposed, even if they were not consciously aware of the exposures, and cannot consciously remember them. Although most authorities currently hold the view that unconscious feelings and desires are almost contradictions in terms, the general acceptance of the cognitive unconscious suggests that we should be prepared to accept the existence of an affective unconscious, and of a conative unconscious, as well. Evidence in this regard is just beginning to trickle in, and at this point neither unconscious emotion nor unconscious motivation has been as well established as unconscious memory and perception.
Attention and Consciousness Revisited One implication of research on implicit perception, memory, and thought is that consciousness is not as closely tied to attention as the distinction between automatic and controlled processing implies. It may seem obvious that attention is the means by which objects and events are brought into consciousness, and that consciousness entails paying attention to such things. But in the case of implicit perception, subjects may pay close attention to the spatial location where a stimulus will appear, without being aware of the stimulus itself. In implicit memory, subjects may have attended to the study list at the time of learning, but fail to consciously recollect it later. In implicit learning, subjects are attending to the manifest task which has been set for them, while being unaware of knowledge they may have acquired in the course of task performance. And in implicit thought, an insight may emerge into consciousness
419
even though the subject has diverted attention away from the problem he is trying to solve. Attention, then, is neither necessary nor sufficient for consciousness.
The Cognitive Unconscious and Conscious Shyness The discovery of the unconscious is commonly (and erroneously) attributed to Sigmund Freud, and there has been some tendency to claim that the literature on the cognitive unconscious provides support for the psychoanalytic conception of unconscious mental life. This is not the case: the unconscious percepts and memories revealed by contemporary cognitive psychology bear no resemblance to Freud’s primitive, infantile, sexual and aggressive urges. Moreover, the distinction between conscious and unconscious mental life appears to be a matter of cognitive architecture, rather than of repression. Freud’s metapsychological proposition that some aspects of mental life are unconscious is correct, but this is a proposition that is not unique to psychoanalytic theory. The more specific propositions of psychoanalytic theory, to the extent that they are testable at all, have all failed to find empirical support. Still, just as Freud believed that unconscious processes dominated mental life, so some psychologists and other cognitive scientists have taken evidence of unconscious perception, learning, and memory as reason for thinking that conscious mental activity plays little or no role in behavior. For example, some personality and social psychologists have shifted from ‘dual-process’ theories that some processes in social cognition and behavior are consciously controlled, and others unconscious and automatic, to a view that social cognition and behavior is dominated by unconscious, automatic processes, to the virtual exclusion of conscious, controlled ones. Some writers have gone so far as to replace Freud’s view of the primacy of unconscious sexual and aggressive motives with the more modern concept of automaticity. Actually, there is no evidence (such as might be provided by the process-dissociation procedure) supporting the proposition that automaticity dominates the conscious control of experience, thought,
420
Unconscious Cognition
and action. Accordingly, the popularity of the view that it does reflects a persisting ‘shyness’ about consciousness noted by the philosopher Owen Flanagan. At the very least, evidence for automaticity, and for unconscious perception, learning, and memory provide support for conscious inessentialism, which holds that consciousness is not necessary for adaptive cognition and action to occur. Conscious inessentialism, in turn, can lead to epiphenomenalism, or the view that consciousness has no causal function at all. But it should be understood that the literature summarized here only reveals the reality of unconscious cognition. It does not by any means negate consciousness.
Neural and Psychological Correlates of Consciousness The advent of brain-imaging research has added a new dimension to research on unconscious mental life, as researchers search for the neural correlates of conscious and unconscious processing. It seems clear, for example, that the striate cortex is required for conscious vision, but not for the unconscious vision revealed by blindsight. And the hippocampus and other portions of the medial temporal lobe appear to be necessary for conscious remembering, but not for priming and other aspects of unconscious memory. More generally, comparisons of automatic versus conscious processing, and of implicit versus explicit perception, learning, and memory, may provide the comparison conditions necessary to identify the neural correlates of consciousness – which is to say, not the neural correlates of conscious wakefulness (as opposed to general anesthesia or sleep), but rather the neural correlates of conscious acts of perception, learning, memory, and thought. For example, automatic processing appears to activate the sensory and motor representation areas of the brain, such as the primary auditory cortex, while controlled processing is associated with areas of association cortex. Similarly, presentation of a masked visual stimulus activates areas of the visual cortex (naturally), while processing of the same stimulus unmasked, activates a wider network of brain areas including parietal and frontal regions.
At the psychological level of analysis, unconscious priming effects underscore the first-person nature of the intentionality that underlies conscious perception and memory. As James noted in the Principles, every conscious percept, memory, or thought is part of a personal consciousness: ‘‘The universal conscious fact is not ‘feelings exist’ or ‘thoughts exist’ but ‘I think’ and ‘I feel.’ ’’ Similarly, when subjects complete tests of explicit perception and memory, their answers are always couched in the first person: I see, I hear, I recall, or I recognize ‘X.’ By contrast, tests of implicit perception and memory are conducted in the third person: subjects are asked to complete a word stem or fragment, or identify a word, and by and large their answers lack this element of self-reference: that word is ‘X.’ Certainly, the subjects are conscious, and conscious of what they are doing, but in general implicit expressions of perception, memory, learning, and thought lack the integration of the percept, memory, or thought with the self that seems critical for conscious awareness. See also: Implicit Social Cognition; Unconscious Goals and Motivation.
Suggested Readings Bargh JA and Ferguson MJ (2000) Beyond behaviorism: On the automaticity of higher mental processes. Psychological Bulletin 126(6 special issue): 925–945. Dehaene S, Changeux JP, Naccache L, Sackur J, and Sergent C (2006) Conscious, preconscious, and subliminal processing: A testable taxonomy. Trends in Cognitive Sciences 10(5): 204–211. Dijksterhuis A and Nordgren LF (2006) A theory of unconscious thought. Perspectives on Psychological Science 1(2): 95–109. Ellenberger HF (1970) The Discovery of the Unconscious: The History and Evolution of Dynamic Psychiatry. New York: Basic Books. Kihlstrom JF, Barnhardt TM, and Tataryn DJ (1992) Implicit perception. In: Bornstein RF and Pittman TS (eds.) Perception without Awareness: Cognitive, Clinical, and Social Perspectives, pp. 17–54. New York: The Guilford Press. Kihlstrom JF, Mulvaney S, Tobias BA, and Tobis IP (2000) The emotional unconscious. In: Eich E, Kihlstrom JF, Bower GH, Forgas JP, and Niedenthal PM (eds.) Cognition and Emotion, pp. 30–86. New York: Oxford University Press. Kihlstrom JF, Shames VA, and Dorfman J (1996) Intimations of memory and thought. In: Reder LM (ed.) Implicit
Unconscious Cognition Memory and Metacognition, pp. 1–23. Mahwah, NJ: Lawrence Erlbaum Associates. Koch C and Tsuchiya N (2007) Attention and consciousness: Two distinct brain processes. Trends in Cognitive Sciences 11(1): 16–22. Moors A and DeHouwer J (2006) Automaticity: A theoretical and conceptual analysis. Psychological Bulletin 132(2): 297–326. Schacter DL (1987) Implicit memory: History and current status. Journal of Experimental Psychology: Learning, Memory, and Cognition 13: 501–518. Schultheiss OC (2008) Implicit motives. In: John OP, Robins RW, and Pervin LA (eds.) Handbook of
421
Personality: Theory and Research, 3rd edn. New York: Guilford. Stadler MA and Frensch PA (eds.) (1998) Handbook of Implicit Learning. Thousand Oaks, CA: Sage. Underwood G (ed.) (1996) Implicit Cognition. Oxford: Oxford University Press. Weiskrantz L (1997) Consciousness Lost and Found: A Neuropsychological Exploration. Oxford: Oxford University Press. Westen D (1999) The scientific status of unconscious processes: Is Freud really dead? Journal of the American Psychoanalytic Association 47(4): 1061–1106.
Biographical Sketch
John F. Kihlstrom received his PhD in personality and experimental psychopathology from the University of Pennsylvania in 1975, and is currently a professor in the Department of Psychology, University of California, Berkeley, where he also directs the undergraduate program in cognitive science. Kihlstrom previously held faculty positions at Harvard, Wisconsin, Arizona, and Yale. He is perhaps best known for his research on hypnosis, a field in which he has won many awards. His 1987 Science paper on ‘The cognitive unconscious’ is often credited with consolidating the emerging literature on unconscious mental life. He also expanded the concept of implicit memory to perception, thinking, and emotion.
Unconscious Goals and Motivation K C A Bongers and A Dijksterhuis, Radboud University Nijmegen, Nijmegen, the Netherlands ã 2009 Elsevier Inc. All rights reserved.
Glossary Chronic goals – Goals which are chronically accessible through their frequent and committed pursuit. Lexical decision task – Experimental procedure measuring how quickly people classify stimuli as words or nonwords. The underlying idea is that more accessible words are recognized faster than less accessible words. If participants are primed with a stimulus directly before making the lexical decision, they are faster in recognizing words that are semantically related to the prime compared with words that are unrelated to the prime. Subliminal priming – Very brief exposure of stimuli that cannot be perceived consciously but can be perceived unconsciously. Unconscious thought – Object relevant or task relevant cognitive or affective thought processes that occur while conscious attention is directed elsewhere.
Introduction Most behavior is goal directed. We set our alarm clock to get up early, we drink coffee to quench thirst (or to wake up), and we exercise in the spring to look presentable in a bathing suit in the summer. Traditionally, theories about self-regulation have emphasized the role-conscious choice or conscious guidance in goal pursuit. Based on homeostatic processes and cybernetics, self-regulation theories have described a goal as a reference point toward which an individual intends to move (or intends to move away from). A comparator compares the current behavior with the reference point, and if a discrepancy is detected, behavior has to be
stopped, changed, or even reversed. Subsequently, the comparator again compares the altered behavior with the reference point, and this feedback loop continues until discrepancies are no longer detected. When this happens, a goal is reached. The process of goal-directed behavior can be meaningfully divided into various stages. The first stage is the establishment stage. During goal establishment a goal is chosen or selected: People choose the reference point toward which they try to move (or move away from). The second stage is the planning stage, in which people develop strategies to attain the goal. The third stage is the goal-striving stage. During this stage people engage in goal-directed behavior, and progress toward the reference point is monitored. The last stage is the revision stage, in which it is tested whether or not the goal has been attained. In other words, in this stage the comparator compares the behavior with the chosen reference point. If there are discrepancies between the behavior and the reference point, the individual has to decide whether to quit, revise, (temporarily) disengage, or persist in goal-striving. If there are no discrepancies between the behavior and the reference point, the goal has been attained and the feedback loop is left. Traditionally, people have assumed that conscious awareness of the goal is a prerequisite to maneuver successfully through all stages. However, recently it has been proposed that goals can be activated (established) outside awareness and that in fact the whole process of goal-pursuit can ensue without conscious awareness of the goal. The idea that goals can be activated automatically is based on two ideas. First, goals are mentally represented as desired states in a hierarchically ordered knowledge structure. Such hierarchical knowledge structures include desired states, actions, and means to reach the desired states. Hence, goals can be activated automatically just like other mental representations such as semantic concepts and stereotypes. Second, consciously chosen goals can
423
424
Unconscious Goals and Motivation
develop the capacity of becoming unconsciously activated. When goals are activated consistently and repeatedly in a specific situation, these goals will be activated automatically whenever the person encounters that same situation. In this article, we describe for each of the stages in the process of goal pursuit whether conscious attention is necessary or whether it may run entirely without conscious intervention.
Goal Establishment During goal establishment a goal is chosen or selected. Goals can originate internally as well as externally. Goals that originate externally are assigned goals, which have to be accepted by the individual. Whether a goal is accepted or not depends on the evaluation processes that are invoked. If people evaluate a goal positively, they may commit to the goal. When a goal is accepted, it is translated into an internally meaningful representation that is connected to other goals in the goal hierarchy. If people evaluate the goal negatively they will not accept the goal and do not commit to it. Goals that originate internally are invoked when goals are not met. This can occur as a top–down process as well as a bottom–up process. According to a top–down process, a chronically unmet higher-level goal is detected, the establishment process for a new middle-level goal is activated, and lower-level goals are selected. For example, if the higher-level goal to achieve academically is not met because one failed an exam, the goal to pass the retest will be activated, and hence, the lower-level goal to study will be selected. According to a bottom–up process, people experience affective responses to inputs. These inputs are associated with higher-level goals as either a threat or an opportunity. Then a goal representation is established, and a current state that can potentially match the goal is assessed. Imagine, for example, that you are telling a joke, but nobody laughs. This will be associated as a threat with the higher-level goal to affiliate; hence, the goal representation of affiliation is activated. The question we want to address here is whether goals can be established without conscious awareness. As described earlier, people commit to assigned
goals only if they evaluate that goal positively. There is considerable evidence (for instance from the literature on evaluative conditioning) that the valence of stimuli can be changed without awareness. If neutral stimuli are paired with positive stimuli, these neutral stimuli become more positive, whereas if neutral stimuli are paired with negative stimuli, these neutral stimuli become more negative. Such effects occur even if these stimuli are presented subliminally. Although it has been demonstrated that the valence of stimuli can be changed unconsciously, this does not yet show that people will commit to goals that were created this way. That is, although people may evaluate an activity positively, it does not automatically mean that they want to engage in that activity. For instance, being very positive about running a marathon, does not necessarily mean that you really want to join the next marathon. However, recent research disentangled these ‘liking’ effects from ‘wanting’ effects. In some clever experiments, neutral behavioral states (i.e., activities such as doing puzzles or going for a walk) were paired unconsciously with positive, neutral, or negative words, and subsequently, participants were asked to evaluate these activities and to indicate whether they wanted to engage in these activities. Consistent with the evaluative conditioning literature, activities paired with positive words were evaluated more positively and activities paired with negative words were evaluated more negatively, compared with activities paired with neutral words. More importantly, wanting was higher for activities that were paired with positive words, compared with activities that were paired with neutral and negative words, but there was no difference in wanting between activities that were paired with neutral words and activities that were paired with negative words. These data demonstrate that goals can be established unconsciously. If neutral behavioral states are paired unconsciously with positive stimuli, people evaluate these states more positively and they indicate that they want to attain these states, showing that they commit to the goal. As mentioned earlier, goals that originate internally are invoked when goals are not met. Detecting a discrepancy between a current state and a chronic goal state will lead to the activation of
Unconscious Goals and Motivation
middle-level goals and a selection of lower-level goals. Recent research has demonstrated that this can occur automatically as well. In one of the relevant experiments, participants were presented with 25 sentences that were immediately followed by probe words. In one condition, five of these sentences described actual states or situations that were discrepant with the goal to look wellgroomed (i.e., ‘‘The shoes you put on look dirty.’’), and in the other condition these five sentences were nondiscrepant with that goal (i.e., ‘‘The shoes you put on have laces.’’). After presenting each sentence, participants were asked to indicate whether the probe word appeared in that sentence. The probe word referred to an applicable instrumental action (i.e., to polish) that never appeared in the sentence. If people automatically activate goaldirected actions upon reading goal-discrepant sentences, the reaction time to indicate that the probe word did not appear in the sentence will be higher. The results showed that only for people who frequently pursue the goal, there was a difference in reaction time for indicating whether a probe word appeared in the sentence between discrepant and nondiscrepant sentences. These data demonstrated that detecting a discrepancy between a current state and a chronic goal state will automatically lead to the activation of middle-level goals. To recapitulate, we can conclude that conscious attention or awareness is not necessarily required in the establishment stage of goal-directed behavior. This accounts for goals that originate externally as well as goals that originate internally.
Planning The planning stage refers to the development of strategies to attain a goal. Once a goal has been established, alternative behavioral paths by which a goal can be attained must be developed, evaluated, and selected. Setting the goal to live healthy, for instance, can be achieved by quitting smoking, by exercising more often, by eating more healthy food and less unhealthy food, by reducing alcohol intake, etc., or by any combination of these. As alluded to above, goals are mentally represented as desired states in a hierarchically ordered knowledge structure. Such hierarchical knowledge
425
structures include goals (or higher-level goals) and means to reach these goals (or lower-level goals), sometimes with subgoals (or middle-level goals) in between. Higher-level goals are connected with its middle and lower-level goals with facilitative links. As a consequence, the activation of a higher-level goal will spread out to its middle and lowerlevel goals, resulting in higher accessibility of all middle and lower-level goals. However, for successful goal pursuit people need to select some lower-level goals. Imagine, for example, that you have decided to live a more healthy life. Trying to attain this goal by pursuing all lower-level goals described above simultaneously is extremely difficult, if not impossible. Therefore, people need to weigh the various possibilities and potential consequences and select which lower-level goals they want to pursue in order to attain the higherorder goal. The establishment of higher-level goals can be called goal intentions. If people formulate goal intentions, they specify a particular desired end state (i.e., ‘‘I intend to be healthy’’). One way of selecting middle or lower-level goals is by forming implementation intentions. With implementation intentions, anticipated situations are associated with goal-directed behaviors and they specify the when, where, and how of goal-directed behavior (i.e., ‘‘If my colleague offers me a cigarette after lunch, I will reject that offer’’). Research has shown that actively forming implementation intentions can double the percentage of successful goalpursuit. In one of the experiments, for example, participants (students) were asked to write a report during Christmas break. Half of the participants were asked to form implementation intentions by specifying the when and where of their goaldirected behavior, whereas the other half was were not asked to form implementation intentions. Two-third of the participants who formed implementation intentions pursued their goal successfully compared with one-third of the participants who did not form these intentions. The most important reason for the success of implementation intentions is that goal pursuit is connected to situations rather than to the person. If people specify what they want to do in a particular situation, they delegate control to the environment. Therefore, the specified situation is
426
Unconscious Goals and Motivation
capable of eliciting goal-directed behavior immediately, efficiently, and without conscious intent. In other words, connecting your intentions to environmental cues will lead to higher accessibility of your intentions when encountering that specific situation, resulting in automatic goal-directed responses. Although implementation intentions lead to automatic goal-directed behavior, the forming of implementation intentions itself is generally a conscious and effortful process. Therefore, we cannot yet conclude that conscious attention is not necessary in the planning stage. If people consistently and repeatedly choose the same lower-order goal to attain a higher-order goal (such as always taking the stairs in order to stay healthy), activation of this lower-order goal can become automatic. This happens, for instance, in habitual behavior. If people (almost) always go to the university by bicycle, the cognitive association between ‘bike’ and ‘university’ becomes very strong. As a consequence, the activation of the goal to go to university may automatically lead to the activation of the associated action or habit of cycling. Evidence for this can be found in research in which habitual and nonhabitual bicycle users were primed or not primed with travel goals (e.g., having to attend lectures). Participants were asked to respond to the word bicycle after being presented with locations that corresponded to the earlieractivated travel goal (e.g., university). The results showed that habitual bicycle users responded faster to the word bicycle than nonhabitual bicycle users, but only when the travel goal was activated. These data demonstrate that only habitual bicycle users automatically activate the concept of bicycle upon activating the higher-level goal (going to university to attend lectures). That is, habitual bicycle users did not need to select the lower-order goal consciously when the higher-order goal was activated. More evidence can be found in research showing that the suppression of habitual responses is more difficult under conditions of cognitive load than the suppression of nonhabitual responses. In relevant research, different locations were presented to participants and half of the participants were asked to respond with the travel mode they would normally use for the presented destinations, whereas the other half were asked to respond with
an alternative travel mode (i.e., they were not permitted to respond with the travel mode they would normally use; in the Netherlands the habitual way of traveling for most students is cycling). Half of the participants performed the task under cognitive load and the other half performed the task without cognitive load. The results showed that when participants were not allowed to respond with bicycle the proportion of bicycle responses in the habitual trials was higher in the load condition compared with the no load condition, whereas in the nonhabitual trials there was no such difference. These data showed that habitual responses are automatically triggered by higher-order goals (i.e., suppression of these responses was difficult under cognitive load), whereas this does not pertain to nonhabitual responses. Based on the experiments described earlier, we can conclude that conscious attention is not always necessary in the planning stage. When people consistently and frequently use the same lower-level goal in order to attain one higherlevel goal, as is the case in habitual behavior, this lower-level goal is activated upon the higher-level goal, leading to a higher probability of pursuing that goal. As described earlier, accessibility alone is not always enough to lead to more ‘wanting.’ Positive affect plays an important role as well. In another line of research, it was investigated whether people automatically evaluate objects or means (i.e., lowerlevel goals) that are useful for current goal pursuit (i.e., an activated higher-level goal) positively, leading to more ‘wanting’ to attain or pursue these means or activities. In one of the experiments, half of the participants were made very thirsty, whereas the other half were allowed to quench their thirst. Subsequently, participants’ automatic evaluations of different objects were measured with an implicit evaluation measure (i.e., a sequential evaluative priming paradigm). The objects were highly relevant to goal-attainment (i.e., water), weakly relevant (i.e., coffee), indirectly relevant (cup), or irrelevant (tree). The findings showed that there was no difference in automatic evaluation of the weakly relevant objects, the indirectly relevant objects, and the irrelevant objects between thirsty and nonthirsty participants. However, thirsty participants evaluated the highly relevant objects more positively
Unconscious Goals and Motivation
than nonthirsty participants. These data demonstrate that people automatically evaluate means (low-level goals) that are highly relevant for current goal pursuit (activated high-level goal) positively. Together, these findings demonstrate that no conscious attention is necessary for selecting lowerlevel goals to attain higher-level goals, meaning that no conscious attention is necessary in the planning stage.
Goal Striving and Monitoring During the goal-striving stage, an individual engages in goal-directed behavior and the progress toward goal-attainment is monitored. The discussion on goal-striving is subdivided into two sections. In the first section, we focus on goal-striving in general, and in the second section, we focus on monitoring processes. In the first section, we review evidence showing that people engage in goal-directed behavior without being aware of it. We start describing research concerning unconscious activation of goals and its effects on goal-directed behavior. We continue with research showing that people are sometimes better in their goal pursuit if they engage in it unconsciously rather than consciously. In the second section, we have a look at monitoring processes. We start reviewing evidence demonstrating that people monitor their behavior without conscious attention. We continue this section reviewing research concerning automatic inhibition of alternatives and temptations and we finish this section with research examining what happens when goal pursuit is obstructed.
Goal Striving Converging evidence has shown that goals can be activated and run to completion without the person being aware at any point. Imagine, for instance, that you buy a starbucks coffee every day you are driving to your work. Over time, the goal of buying a starbucks coffee can be activated and performed automatically, as is demonstrated by finding yourself enjoying your starbucks coffee in the car without even noticing that you decided to buy and indeed bought the coffee that day.
427
One of the first studies on the unconscious activation of goals was conducted in the beginning of this century. In these studies, an achievement goal was activated by priming half of the participants subliminally with words related to achievement. After the priming procedure, these participants performed better on an achievement task than participants who were not exposed to these subliminal primes. To ensure that people were really engaging in goal-directed behavior, the authors designed three additional experiments investigating whether the behavior contained motivational aspects as well. The data demonstrated that the effects of goal priming increased over time, that participants primed with an achievement goal persisted in the face of obstacles, and that participants primed with an achievement goal were more likely to return to an incomplete achievement task after interruption than participants not primed with that goal. Since these first set of studies, a lot of research has shown that goals can be activated outside of awareness. There are many ways in which goals can be activated unconsciously. One such way is through goal contagion. Individuals can automatically adopt and pursue a goal that is implied by another person’s behavior. For instance, in one of the experiments on goal contagion, half of the participants read a story implying the goal of earning money and the other half read a control story. Then, participants were told that they had to perform one more task and if there was sufficient time left they could perform a final task in which they could earn some money. Participants who read the story implying the goal of earning money were faster during the intervening task than participants who read the control story, presumably because they were motivated to quickly start with task with which they could earn money. Another way goals can be activated unconsciously is by significant others. If a significant other has a goal in mind for someone, then this goal can be activated whenever that person thinks about the significant other and even when the person is merely primed with that significant other. For example, research has shown that priming participants subliminally with words related to their father led to a higher commitment, more persistence, and a better performance on a task assessing analytic reasoning. This effect occurred especially when participants
428
Unconscious Goals and Motivation
were close to their father and when their father valued that particular goal. Interestingly, there is also evidence that goals can be activated unconsciously by scent. In a recent experiment, participants were asked to fill out a filler questionnaire. Some participants were seated in a room where a bucket was hidden behind a cupboard. The bucket contained water with citrusscented all-purpose cleaner. Subsequently, participants were asked to move to another room and were instructed to eat a very crumbly biscuit. Participants who first filled out the questionnaire in a room with the scent of all-purpose cleaner removed the crumbs substantially more often than participants who completed the questionnaire in a room without the scent of all-purpose cleaner. In the experiments described above, the authors asked their participants whether they had noticed something special or something strange during the experiments. It was demonstrated that participants were not aware of the priming manipulation, nor were they aware of the influence of the priming manipulation on their behavior. Thus far, this section only focused on the effects of unconscious activation of goals on goal-directed behavior. Although people were not aware of the origin of their behavior (i.e., the goals were activated unconsciously), the behavior will have been conscious. For instance, people who were unconsciously primed with an achievement goal were not aware of the activation of the goal, but they may have been aware that they were trying to do their best on the achievement task. However, recently it has been demonstrated that people can engage in goal pursuit unconsciously, and interestingly, this research showed that people sometimes perform even better when pursuing their goals unconsciously rather than consciously. So far, this research has mainly focused on decision making. If people have the goal to make a sound decision, they need to process information concerning this decision. According to unconscious thought theory, people can process the information consciously as well as unconsciously. In one of the experiments in this line of research, participants were given information about four apartments that were each described with 12 aspects. One apartment had more positive aspects than negative aspects (the best apartment), and one apartment had more negative aspects than
positive ones (the worst apartment). The other two apartments were more neutral and had an equal number of positive and negative attributes. All 48 pieces of information were randomly presented on the computer screen. The goal of the participants was to evaluate each apartment. One-third of the participants were not allowed to think about the decision and had to decide immediately after all information was presented. Another one-third of the participants were allowed to think consciously about the decision for 3 min after all the information was presented. The last one-third of the participants were able to think about the decision unconsciously for 3 min. After all information was presented, participants in this condition were distracted for 3 min with a cognitive load task. As a consequence, these participants were not able to think about the decision consciously, but were able to process it unconsciously during the cognitive load task. The data showed that only participants who thought about the apartments unconsciously evaluated the best apartment more positively than the worst apartment. Participants who thought about it consciously and those who had to decide immediately did not show this preference for the best apartment. The reason for why people who thought about the decision unconsciously performed better than people who thought about it consciously is that these latter were simply not able to correctly process all the information about the four apartments. Especially, when things are very complex (i.e., the decision is difficult or requires the processing of a lot of information), it is better to process it unconsciously rather than consciously. However, one problem of the experiment described above is that you cannot rule out that people in the unconscious thought condition made a better decision because they were distracted and not necessarily because they thought about it unconsciously. That is, it may be the case that these people made a better decision because they could look at the problem with a so-called fresh look. If that were the case, there is no evidence that people in the unconscious thought condition engage in goal pursuit. Fortunately, in another experiment this alternative explanation was ruled out. This experiment was quite similar as the experiment described above. However, in this experiment, participants were given information about four cars
Unconscious Goals and Motivation
that were each described with 12 aspects. One car had more positive aspects than negative aspects (best car) and one car had more negative aspects than positive aspects (worst car). The other two cars had an equal number of positive and negative aspects. All 48 pieces of information were randomly presented on the computer screen. Then, participants were assigned to one of the three conditions. Two conditions, the unconscious thought condition and the conscious thought condition, were similar as in the experiment described above. In these two conditions the goal of the participants was to evaluate each car. Participants in the conscious thought condition were allowed to think about the decision for 4 min and participants in the unconscious thought condition were able to think about the decision unconsciously for 4 min. In the third condition, the mere distraction condition, participants were also distracted for 4 min, just as in the unconscious thought condition. However, these participants did not have the goal to evaluate each car. The data again showed that participants in the unconscious thought condition outperformed participants in the conscious thought condition in distinguishing between the best and the worst car. Moreover, participants in the unconscious thought condition outperformed participants in the mere distraction condition as well. These findings rule out the alternative explanation that people in the unconscious thought condition outperform those in the conscious thought condition purely because they were distracted for a while and then continued with the decision problem with a fresh look. Based on the experiments described above, we can conclude that people can engage in goaldirected behavior without being aware of it. Goals can be activated unconsciously and subsequently affect goal-directed behavior. Furthermore, people are not necessarily aware of their goal-directed behavior. That is, they can process information unconsciously in order to make a good decision. And interestingly, this decision often is even better than after conscious thought. Monitoring A second characteristic of the goal-striving stage is that progress toward goal-attainment is monitored. This monitoring process serves to increase the
429
likelihood of goal-attainment by scanning the environment for threats and opportunities. If an opportunity for goal pursuit is detected, the monitoring process increases the salience of this opportunity by heightening its accessibility or by signaling the person that an opportunity is detected. If a threat (i.e., alternative goals or temptations) is detected, the monitoring process inhibits this threat by decreasing its accessibility. If inhibition failed, for instance if the threat is too strong or cannot be avoided, the monitoring process signals the person that goal pursuit might be under threat. In this section, we report some examples of unconscious monitoring processes. One famous example of unconscious monitoring is the cocktail-party effect. This effect is based on the idea that people have a chronic goal to process self-relevant information. Imagine you are talking to someone at a noisy party. You consciously focus on that conversation and the noise of others at that party is somehow filtered out and is consigned to background babble. However, if someone in your environment (not your conversation partner) speaks out your name, then your name will capture your attention, even when you did not hear anything of that conversation. So, while you are talking to someone, other conversations are being processed unconsciously. And as soon as something relevant or something important is said (i.e., goal-related or self-relevant information), this unconscious monitoring process brings that information into your attention and into your consciousness. Other examples of unconscious monitoring processes can be found in the literature concerning suppression. When people are not allowed to think about something, the ‘forbidden’ thought is, ironically, more likely to pop into consciousness. For instance, when participants are asked not to think about white bears (i.e., they are asked to suppress these thoughts) for a period of time, after that period they report more thoughts about white bears than participants who were not asked to suppress thoughts about white bears in the first place. The reasoning behind these findings is that for participants who were not allowed to think about white bears, a monitoring process starts to search for failures not to think about white bears. In other words, a monitoring process keeps track of
430
Unconscious Goals and Motivation
whether someone is still engaging in goal pursuit, and threats are signaled to the person. The two examples described above show that the monitoring process indeed signals the person that opportunities (i.e., self-relevant information in your environment) or threats (i.e., ‘forbidden’ thoughts) are detected in the environment, without the person being aware of this monitoring process. Moreover, there is also evidence that threats to current goal pursuit are inhibited automatically. In recent research, participants were asked to provide three positive attributes they want to possess (i.e., current goals) and one positive attribute they do not want to posses (i.e., nongoal). Participants continued with a modified lexical decision task in which their three current goals and neutral control words were primed unconsciously. After these primes, the target words were presented. These target words were again participants’ three current goals, the nongoal and two nonattributes. Participants were asked to indicate as quickly as possible whether the target word was a personal attribute or not. The results showed that the responses to their current goals slowed down when the preceding prime was one of the other two current goals compared with when the preceding prime was a nonattribute. There was no difference in response time when the target word was a nongoal. These data indicate that the activation of a current goal results in inhibition of the alternative current goals. In another line of research, it was investigated whether the activation of a goal will also lead to inhibition of temptations. In that research, participants were given a lexical decision task to measure accessibility of goals and temptations. In the lexical decision task words related to religious goals (i.e., the Bible) and temptation- or sin-related words (i.e., sex) were used as target words among other neutral filler words and filler nonwords. These target words were preceded by goal-related words and temptation-related words in the relevant trials and neutral words in the irrelevant trials, resulting in four combinations: goals as target words preceded by relevant temptation primes; goals as target words preceded by irrelevant neutral primes; temptation primes preceded by relevant goal primes; and temptation primes preceded by irrelevant neutral primes. The data showed that participants were faster in recognizing goal-related
words when preceded by relevant temptationrelated words compared with goal-related words that were preceded by neutral words. Furthermore, participants were slower in recognizing temptation-related words when preceded by relevant goal-related words compared with temptationrelated words preceded by neutral words. These findings demonstrate that people automatically activate relevant goals upon activation of the temptation, whereas they automatically inhibit relevant temptations upon activation of the goal. In other words, the monitoring process increased the salience of an opportunity to attain a goal (by heightening the accessibility of the goal when a temptation was activated) and decreased the salience of a threat to goal-attainment (by lowering the accessibility of the temptation when a goal was activated). In sum, monitoring can occur outside of awareness. If opportunities for successful goal pursuit are detected in the environment, these opportunities increase in accessibility, whereas if threats for successful goal pursuit are detected, these threats decrease in accessibility. Taken together, there is ample evidence that conscious attention is not necessary in the goal-striving stage.
Revision In the revision stage, people test whether they have attained the activated goal or not. To do so, people test their present behavior against the selected goal they want to achieve. If no discrepancies are detected no further action is necessary. However, if discrepancies are detected between the present behavior and the selected goal, people have to decide whether to quit, revise, temporarily disengage, or persist in goal-striving. What people decide to do will depend on many different factors. Here, we name only a few. Expectancies people have about attaining the goal, for instance, may play an important role. If people expect that they can still attain the goal, in other words when expectancies are high, they will most likely persist in goal-striving. On the other hand, if people expect that they will not attain the goal, that is, when expectancies are low, they will most likely quit. Another factor that may influence whether people decide to quit, revise, temporarily disengage, or
Unconscious Goals and Motivation
persist in goal-striving is the difficulty of attaining the goal. Of course this factor is closely related to the factor described earlier. However, the factors are not necessarily identical. Finishing a 5-star sudoku puzzle, for example, might be very difficult to do, but for experts in making sudoku puzzles, expectancies for solving that puzzle will be high. On the other hand, although attaining a goal might be easy, it is not necessarily easy for you to attain. Walking to the nearby supermarket is usually very easy, but not if you just sprained your ankle. If goals are difficult to attain, people may be less motivated to attain the goal and quit pursuing the goal, whereas if goals are easy to attain, people may be more motivated to attain the goal and persist in their goal-striving. Besides these motivational effects, discrepancy detection also has effects on feelings and selfesteem. If people observe discrepancies between their behavior and their goals, a negative evaluation ensues, leading to a negative mood and lower self-esteem, whereas observing no discrepancies will lead to a positive evaluation, resulting in a positive mood and higher self-esteem. Recently, it has been investigated whether success and failure to attain unconsciously activated goals may affect mood states. Do people have a negative mood after failing to attain an unconsciously activated goal and a positive mood after succeeding to attain an unconsciously activated goal? In one of the experiments, participants were unconsciously primed with an achievement goal or not. Subsequently, they were given an easy anagram task to manipulate success or a difficult anagram task to manipulate failure. It was hypothesized that participants in the success condition would not detect any discrepancy between their behavior and their unconsciously activated goal (which would then lead to a positive mood), whereas participants in the failure condition would detect a discrepancy between their behavior and their unconsciously activated goal (which would lead to a negative mood). Furthermore, this effect was only expected for participants who were primed unconsciously with an achievement goal and not for participants who were not primed with that goal. After all, participants who were not primed with an achievement goal would not evaluate their behavior, and therefore would not engage in discrepancy detection. The findings were consistent with these expectations.
431
In another line of research, the experiment explained above was replicated with self-esteem rather than mood as the dependent variable. Consistent with the experiment described above, it was found that participants in the success conditions reported higher self-esteem than participants in the failure condition. And again, these effects only emerged for participants who were primed unconsciously with an achievement goal and not for participants who were not primed with that goal. These two experiments support the idea that people are able to discover whether goal pursuit operates smoothly or not even when these goals were activated unconsciously. However, mood and self-esteem were measured explicitly, leaving open the question whether this discovery occurs outside of awareness as well. Imagine you are teaching your students and your lecture meets with disaster. You cannot explain the major issues of the lecture and your students do not seem interested at all! Students are not attending and most of them start their own little conversation and those who keep quiet even fell asleep. Immediately after this horrible fiasco, thoughts about the goal you actually pursued, the goal of making an excellent impression on your students spontaneously keeps popping into consciousness. Recently, it has been investigated whether people become aware of unconsciously activated goals if goal pursuit is threatened. Based on literature concerning ruminative thoughts, it was hypothesized that people who are not able to attain their unconsciously activated goals will start to think about it consciously. If people are not able to attain consciously selected goals, they often ruminate about these unattained goals. Such rumination is motivationally driven and thoughts concerning the unattained goal may enter consciousness repeatedly and unintentionally. Furthermore, these thoughts are likely to keep intruding into consciousness until the goal is either attained or abandoned. To investigate whether people also start to think consciously about unconsciously activated goals when goal pursuit is frustrated, several experiments were conducted. In one experiment, half of the participants were primed with an achievement goal and the other half were not. Subsequently, they were given either an easy version of the remote associates test to manipulate success or a
432
Unconscious Goals and Motivation
difficult version to manipulate failure. During the test, participants were encouraged to say everything they were thinking about out loud, regardless of the relation to the task. Of interest was whether participants would report thoughts about the goal to achieve during the test. If so, it means that participants become aware of the unconsciously activated goal. It was expected that only participants who failed to attain their unconsciously activated goal would become aware of their goal pursuit. The findings were consistent with this prediction. Participants who were primed with an achievement goal and who failed to attain that goal reported more goal-related thoughts than participants in all other conditions. This experiment shows that participants become aware of their unconscious goals if goal pursuit is frustrated, but not if goal pursuit runs smoothly.
Summary We investigated whether consciousness is necessary in each of the stages of goal pursuit or, alternatively, whether all these stages can run completely outside of awareness. The stages of goal pursuit that can be identified are goal establishment, planning, goal striving, and revision. The first three stages of goal pursuit can indeed occur without conscious attention. In the revision stage, however, some conscious attention may be necessary. In this stage people test whether their behavior is in line with their current goal pursuit, and if discrepancies are detected, people need to revise their goals. Although conscious attention is not necessary for discrepancy detection, there is no evidence that people can revise their goals unconsciously. See also: Unconscious Cognition.
Suggested Readings Aarts H and Dijksterhuis A (2000) Habits as knowledge structures: Automaticity in goal-directed behavior. Journal of Personality and Social Psychology 78(1): 53–63. Aarts H, Gollwitzer PM, and Hassin RR (2004) Goal contagion: Perceiving is for pursuing. Journal of Personality and Social Psychology 87(1): 23–37.
Austin JT and Vancouver JB (1996) Goal constructs in psychology: Structure, process, and content. Psychological Bulletin 120(30): 338–375. Bargh JA, Gollwitzer PM, Lee-Chai A, Barndollar K, and Tro¨tschel R (2001) The automated will: Nonconscious activation and pursuit of behavioral goals. Journal of Personality and Social Psychology 81(6): 1014–1027. Bongers KCA and Dijksterhuis A (2009) Consciousness as a trouble shooting device? The role of consciousness in goal-pursuit. In: Morsella E, Bargh JA, and Gollwitzer P (eds.) The Oxford Handbook of Human Action, pp. 589–604. New York: Oxford University Press, Inc. Bos MW, Dijksterhuis A, and Van Baaren RB (2008) On the goal-dependency of unconscious thought. Journal of Experimental Social Psychology 44(4): 1114–1120. doi:10.1016/j.jesp.2008.01.001. Brandstaetter V, Lengfelder A, and Gollwitzer PM (2001) Implementation intentions and efficient action initiation. Journal of Personality and Social Psychology 81(5): 946–960. Chartrand TJ and Bargh JA (2002) Nonconscious motivations: Their activation, operation, and consequences. In: Tesser A, Stapel DA, and Wood JV (eds.) Self and Motivation: Emerging Psychological Perspectives, pp. 13–41. Washington, DC: American Psychological Association. Custers R and Aarts H (2005) Positive affect as implicit motivator: On the nonconscious operation of behavioral goals. Journal of Personality and Social Psychology 89(2): 129–142. Custers R and Aarts H (2007) Goal-discrepant situations prime goal-directed actions if goals are temporarily or chronically accessible. Personality and Social Psychology Bulletin 33(5): 623–633. Dijksterhuis A and Nordgren LF (2006) A theory of unconscious thought. Perspectives on Psychological Science 1(2): 95–109. Ferguson MJ and Bargh JA (2004) Liking is for doing: The effects of goal pursuit on automatic evaluation. Journal of Personality and Social Psychology 87(5): 557–572. Fishbach A, Friedman RS, and Kruglanski AW (2003) Leading us not unto temptation: Momentary allurements elicit overriding goal activation. Journal of Personality and Social Psychology 84(2): 296–309. Holland RW, Hendriks M, and Aarts H (2005) Smells like clean spirit: Nonconscious effects of scent on cognition and behavior. Psychological Science 16(9): 689–693. Kruglanski AW, Shah JY, Fishbach A, Friedman RS, Young Chun W, and Sleeth-Keppler D (2002) A theory of goal systems. In: Zanna MP (ed.) Advances in Experimental Social Psychology, 34, pp. 331–378. San Diego: Academic Press, Inc. Shah JY (2002) Forgetting all else: On the antecedents and consequences of goal shielding. Journal of Personality and Social Psychology 83(6): 1261–1280. Shah JY (2003) Automatic for the people: How representations of significant others implicitly affect goal pursuit. Journal of Personality and Social Psychology 84(4): 661–681. Wegner DM (1994) Ironic processes of mental control. Psychological Review 101(1): 34–52.
Unconscious Goals and Motivation
433
Biographical Sketch
Karin Bongers is a postdoctoral researcher at the Radboud University Nijmegen. Her main research interests center on selfregulation, motivation, and goal-directed behavior. Specifically, she has focused on consequences of unconscious goal-pursuit failure for self-esteem and self-protecting mechanisms. Furthermore, she is interested in the interplay between conscious and unconscious processes. Her research has focused on how and when unconscious processes appear in consciousness. She has recently started to investigate the underlying mechanisms in unconscious thought in the domain of creativity.
Ap Dijksterhuis is professor of psychology at the Radboud University Nijmegen. His laboratory is focused on unconscious processes and on the role these processes play in social behavior, decision making, and creativity. In 2006, he published the Unconscious Thought Theory, a theory that posits that unconscious processes greatly help us to solve problems and that decisions made after a period of unconscious activity are sometimes better than decisions made after conscious deliberation. He won various awards, including the APA Early Career Award and the SESP Career Trajectory Award.
Visual Experience and Immediate Memory J T Enns, University of British Columbia, Vancouver, BC, Canada ã 2009 Elsevier Inc. All rights reserved.
Glossary Attention – Any influence causing some information to be selected for more detailed processing at the expense of other information. Cartoon advantage – Improved recognition for pictures exaggerating distinctive features of a familiar object. Change blindness – Being insensitive to a difference between two images. Comparison failure – Failure to actively compare one region in a scene with its counterpart in another scene. Constraint – A physical regularity of the world in which our visual systems have evolved, for example, gravity. Convexity bias – A tendency to assume, in the absence of contradictory information, that shapes with bulges are nearer to you than shapes with indentations. Edge alignment – A tendency to assume, in the absence of contradictory information, that aligned edges belong to the same surface. Ganglion cells – Neurons in the eye that are organized in their circuitry to enhance edge detection. Heuristic – A rule of thumb that relates image interpretation to an expected constraint of the visual world Lateral inhibition – Circuitry found in ganglion cells that enhances edge detection for one ganglion cell by inhibiting neighboring ganglion cells as a direct function of its degree of activation. Own-race advantage (equivalently, the cross-race disadvantage) – Improved ability to learn the faces of previously unfamiliar individuals from one’s own race. Receptive field – A description of the behavior of a sensory neuron in terms of a filter or template.
Subjective edge (equivalently, virtual edge) – The appearance of a faint edge at a surface boundary in the absence of any physical evidence for such an edge.
Introduction Healthy people with fully functioning visual systems take many things for granted. But what they may take most for granted, even more than the air they breathe, is that the world they experience through vision is vivid in color, rich in texture, and detailed in shape. There is no longer any doubt that this experience is created by the neurons in our brains. However, how that experience is created by our neurons is still largely a mystery. This article will review what scientists from a number of related academic disciplines currently understand about how visual experience is constructed by our mind when it confronts the pattern of light that reach our eyes. To help the reader understand some of the questions that are still unanswered in this process, the review will be organized around three primary puzzles that everyone faces when they try to understand how vision works. Each puzzle is in fact a ‘paradox,’ because each one involves strong scientific evidence about our visual experience, on the one hand, that flies in the face of our common sense beliefs, on the other hand. These common sense beliefs are themselves presumably based on our own experiences, which only deepens the mystery.
Paradox One: Rich Experience from Impoverished Input When we see a cat through a picket fence we have no doubt that we are experiencing the cat as an entire animate object, even though the light 435
436
Visual Experience and Immediate Memory
reflecting from the cat is only visible to us through the openings between the pickets. Moreover, we experience the cat as a solid object with volume, even though the individual light patterns between the pickets is in reality only two-dimensional at any point in time. And over time, as we move past the picket fence, or the cat moves behind it, we experience a unique cat of a given size, with a specific color pattern, and with a unique posture, even though the information from between the pickets is constantly changing. As this example shows, our visual brain is able to take fragmentary, ever-changing patterns of two-dimensional light and transform them into an experience of individual objects interacting with one another in a stable world. At no point in this process do we gain insight from our own experiences of the seen world about how our brain accomplishes this transformation. We simply have the experience of seeing the world, out there, and for the most part, as it really is. Our common sense understanding of vision, based on these experiences, thus vastly underestimates the real difficulty of the problem of converting fragmentary two-dimensional patterns of light into holistic three-dimensional visual experiences. An understanding of that only comes from treating other humans (and often even ourselves) as temporary objects of scientific interest. When we adopt this stance of the experimenter, we are able to conduct experiments on the transformation of light to experience by observing how human awareness and behavior (considered as the output of the scientific object) are related to the physical patterns of light (considered as the input to the object) that give rise to these experiences in us. Experiments studying the relationship between the physical conditions of vision and the associated experiences in humans are traditionally referred as ‘psychophysics.’ This branch of science, also known as experimental psychology, had its beginning over 100 years ago in the laboratories of Hermann Helmholtz, Wilhelm Wundt, and Gustav Fechner. In the twenty-first century, researchers who conduct psychophysical experiments often refer to themselves as ‘cognitive neuroscientists,’ if they have a special interest in the brain mechanisms associated with the experiences they are studying, ‘cognitive scientists,’ if they are interested in
generalizing what they know about the problem of extracting information from light to nonbiological systems, or ‘vision scientists,’ if they are interested in comparing how different visual systems, both biological and nonbiological, solve the problem of creating experience from light.
Edge enhancement Psychophysical experiments over the past century have revealed that the tendency for the visual brain to go beyond the information given applies equally well to its very earliest and simplest processes as it does to the highest levels of visual cognition and imagination. Consider one of the lowest level functions performed by the eye, that of registering the existence of an ‘edge’ (the difference between light and dark) at a particular location in an image. The daylight neural receptors in the eye, known as ‘cones,’ do not merely record edges as faithfully as they can. Rather, they are organized to deliberately exaggerate the contrast of edges they encounter, through a system of circuitry known as ‘lateral inhibition.’ The net effect of this circuitry is to highlight the existence of edges that are detected, so as to help bridge and join edges in regions of the image in which the lighting conditions make their detection difficult. The ‘Hermann grid’ shown in Figure 1(a) provides first-hand experience of the effects of the lateral inhibitory circuitry in the eye. Notice that when you first glance at the grid, you see gray smudges at most of the intersections. However, when you fix your gaze directly on one of the intersections, the gray smudge you previously saw out of
– – – –+ –– + ––– +++++ –– – – ––
– – – –+ –– + ––– +++++ –– – – ––
(a)
(b)
Figure 1 Hermann grid (a) and the differential effect of
intersections vs. spaces between columns on the receptive fields of ganglion cells in the eye (b). Illusory smudges are seen at the intersections of the grid because receptive fields centered there receive more light falling onto their inhibitory surrounds.
Visual Experience and Immediate Memory
the corner of your eye disappears. The gray smudges seem to be present only at the intersections that are not currently at the center of your gaze. Figure 1(b) illustrates schematically how the circuitry of the eye causes these smudges to appear in our experience. Many ‘cones’ feed their information into a smaller number of ‘ganglion cells,’ and the ganglion cells are organized to inhibit their neighbors as a direct function of how active each one is. The plus signs in Figure 1(b) represent activation by light; the minus signs inhibition. So the stronger the input to a given ganglion cell, the more that cell will reduce the activity in neighboring cells. The net effect is that patterns of light are registered by ganglion cells that are organized as small filters, optimally sensitive to a discrete spot of light that is surrounded by a region of reduced light. This is illustrated in Figure 1(b) by the circular regions of plus signs, surrounded by an annulus of minus signs. When one of these filters is centered on an intersection in the grid, its net activity (the sum of plusses and minuses exposed to the light) will be less than when it is centered on a white space between the squares. The result is that we experience a reduced sensation of light at the intersections. But why do the gray smudges disappear every time we fix our gaze directly on them? This relates to the way in which visual acuity diminishes as it moves away from the center of the eye. Because ganglion cells have very small ‘receptive fields’ (circular-surround filters) at the center of the eye, they are smaller than the size of the intersections and so illusory gray smudges are not seen. As we move away from the center, the receptive fields grow increasingly larger, ensuring that there will be a match between intersection size and the optimal size of receptive field at some distance from the center of gaze. Edge interactions Once the pattern of signals from the eyes has reached the brain, similar processes in the circuitry there operate to group edges together that belong to the same object and to segment edges away from each other that belong to different objects. Psychophysical experiments designed to study this process have used displays such as those shown in Figure 2.
(a)
437
(b)
Figure 2 Identical S-shape configurations, consisting
of short line segments aligned at their endpoints (a) or short line segments at right angles to these and aligned at their midpoints (b). The fact that the S-shape is more easily visible in (a) illustrates the cooperative effects between aligned edges and the inhibitory effects of neighboring edges of different orientation.
Both of these patterns contain the same number of short line segments arranged to form an ‘S’ shape. Yet it is easy to see the ‘S’ in the pattern in Figure 2(a) and almost impossible to see it in Figure 2(b). The only difference between the two patterns is that the short segments that make up the ‘S’ in Figure 2(a) are aligned with one another along their endpoints, whereas the same segments forming the ‘S’ in Figure 2(b) are aligned along their midpoints, placing each one of them at right angles to the imaginary curve forming the ‘S.’ You will be able to confirm their existence by looking at each line segment individually and comparing it to its neighbor, even if you are unable to see the large S-shape as a whole. The visibility of the S-shape in Figure 2(a) points to the existence of cooperative mechanisms that group neighboring edges of similar orientation and alignment. The invisibility of the S-shape in Figure 2(b) points to the existence of competitive mechanisms that keep neighboring edges of different orientation distinct from one another. Both cooperative and competitive mechanisms of this kind have been confirmed in electrophysiological studies of animals and in brain imaging studies in humans. Filling in edges and colors Everyone who looks at the shapes in Figure 3(a) says they see a white square shape in addition to some black round and oval shapes. Moreover, the square shape seems to be lying on top (or in front) of the oval shapes. When it comes to the color of
438
(a)
Visual Experience and Immediate Memory
(b)
Figure 3 An illusory square shape formed by subjective contours (a) and three illusory orange-tinged surfaces formed using the watercolor illusion (b). The visual system uses heuristics to interpret ambiguous images.
the white square, it seems to be even brighter than the whiteness of the surrounding paper, causing faint edges to appear where the square lies in front of the surrounding background. The faint edges of the square are known as subjective (or virtual) edges because they exist only in your brain. They do not exist on the page, as you can prove to yourself by blocking your view of one of the ovals with another piece of paper. These edges illustrate the way the visual system is constantly trying to find a reasonable interpretation of an ambiguous pattern of input. To do this the visual system uses heuristics (rules of thumb that work most of the time) to assign plausible interpretations to what it sees. In this case, it seems to be using a ‘convexity bias’ (in the absence of contradictory information, assume shapes with bulges are nearer to you than shapes with indentations) and ‘edge alignment’ (in the absence of contradictory information, assume that edges that are aligned belong to the same surface). Everyone who looks at Figure 3(b) says they see regions of orange separated by regions of white. Yet, when you look closely you can see that the pattern really only consists of two thin lines of equal thickness. One line is purple and much darker than the white of the background page; the other is orange and therefore more similar in its brightness to the background white. The illusion created by these two thin lines is that the color of the lighter line (the one more similar in brightness to the background) fills-in its region with a faint tinge of color whereas the darker of
the two lines creates a sharp edge, giving the appearance its regions may even be more intensely white than the background page. This ‘watercolor illusion’ occurs because the visual system follows several other heuristics when it comes to assigning color values to surfaces and to lighting conditions. In brief, sharp or abrupt changes in color or brightness are assumed to signal the edges of surfaces. More gradual changes are assumed to arise from surface pigmentation and lighting conditions, which includes coloration, shading, shadows, and other factors that contribute to the uneven distribution of light. The application of these assumptions then lead to the perception in Figure 3(b) of three surfaces with an orangish color, separated by two background regions of white separating these surfaces. Using heuristics and constraints There are many other heuristics used by the visual system in addition to those used in interpreting the boundaries and colors of surfaces. Figure 4(a) gives an example of a consequence of critical heuristic used to interpret whether a shaded region of an image is bulging out toward the viewer or is dented inward, away from the viewer. This is the heuristic that light generally shines from overhead. With this assumption, a disc that is lightly shaded on top, and gradually becoming darker toward the bottom, will be seen as a bump on a surface, whereas a disc with the reverse pattern of shading will be seen as a divot. The heuristic is effective, meaning it usually brings us to the correct interpretation, because we live in a world in which our most important source of light, the sun, shines from above. Thus, a heuristic can be thought of as our visual systems’ taking into account of a ‘constraint’ (i.e., a physical regularity) of the world in which our visual systems have evolved and in which we have spent a lifetime of experience. Gravity is a constraint that is also used in a number of heuristics employed by the visual system to interpret the pattern of light given to the eye. Because of gravity, objects are rarely able to hover in mid air, unless they are lighter than air or have a source of their own energy that permits them to fly. A very useful heuristic then is that objects are usually resting on or attached to
Visual Experience and Immediate Memory
(a)
439
(b)
Figure 4 Two shapes defined by shading to illustrate perceptual heuristics based on the constraints that light
generally shines from overhead (a) and that objects are usually resting on or attached to other surfaces (b).
other surfaces. This heuristic prompts us at first glance to interpret Figure 4(b) as a small block resting on a larger block. Closer inspection will reveal that there is another perfectly reasonable interpretation – it could be a small block hovering over the front right hand corner of the larger block – but this interpretation is more rare simply because our visual system seems to take the constraint of gravity into account.
for the category as a whole. One advantage of this type of coding is that it then benefits the task of recognizing a familiar object when it is seen from a novel vantage point or under unusual viewing conditions. This is because differences between any specific object and the average or prototype object will usually be preserved in the face of conditions of lighting or of viewpoint that affect the specific object and the average object in the same way.
Feature exaggeration
Enhancing vision with knowledge
Sketch artists and portrait painters have known for a long time that in order to communicate quickly and effectively with the brain of the viewer, it is best to make a drawing of a familiar face that is not quite faithful to its actual appearance. For best results, one should exaggerate those features of a face that distinguish it from the average or prototypical face. This means, for example, overemphasizing the chin and beard of Abraham Lincoln and the protruding ears of Prince Charles. Pictures drawn with these distinctive exaggerations are actually identified more quickly and accurately by viewers than are more realistic photographs of the same people, an effect referred to as the ‘cartoon advantage’ in object recognition. The cartoon advantage illustrates that the brain has some very efficient ways of storing and retrieving information in memory. Rather than keeping copies of all the image and scenes it encounters in their original form, it seems to store images in terms of their differences from some prototype or average
Constraints that arise from the physical conditions under which vision occurs are not the only factors that determine what we see. Knowledge that we have acquired about the peculiar characteristics of objects and the contexts in which they are most likely to occur also play a large role in our immediate visual experience. Experts see things at a glance that novices miss entirely or are only able to see with external guidance or more time. For example, experienced bird watchers and specialized dog breeders directly experience the specific species in a glance (in the case of bird watchers) and specific breeds in a glance (in the case of dog breeders) while novices directly experience only the higher-order category of bird or dog in a single glance. The same seems to be true for those trained in medical diagnosis, automotive repair, and rock climbing. One domain in which most humans have expertise is that of face recognition. This means they tend to see individual identities and specific emotional
440
Visual Experience and Immediate Memory
expressions in a single glance rather than simply experiencing a face. But because humans also tend to live in groups of similar looking individuals, they also tend to have expertise that is uneven when it comes to the identification of individuals of races that differ from their own. This means that when individuals from one racial group are given the task of learning to recognize individuals they have never seen before from either their own race, or from another race with which they have less experience, they tend to learn new faces from their own race more quickly. This is known as the ‘own-race advantage’ (or equivalently, the ‘cross-race disadvantage’) in face recognition. It is as though they experience faces of their own race at the level of individual identity, whereas they experience faces of another race at the level of group identity. Training and experience with faces of another race can overcome the cross-race disadvantage, just as training and experience with different bird species can turn the novice into an expert bird watcher.
Paradox Two: Rich Experience but Limited Report The previous section discussed how the visual system goes far beyond the information given in creating a rich visual experience from visual input that is, by comparison, much more impoverished and ambiguous than our corresponding experience of it. The present section will contrast this rich sensory experience with the surprisingly limited nature of what we are able to report about our experience at any moment in time. The severely restricted nature of our ability to report what we see was first studied systematically by George Sperling. His experiments involved flashing an array of English letters on a screen (several rows and columns) and asking participants to report as many letters as they could. The main result was that their report was limited to only three or four letters, even though it felt to the participants that they could see all of them quite clearly. This aspect of participants’ experience was confirmed in another study where they were asked to report only one row of the letters, depending on an auditory tone (high, medium, or low in pitch, corresponding to the top, middle, or bottom row)
that occurred immediately after the letter array had been erased from the screen. Under these conditions, participants could still report three or four letters from any given row, indicating that immediately following the flash of the letters they had conscious access to all of them. The problem with trying to report all of the letters was that in order to do so, some of the letters had to be committed to short-term memory while other letters were being reported. Short-term visual memory therefore placed restrictions on what participants could report about their immediate visual experiences. Since the original experiments of Sperling, the nature and limits of short-term visual memory have been studied in a large number of ways. Studies focusing on the question of its ‘capacity’ (the size of the memory store) indicate that it is large (virtually limitless) for a period of about 200 ms but then, following that brief period, it is limited to only three or four individual objects or events. Studies focusing on the ‘content’ (the nature of the representations in memory) suggest that each of the three or four items being held in short-term memory are coded quite richly, meaning that several of their features, including size, color, and shape, can be accessed for report from any of these objects if required. However, detailed comparisons of how much can be reported when the system is stressed to its maximum reveal that there are still deficits in reporting multiple features from two objects when compared with reporting the same number of features from only a single object. Thus, our ability to talk about our immediate visual experiences is greatly limited by the requirement to hold this information in shortterm memory while we are generating utterances about them that unfold over time. A popular laboratory task to study immediate visual experience that places only very minimal demands on the participant’s ability to report their experience is the ‘change detection task.’ In typical studies of change detection, two images are presented to the participant, either in temporal alternation, or beside each other in space, and the participant is asked to detect, locate, or identify any difference they can find in the two images. If the difference in the two images is detected immediately, then the experimenter assumes that the
Visual Experience and Immediate Memory
region in which the difference occurred in each image was represented in the visual experience of the participant. On the other hand, if the difference is only detected after some time, then the experimenter assumes that the region of the difference was not immediately accessible in the experience of the participant. Furthermore, the time required to detect the change can be used to index how long it takes before the changed region of the image has entered the consciousness of the participant. Many studies of change detection have shown that participants are surprisingly insensitive to large physical differences in two images (e.g., a large passenger jet that is missing an engine at the center of one of the two images), being able to detect the difference only after dozens of image alternations, which can take upward of 10 or 20 s. This profound insensitivity to change is therefore often referred to as ‘change blindness’ to help emphasize how long it takes for some aspects of the images to enter the consciousness of the participant. The primary factor that predicts success in a change detection task is whether the region of change is at the focus of the participant’s visual ‘attention,’ where this term refers to all those influences causing some information to be selected for more detailed processing at the expense of other information. One obvious influence on attention is the content of the participant’s shortterm memory. If the participant has selected three or four specific objects in a scene to be the focus of their immediate experience, then a change to one of those objects will be detected much more quickly than a change to an object that is not currently being held in short-term memory. What is selected as the focus of attention can itself be influenced by a wide range of factors, including the capture of attention by an abrupt physical event in a specific region of the image, the interest of the participant in the content of the images, the relevance of the information in the images to primary task of the participant, and the expertise of the participant for the scenes that are depicted. Many researchers believe that change blindness occurs because participants do not have the conscious access to the information in the scenes that they assume they have. In other words, change blindness represents a gap between our experience
441
of a rich visual world and the reality of much more impoverished representations. Some have even referred to this gap as the grand illusion of complete perception. However, others caution that successful change detection requires both a representation of the scene before the change and a comparison of that representation with the scene after the change. This means that change blindness could occur, either because of a failure to represent one of the scenes (as has often been assumed) or because of a failure on the part of the viewer to actively compare a region in one scene with its counterpart in the other scene. As such, comparison failures may have been overlooked in previous theoretical accounts of change blindness. In support of this possibility, it has been reported that viewers can recognize a previously attended object on a memory test even when they have failed to detect a change to the same object and that viewers can recognize both the unchanged and changed object at above-chance levels when they have failed to detect any change. If this interpretation is correct, change blindness may not imply a failure to form initial scene representations as much it points to a failure to make and test predictions between scenes. But so far, this is still only a conjecture that awaits systematic study.
Paradox Three: Experiences Evolve over Time from General to Specific Vision researchers have been puzzled for many years by the type of information that participants can report when given only a very brief glimpse of a scene. Their puzzlement comes from knowing that the visual information processed by the brain proceeds in a certain order that seems at odds with the order in which participants have conscious access to the information. In brief, the brain proceeds in its analysis of a visual image by first processing very specific details at each location in the image, before coalescing and abstracting this information in order to form higher-order concepts. This means that first there are neurons in our eye and then other neurons in the early visual centers of our brain that respond simultaneously to the minute details of what we are seeing. Only higher up in the anatomical hierarchy of processing do the neurons respond to familiar objects and entire scenes. Yet,
442
Visual Experience and Immediate Memory
when it comes to our experience of an image, things seem to proceed in exactly the opposite direction. We seem to be aware of the general content of a scene and its overall meaning to us well before we are aware of any of the specific details that underlie the formation of this general concept. Some vision researchers have referred to this as our ability to ‘see the forest before the trees.’ Others discuss this puzzle in terms of the ‘reverse hierarchy’ of visual experience. Some of the first studies of this puzzle showed participants brief glimpses of naturalistic images and asked them a variety of questions about the scenes’ content. Although questions about any of the specific details, such as the particular shapes or colors depicted, could only be answered for the small region of the scene that the participant happened to focus on in any glimpse, questions about the ‘gist’ (meaning) of the entire scene could be answered quite accurately. For example, participants were easily able to say whether the scene depicted a seashore, a jungle scene, or the interior of an office. Participants also had a very good idea of the general ‘layout’ (spatial arrangement) of the large objects in the scene. They even had very accurate impressions of the emotional content that was depicted, being able to say whether the scene was generally an angry or a happy one. Only with additional time and sometimes only with hard-won expertise were participants able to experience the specific details in a glance. Another way this issue has been studied involves creating images that depict different content at the level of detail and at the level of the overall configuration. Some famous art involving this principle is shown in Figure 5(a), while some displays from laboratory experiments are shown in Figure 5(b)). In both cases, it is easier to see the larger configurations (e.g., the face in Figure 5(a) and the letter ‘S’ in Figure 5(b) than it is to see the components of the configurations (e.g., the vegetables in Figure 5(a) and the many letter Es in Figure 5(b)). Experiencing these details takes additional time. This suggests that conscious awareness of what we are seeing begins high in the anatomical hierarchy of brain processes and is only able to access the component details that are represented lower in the hierarchy with additional time and mental effort.
EE EE E E E EE EE E E E EE EE E (a)
(b)
Figure 5 The portrait of a face composed of
vegetables (a) and the letter ‘S’ composed of numerous copies of the smaller letter ‘E’ (b). Visual experience usually begins with the larger configuration and only has access to the component details with additional time and mental effort.
Conclusion Visual experience is a construction by the neurons in our brain. To understand how this is accomplished, scientists in a number of related disciplines study three paradoxes. These include (1) that our phenomenally rich visual experience seems to derive from relatively impoverished visual input, (2) that although our experience may be rich in color, shape, and texture, we are severely limited in our ability to make an immediate report on it to others, and (3) that our visual experiences seem to evolve over time in an order that is reversed with respect to the anatomical order of processing, allowing us relatively quick access to the more abstract information conveyed in an image, but only slower access to specific details on which these abstractions are based. See also: Bistable Perception and Consciousness; The Neural Basis of Perceptual Awareness; Perception, Action, and Consciousness; Perception: The Binding Problem and the Coherence of Perception; Perception: Unconscious Influences on Perceptual Interpretation; Sensory and Immediate Memory.
Suggested Readings Biederman I (1981) On the semantics of a glance at a scene. In: Kubovy M and Pomerantz JR (eds.) Perceptual Organization, pp. 213–253. Hillsdale, NJ: Erlbaum. Bruner J (1973) Going Beyond the Information Given. New York: Norton. Crick F (1994) The Astonishing Hypothesis. New York: Charles Schribner’s Sons.
Visual Experience and Immediate Memory Duncan J (1984) Selective attention and the organization of visual information. Journal of Experimental Psychology: General 113: 501–517. Enns JT (2004) The Thinking Eye, the Seeing Brain. New York: W.W. Norton. Enns JT and Austen EL (2007) Mental schemata and the limits of perception. In: Peterson MA, Gillam B, and Sedgwick HA (eds.) In the Mind’s Eye: Julian Hochberg on the Perception of Pictures, Film, and the World, pp. 439–447. New York: Oxford University Press. Hochstein S and Ahissar M (2002) View from the top: Hierarchies and reverse hierarchies in the visual system. Neuron 36: 791–804. Luck SJ and Vogel EK (1997) The capacity of visual working memory for features and conjunctions. Nature 390: 279–281.
443
Navon D (1977) Forest before trees: The precedence of global features in visual perception. Cognitive Psychology 9: 353–383. Noe¨ A and O’Regan JK (2000) Perception, attention and the grand Illusion. Psyche 6(15). http://psyche.cs.monash. edu.au/v6/psyche-6–15-noe.html. Pinna B, Brelstaff G, and Spillmann L (2001) Surface color from boundaries: A new ‘watercolor’ illusion. Vision Research 41(20): 2669–2676. Simons DJ and Ambinder MS (2005) Change blindness: Theory and consequences. Current Directions in Psychological Science 14(1): 44–48. Sperling G (1960) The information available in brief visual presentations. Psychological Monographs 74: 1–29. (11, Whole No. 498).
Biographical Sketch
James T Enns is a professor at the University of British Columbia in the Department of Psychology and the Graduate Program in Neuroscience and the Director of the UBC Vision Lab (http://www.psych.ubc.ca/ennslab/). A central theme in his research is the role of attention in human vision. This includes studies of how the visual world is represented inside and outside of focused attention, how attention changes the contents of consciousness, how perception changes with development, and how to design visual displays for optimal human performance. He has served as an associate editor for the journals Psychological Science, Perception & Psychophysics, and Consciousness & Cognition. His research has been supported by grants from NSERC, BC Health, and NATO. He has authored books on Visual Cognition, The Thinking Eye, The Seeing Brain (W.W. Norton, 2004), edited two volumes on the Development of Attention, coauthored a textbook, Sensation and Perception (Wiley, 2004), and published numerous scientific articles. He was named a Distinguished University Scholar at UBC in 2003 and was the recipient of the Robert Knox Master Teaching Award in 2004. His PhD is from Princeton University and he is a Fellow of the Royal Society of Canada.
Visual Imagery and Consciousness N J T Thomas, California State University, Los Angeles, CA, USA ã 2009 N J T Thomas. Published by Elsevier Inc. All rights reserved.
Glossary Hard problem – In this article, the ‘hard problem’ of consciousness refers to the problem of understanding how a brain state (a spatiotemporal pattern of neuronal excitation, itself reducible to a complex dance of molecules and ions around membranes) could possibly be a conscious experience. Posed this way, the problem is unsolved, and may well be insoluble. It does not necessarily follow that consciousness cannot be scientifically understood, however. The problem may be ill-posed. Imageless thought – A thought that is (1) consciously experienced, but (2), unlike mental imagery, has no sensory or ‘perception-like’ character. The existence of such thoughts has been controversial ever since the notion was first introduced into psychology in the early twentieth century. Intentionality – The property, possessed by many, perhaps all, mental acts or entities (such as mental images, beliefs, desires, and thoughts in general), of being about, of, or directed at something. The ‘thing’ in question (sometimes referred to as the ‘intentional object’ of the image, thought, or whatever) may be real or imaginary (one may have a thought about a unicorn quite as well as a thought about a horse). This technical, philosophical concept of intentionality is only indirectly (if at all) related to the ordinary language notion of having an intention (to do something), or of doing something intentionally (i.e., on purpose rather than inadvertently). Mentalese – The hypothetical, innate, and unconscious, ‘language of thought,’ held by some cognitive scientists to be the symbolic format in which information is represented
within the brain, and which the brain, considered as a computer, uses in its computations. It supposedly resembles the languages people actually speak in having a combinatorial syntax and an arbitrary semantics. Clearly we are not consciously aware of our mentalese representations as such, but if they do indeed exist they may be the substrate of the conscious thoughts that we experience as mental imagery and ‘inner’ speech. Mental imagery – Quasi-perceptual experience: that is, experience that subjectively resembles perceptual experience, but which occurs in the absence of the relevant perceptual stimuli. It is generally acknowledged that imagery may occur in any sense mode – visual, auditory (including what is sometimes called ‘inner speech,’ or ‘thinking in words’), olfactory, kinesthetic, etc., – or even in several simultaneously. However, visual mental imagery, also colloquially referred to as ‘visualization,’ ‘seeing in the mind’s eye,’ ‘picturing,’ etc., has been by far the most extensively studied and discussed. Retinotopically mapped visual cortex – Certain visual processing areas of the cortex of the brain, most notably the primary visual cortex (V1), that are structured as (rather distorted and low resolution) maps of the light-sensitive retina upon which light is focused at the back of the eye. Adjacent regions of cortex, in these areas, correspond to adjacent areas on the retina such that, during vision, the two-dimensional spatial pattern of excitation of the cortical neurons corresponds topologically to the pattern of illumination, the optical image, on the retina.
445
446
Visual Imagery and Consciousness
Introduction For most people visual mental imagery is a common, frequent experience. We often recall past events, or imagine possible ones, by forming mental images. Our dreams may also consist largely of mental imagery. Indeed, many philosophers and psychologists have held that, together with immediate perceptual experience, imagery makes up the entirety of consciousness (although clearly not just visual imagery, but also imagery in other sense modes, especially inner speech, auditory, or vocal-motor imagery of spoken words). If we are to fully understand consciousness, we will certainly need to understand mental imagery. Philosophers have studied mental imagery for many centuries, experimental psychologists have studied it for well over a 100 years, and more recently it has attracted the attention of cognitive scientists and neuroscientists, but many basic questions still remain unresolved. There is controversy not only over what cognitive and neural mechanisms are responsible for imagery, but also over its function (if any) in cognition. Some regard mental imagery (or some closely related notion, such as ‘perceptual symbols’ or ‘image schemata’) to be the necessary vehicle for all thought; some regard it as important only for certain types of cognitive task (such as judging spatial relationships); and yet others regard it as a functionally insignificant conscious epiphenomenon of the unconscious cognitive processes that really constitute our thinking. Like percepts, mental images bear intentionality. That is to say, they are always images (or percepts) of something or other, of some (real or imaginary) ‘object.’ (Some philosophers regard intentionality and consciousness as very closely related phenomena.) However, unlike percepts, images occur in the absence of their object. You cannot perceive a cat when no cat is present, but you can imagine a cat (or any other perceptible thing) at any time. Furthermore, (mistakes aside) you cannot perceive a cat to be other than where and how it actually is, but, whenever you want to, you can imagine (i.e., form an image of ) a cat as anywhere, or in any condition. Images thus seem well suited to function as mental representations, allowing us to think of things as they currently are not, and thus to recall the past, plan for the future, fantasize about the unreal, and
speculate about the unknown. Since ancient times, this has generally been believed to be their cognitive function.
Defining Imagery: Experience or Representation? Mental imagery is commonly defined as a form of experience: quasi-perceptual experience, experience that subjectively resembles the experience we have when we actually perceive something (see ‘Glossary’). This implies that we are unlikely to be able to understand imaginal consciousness unless we understand perceptual consciousness (and perhaps vice versa). Unfortunately, we do not yet have such an understanding (or, at least, one that is generally agreed upon). It also, however, implies that imagery is always and necessarily conscious: if something is not consciously experienced it cannot be mental imagery. If, on the other hand, mental imagery is defined as being a form of mental representation, as some contemporary cognitive theorists suggest, the tight conceptual linkage between imagery and consciousness is broken, and it becomes conceivable that images (imaginal representations) might sometimes play a role in cognition without our being consciously aware of them. There is some evidence suggesting that this does indeed occur. For instance, experimental studies of verbal memory have found that nouns for which it is easy to think of a corresponding image (mostly words for concrete things, such as ‘dog,’ ‘ship,’ or ‘skyscraper’) are more readily remembered than nouns for which it is difficult to think of an image (mostly abstractions, such as ‘truth,’ ‘nation,’ or ‘size’). However, this mnemonic effect of ‘imagability’ seems to occur quite regardless of whether any relevant images are actually consciously experienced (or, at least, of whether the subjects report or recall experiencing them). One interpretation of this finding is that image representations may be spontaneously evoked by the concrete words, and may play a role in making these words more memorable, even when they do not rise to consciousness. Unlike the experiential conception of imagery with which we started, this representational conception suggests that we might be able to
Visual Imagery and Consciousness
understand the nature and function of imagery without giving any attention to the fact that it is (at least sometimes) consciously experienced, and, in fact, most empirical and theoretical cognitive science research on imagery over the past few decades has proceeded on that assumption. The problem of the imaginal consciousness has, very largely, been ignored or set aside, while questions and controversies about its representational function (and, more recently, its neural instantiation) have been pursued enthusiastically. Furthermore, little if any attention has been given to the question of why we should be conscious of some of our image representations and (supposedly) not others. But in any case, imagery probably cannot be satisfactorily understood purely in terms of its representational function. It is difficult or impossible to differentiate it from other forms of actual or possible mental representation, without either, on the one hand, begging some very controversial questions about its nature (about whether, e.g., the relevant representations are somehow picture-like), or, on the other hand, appealing to the experiential conception of imagery. We could (and perhaps should) say that mental images are just those mental representations whose presence to mind has the potential to give rise to quasi-perceptual experiences. If so, however, the representational conception of imagery becomes conceptually dependent upon the experiential conception, and we cannot even begin truly to understand imagery unless we take its conscious nature into account.
Historical Development of Ideas about Imagery Contemporary scientific controversies about mental imagery by no means turn entirely on matters of empirical fact. Disagreements turn largely upon conceptual issues, and upon differing ideas about what questions a theory of imagery most needs to answer. For example, is it crucial to consider the conscious and intentional nature of imagery when we seek to understand its cognitive mechanisms and functions, or are such considerations irrelevant and distracting? An awareness of the historical contexts from which the various contemporary research programs emerged is indispensable in
447
understanding and adjudicating between such differing perspectives. Scholars disagree as to whether ancient thinkers such as Plato and Aristotle, whose works contain the earliest known discussions of cognition, had anything like our modern concept of consciousness. However, it is clear that they did have the concept of mental imagery. Plato tentatively suggests that memory might be analogous to a block of wax into which our perceptions and thoughts stamp impressions (i.e., memory images); he also speaks, metaphorically, of an inner artist who paints pictures in the soul, and suggests that imagery may be involved in the mechanisms by which the rational mind exerts its control over the animal appetites. However, it is Plato’s successor Aristotle who provides the first systematic account of the role of imagery in cognition. In Aristotelian psychological theory, images play much the same role that the rather broader notion of mental representation plays in modern cognitive science. He held that mental images play an essential role in memory and thought: memory is the recall to mind of images of past events, and ‘It is impossible to think without an image.’ He also held that images underpin the meaningfulness of language, and play a key role in motivation. Were it not for mental images, he thinks, our speech would be empty noise, like coughing, and something could only excite our desire or fear while it was actually present to our senses. However, a mental image of the desired (or feared) thing enables us to think about it in its absence, and thus can sustain our motivation to obtain or avoid it at other times. Aristotle also posits a mental faculty, phantasia (usually translated as ‘imagination’), that is closely allied to (or perhaps even structurally identical with) the general faculty of sense perception, and is responsible for creating our mental images. Some suggest that this faculty amounts to his conception of consciousness. In the wake of Aristotle’s work, through later antiquity and the middle ages and into the era of modern philosophy, images continued generally to be seen as the principal vehicles of mental content. In the work of Descartes and his successors the mind was explicitly understood as conscious, and the contents of consciousness were known as ‘ideas.’
448
Visual Imagery and Consciousness
Descartes himself seems to have at least two distinct concepts of ‘idea.’ What he calls a ‘clear and distinct idea’ is a direct (perhaps propositional) mental grasp of the essence of something. It exists within consciousness itself, but is not an image. However, he also hypothesizes that when we see, imagine, or remember something previously seen, a pictorial image is formed deep within the brain (on the surface of the pineal gland). Such images, which he also quite explicitly calls ideas, do not exist within the mind as such (since, for Descartes, the conscious mind is immaterial, and distinct from the brain), but they are somehow presented to the mind and are the immediate causes of our conscious perceptual, memory, and imaginative experiences. The British Empiricist philosophers’ concept of ‘idea’ seems to combine aspects of both of the notions found in Descartes. Although some scholars today question whether John Locke really believed ideas to be picture-like, it is natural to interpret him as thinking so, and there is little room for doubt that his Empiricist successors, most notably Berkeley and Hume, conceived of ideas as images. Hume distinguishes ‘impressions’ (i.e., percepts) from ‘ideas’ that arise from the memory or the fancy, but he thinks that these differ only in the intensity with which we experience them, and he seems to conceive of all visual impressions and ideas as consciously experienced, picturelike images. Unlike the image-ideas of Descartes, however, for the Empiricists’ image-ideas are themselves entirely mental, and exist only inasmuch as we are conscious of them. Indeed, Hume’s so called ‘bundle’ theory of the mind suggests that these images actually constitute consciousness: the mind is nothing but a bundle of impressions and ideas. Although it certainly had its critics, this Empiricist conception of mind continued to be influential up to, and beyond, the emergence of scientific psychology in the late nineteenth century. Pioneering experimental psychologists such as Wilhelm Wundt in Germany and William James in the United States thought of psychology as the study of consciousness and of the images (and emotional ‘feelings’) that populate it. However, in the early twentieth century a group of psychologists based in Wu¨rzburg in Germany, led by Oswald Ku¨lpe, reported that systematic introspection of thought processes under controlled laboratory
conditions had led them to recognize that, as well as images, the mind also contains ‘imageless thoughts.’ These were, supposedly, conscious experiences (they had been discovered by introspection after all) and cognitive (they were evoked during reasoning, and were not merely emotional feelings), but, unlike imagery, they had no sensory character. These claims proved extremely controversial. Some psychologists, such as Wundt, argued that the introspective methodology employed by the Wu¨rzburg psychologists was inherently unscientific and unreliable. Others, such as Edward Titchener, argued that competently conducted introspective investigations revealed no evidence of imageless thoughts. The ‘imageless thought debate’ that ensued proved to be irreconcilable. However, no party to this debate was claiming that thought in general is imageless; rather, the point at issue was whether nonimaginal, but nevertheless conscious, thoughts actually exist. Although the factious arguments about this have long since died down, the issue has never truly been resolved. Some people still hold (although perhaps more often as an implicit assumption, rather than an explicitly defended view) that all the cognitive contents of consciousness have a sensory character, either as actual sensations or percepts, or else as mental images; others hold (again, often implicitly) that there are also nonsensory conscious thought contents. These are sometimes referred to as states of ‘fringe consciousness,’ and may be described as ‘feelings’ of, for example, ‘familiarity,’ ‘unfamiliarity,’ ‘rightness,’ ‘wrongness,’ etc. Historically, however, the upshot of the deadlock of the ‘imageless thought debate’ was a general discrediting of introspective methodologies in psychology, and even of the very idea that consciousness could be studied scientifically. John B. Watson argued that consciousness was an inherently unscientific notion, and, as a central plank of this argument, questioned the very existence of mental imagery. He urged that psychology should be reconceived as the study of behavior rather than the study of consciousness, and the behaviorist movement that he instigated came to dominate the field for the next several decades (very roughly, from c. 1920 to 1960). Few, if any, of the behaviorist psychologists who succeeded Watson went quite as far as he did toward denying the reality of
Visual Imagery and Consciousness
consciousness and imagery, but almost all of them, in practice, treated them as beyond the reach of science. During the same period, the rise of the analytical philosophy movement (with its emphasis on formal logic and language as the keys to understanding knowledge and the mind), and particularly the influence of Ludwig Wittgenstein, also led many philosophers to discount the significance of imagery, or simply to ignore it. The upshot was that neither imagery nor consciousness received much serious attention during the behaviorist era. A revival of interest in imagery, however, was an important component of the cognitive psychology movement that challenged and eventually displaced behaviorism as the dominant psychological paradigm in the 1960s and 1970s. New experimental methods for studying imagery were devised that did not depend on purely subjective introspection, and evidence emerged suggesting that imagery plays a large role both in memory (as shown mainly be experiments on verbal memory), and spatial thinking (as shown by experiments on ‘mental rotation’ and ‘mental scanning’). A strong empirical case was built up for the functional importance of imagery in cognition, and a vigorous (and continuing) debate ensued about the nature of the cognitive mechanisms responsible for imagery. However, there was not a comparable revival of scientific interest in consciousness until the 1990s. As cognitive theories of imagery developed in the 1960s through the 1980s and beyond, only cursory attention was given to the fact that it is a conscious phenomenon (or to the, arguably, closely related fact that it bears intentionality). Even today (the early twenty-first century), cognitive scientists and neuroscientists often treat imagery purely as a form of representation, and offer no substantive account of how images are able to represent (almost certainly it is not because they resemble their objects), or of how they come to be consciously experienced.
Subjective Individual Differences in Imagery Experience Since the pioneering work of Francis Galton in the nineteenth century there has been a tradition of research into individual differences in the
449
subjective experience of imagery. Galton prepared a questionnaire in which he asked his subjects to recall their morning breakfast table, and to consider ‘the picture that rises before your mind’s eye.’ They were then asked to comment on a number of subjective aspects of that image, such as its brightness, its clarity, and the distinctness of the colors. One of the best knowing findings of Galton’s study is that some small minority of the questionnaire respondents reported that they were unable to visualize anything whatsoever. Unfortunately, no systematic research has since been done on this phenomenon, and the issue remains very poorly understood (although recent research has directly contradicted Galton’s related claim that scientists are particularly likely to be weak imagers or ‘non-imagers’). Subsequent researchers have refined Galton’s questionnaire technique in various ways, most importantly, perhaps, by introducing numerical scales along which the subjects can rate the strength of the image attribute of interest. Most often, this attribute is ‘vividness,’ but there are also questionnaires that attempt to measure such things as how easily someone can transform their mental images (by asking them to form an image and then change it in some specified way, and then rate the difficulty of doing so), or how frequently they use images of one or another sense mode in their thinking. In the early twentieth century there was considerable interest in classifying people into ‘imagery types’ according to which sensory mode of imagery (visual, auditory, haptic, etc.) they preferred. It was hoped that these ‘types’ might prove to be correlated with other independently measurable aspects of the person’s psychological makeup. However, no clear cut pattern of findings emerged. Perhaps the most extensively and successfully used imagery questionnaire of recent times is the Vividness of Visual Imagery Questionnaire (VVIQ) devised by David Marks. Taking the VVIQ involves being asked to visualize a series of specified objects and scenes and to rate the vividness of each resulting image on a five-point scale ranging from ‘no image at all’ to ‘as clear and vivid as normal vision.’ An average score is then calculated. Clearly research of this sort deals with imagery as a conscious phenomenon. It relies upon the subjects’ verbal, introspective reports of their
450
Visual Imagery and Consciousness
subjective experience. Furthermore, there has been much effort to find correlations between people’s scores on imagery questionnaires and their other psychological traits. The results, however, have generally been disappointing. Even where reproducible correlations have been found, it has often proven difficult to make much theoretical sense out of them. For example, there is some evidence that vivid imagers (as measured by the VVIQ ) remember pictures (color photographs) rather better than less vivid imagers; however, surprisingly, it has also been found that the less vivid imagers (by the VVIQ again) actually remember specific shades and hues of color better. Some research suggests that when people form mental images they have elevated levels of activity in the retinotopically mapped visual cortex of their brain, and that this activity is greater for more vivid imagers. However, other researchers fail to find elevated activity in these areas during imagery (they find it elsewhere in the brain). Perhaps most surprisingly and disappointingly, however, virtually no sign of any correlation has been found between people’s vividness ratings and their performance (speed or accuracy) in various visuo-spatial thinking and problem solving tasks, even though, subjectively, such tasks seem to depend on imagery. There are also a number of conceptual concerns about the validity of questionnaires of this type. Most obviously, as the rating is necessarily purely subjective, and as nobody can experience another’s image, there is no way to tell whether the scales are being applied in a consistent way. Perhaps an image that one person thinks of as ‘clear and reasonably vivid’ another might rate as ‘vague and dim,’ simply because different subjective standards are being applied. Furthermore, most people will probably agree that the vividness of their imagery can vary markedly from time to time, circumstance to circumstance, and image to image: Some memories may come back to us (often for no very apparent reason) as especially vivid images, whereas others (or even the same ones on other occasions) are recalled only vaguely and dimly. Thus it may be that these tests do not measure a person’s true capacity for having vivid images, but only, at best, the vividness that their images tend to have under the circumstances of filling out an imagery questionnaire.
Finally, it is unclear quite what ‘vividness’ truly means; whether it is really a well defined, coherent attribute of experience. Perhaps different subjective image features such as clarity, apparent brightness, level of discriminable detail, stability of the image (whether it can be held in consciousness for a time, or quickly fades), and so forth, in fact vary independently of one another, but tend to get indiscriminately lumped together as ‘vividness.’ Some recent research suggests that introspective ratings of certain more fine grained aspects of subjective imagery experience may be more meaningful than simple vividness ratings.
Theories of Imagery, and Their Implications for Consciousness There is much controversy in the contemporary scientific and philosophical literature about the underlying nature and mechanisms of imagery. The rival theories fall into three broad classes, each of which have distinctive implications for our understanding of imaginal consciousness:
Picture theory (its contemporary advocates often prefer to call it ‘quasi-pictorial,’ or ‘analog’ theory) Description theory (also, rather misleadingly, known as ‘propositional’ theory) Enactive theory (there is no established name for this type of theory; different versions are known as ‘motor theory,’ ‘sensorimotor theory,’ ‘perceptual cycle theory,’ ‘role taking theory,’ and ‘perceptual activity theory’) Because of the vigorous, high-profile dispute (known as the ‘analog-propositional’ debate, or sometimes just ‘the imagery debate’) that flared up between advocates of the first two theory types in the 1970s, and which continues to rumble on even today, picture and description theories have become by far the better known of these theoretical alternatives. Unfortunately, this debate has been conducted almost entirely without regard to the conscious nature (or, indeed, the intentionality) of imagery. Each of the three theory types purports to be able to account for the full range of established empirical facts about imagery, and, in fact, each is flexibly enough conceived that it is likely to prove difficult to choose between them on
Visual Imagery and Consciousness
narrowly empirical grounds. That being so, it seems reasonable that our evaluation of them should look to broader theoretical concerns, such as their respective implications for the theory of consciousness. Picture Theory Picture theory, as the name suggests, holds that visual mental images may be identified with representations in the mind and/or brain that are in some significant way like pictures, or, at least, that represent things in much the same way that pictures represent things. (Whatever that way may be: Although it is widely believed that pictures represent through resemblance, it does not seem likely that resemblance can be the basis of mental representation. Resemblance is a fundamentally subjective relationship that is recognized when a conscious mind focuses upon certain similarities between things while, at the same time, discounting or ignoring other, equally objectively real dissimilarities. In order to do this, the mind must already be able to represent the things, and their relevant aspects, to itself.) Picture theory is undoubtedly the oldest and most widely accepted theory of imagery. After all, since prehistoric times people have used pictures as a means of causing one another to have visual experiences of things that are not really there, and the analogy with mental imagery (also experience of things that are not really there) is all too easily drawn. The theory goes back at least to Plato, and some rudimentary version of it seems to be deeply entrenched in ‘folk’ (i.e., lay or ‘commonsense’) beliefs about how the mind works. Indeed, many of the expressions used to talk about imagery in ordinary colloquial language (such as, ‘mental picture,’ ‘the mind’s eye,’ and even ‘image’ itself ) seem to be derived from this ‘folk’ version of picture theory. The fact that the theory is embedded in our language in this way can make it difficult to express or understand criticisms of it, or to clearly grasp alternative accounts of the actual phenomena. Nevertheless, many twentieth and twentyfirst century philosophers have raised telling objections against it, leading some of them to be accused (quite unjustifiably) of denying that people have quasi-visual experiences.
451
However, the picture theory today is not just a folk theory. From the 1970s onward, cognitive scientists, most particularly Stephen Kosslyn, have developed the basic idea of the mental image as an inner picture into a sophisticated and detailed scientific model designed to account for a wide range of experimental findings. More recent versions of this theory hold that the visual mental image is embodied as a spatially extended twodimensional pattern of neural excitation in the retinotopically mapped visual cortices of the brain. Such patterns, isomorphic to the pattern of optical stimulation on the retinae of the eyes, are known to occur during vision. Kosslyn holds that our mental images are similar patterns, but generated from internal sources rather than sensory input. This is not the place to review all the alleged empirical shortcomings of quasi-pictorial theory. However, both introspective and empirical evidence suggest a number of ways in which the experience of imagery and the experience of looking at a picture vary. Perhaps the strongest evidence comes from experiments involving pictures that invite more than one interpretation, such as the Necker cube and the duck–rabbit (Figure 1). Subjects find it much harder to see both interpretations in their mental images of these than they do when the picture is physically displayed in front them, and this applies even when the picture is their own drawing, based upon their mental image. Kosslyn maintains that the pattern of neural excitation that (he thinks) constitutes the mental image is only a picture in an extended, metaphorical sense: a ‘quasi picture.’ After all, unlike regular pictures (drawings, photographs, projected optical images, etc.), we do not need to see it with our eyes in order to derive information from it, or consciously experience it. The neural excitation pattern is like a picture only inasmuch as it represents
Figure 1 The Necker cube and the duck–rabbit.
452
Visual Imagery and Consciousness
spatial relationships within the two-dimensional projection of the represented visual scene by topologically equivalent spatial relationships in the spatially extended pattern of excitation. Kosslyn maintains that because his theory does not identify mental images with pictures in the full, literal sense, but only with these ‘quasi pictures,’ it avoids all of the many powerful objections that have been raised against naı¨ve ‘folk’ versions of picture theory. Whether this is so, however, remains controversial. Although the modern version of pictorial theory can account for a large amount of empirical data about how people use mental images as representations, its supporters have given very little attention to the question of how it might account for the fact that imagery is (at least sometimes) consciously experienced. There seem to be two basic ways in which this issue might be approached. Either the quasi picture in the cortex might be conscious in and of itself, or, alternatively, consciousness arises when some other, more inward cognitive structure, a ‘mind’s eye’ as it were, somehow ‘quasi-sees’ the quasipicture. The first option, however, runs immediately into the notorious ‘hard problem’ of consciousness in its most intractable form: we really have no idea how a pattern of excitation in part of the brain (excitation that is ultimately nothing but an elaborately choreographed dance of molecules and ions around and through membranes) could possibly be, or give rise to, a conscious experience of any sort, let alone an experience of a spatially extended structure isomorphic to the excited brain region. In any case, the second option appears to be favored by Kosslyn. He speaks of (and diagrams) a ‘mind’s eye function’ gathering information from the quasi-pictorial image in the brain, and some of his explanations of empirical findings seem to depend upon this idea. Perhaps then, just as ordinary visual consciousness arises when people take in visual information with their eyes, imaginal consciousness arises when the mind’s eye takes in visual information from the inner picture. But the ‘mind’s eye’ metaphor strongly suggests that the theory commits the homunculus fallacy. To whom might this eye belong if not to a conscious little man inside my head, who experiences the inner world of quasi pictures in a way analogous to that in which I, a whole person, experience the
external world through my bodily eyes? Compounding this worry is the fact that, at a formal, structural level (abstracting away from implementational details), Kosslyn’s quasi-pictorial theory of imagery looks very like a more detailed version of Descartes’ theory of imagery. For Descartes too, mental images are material quasi pictures formed within the brain and consciously experienced not simply because they are there, but because they are somehow ‘seen’ by the soul. Kosslyn makes it clear that he does not intend to follow Descartes in attributing our consciousness of these brain pictures to an immaterial soul (a ghostly homunculus) beyond the reach of science; but although quasi-pictorial theory may not necessarily imply Cartesian dualism, it does seem to be at least committed to Cartesian materialism, the notion that subjective experience depends upon some structure in the brain that (like a soul or a homunculus) acts as a conscious inner spectator of internal representations. Quasi pictorialists seem to have two comebacks to the homunculus objection. One is to point out (as Kosslyn has often done) that a computer simulation of quasi-pictorial theory has been implemented, and that there can be no homunculus (at least of any objectionable sort) within a computer. Unfortunately, however, it turns out that the computer program in question makes no attempt whatsoever to model the conscious nature of imagery. Like most of Kosslyn’s work, it is concerned to explore how certain sorts of spatial thinking might be achieved through imagery, via such operations as scanning linearly across a quasi picture, or rotating it into a new orientation. The program performs such operations on internal data structures that are supposed to model, at a functional level, the brain quasi pictures posited by the theory, and it displays its results in the form of actual pictures that are rotated or shifted across the computer’s display screen in relevant ways. No attempt is made to simulate the hypothesized ‘mind’s eye function,’ and consciousness, either real or simulated, does not enter into the matter at all unless and until some human operator looks at the screen. It is true that there is no homunculus in the computer, but, inasmuch as the system models consciousness at all, it models it by co-opting a full sized conscious human being to play the homunculus role.
Visual Imagery and Consciousness
Not coincidentally, the program also omits to model the intentionality of imagery. The fact that the two-dimensional patterns it produces and manipulates may look like pictorial representations to human onlookers (and were designed to look that way by the programmers) is quite irrelevant to the program’s functioning. For the computer, they represent nothing. The second comeback is to suggest that perhaps cognitive science can render the homunculus and its ‘mind’s eye’ innocuous by showing that it can be reduced to a set of computational or neural processes. Quasi-pictorial theory was originally developed in the context of the ‘information processing’ paradigm which dominated perceptual theory at the time (the 1970s), and which still deeply influences the way many cognitive scientists think about perception. In essence, ‘information processing’ theory regards vision as a one way flow of visual information, in through the eyes and then through a series of processing stages in the brain until it is eventually transformed into a representation, or a set of representations, suitable for guiding behavior. (This is an oversimplification, but not, in this context, a misleading one. Developed information processing theories generally call for a significant degree of top-down modulation of the bottom-up flow of information from the sense organs. Nevertheless, the bottom-up flow dominates and drives perception.) A common assumption (usually implicit) is that this final set of representations, this ultimate product of the visual information processing system, is the immediate cause of conscious visual experience. (Some information processing theorists may prefer to think of visual consciousness not as something attaching to representations, but, rather, as something arising from the workings of the processing system as a whole, perhaps even including the sense organs and the muscles that support behavioral response. However, this position is not open to the pictorial imagery theorist, who necessarily holds that we have conscious experience of inner representations.) According to quasi-pictorial imagery theory, one of the earlier stages of this visual information processing is the creation of quasi-pictorial representations in the brain. These may be derived from actual present visual input from the eyes (when we
453
are actually seeing), or they may be created from stored data in memory (when we are remembering or imagining), but in either case they must be passed through several more stages of processing in order to extract useful information from them. It is these further stages that constitute the ‘mind’s eye function.’ Thus, it can be argued, the mind’s eye has a principled and independently motivated role within the broader theory of vision. As scientists are actively investigating and seem to be making progress in understanding the computational and neural mechanisms of visual information processing, this ‘mind’s eye’ function is not so much a nonexplanatory homunculus as a promissory note drawn against the expected success of an ongoing research program. But even if information processing theory does provide the right framework for understanding visual perception (and not everyone thinks it does), the output of all the processing is just more representations, instantiated as patterns of neural activity. Once again, it seems that we must say either that these representations are conscious in and of themselves, which brings us smack up against the ‘hard problem’ once again, or else we need another homunculus to read them and be conscious of what they represent. Perhaps this homunculus, also, might be reduced to a further series of stages of information processing, but this would only lead us to the same place yet again, and so on in unending regress. We still do not begin to understand how quasi pictures could be, or could produce, conscious imaginal experiences. Description Theory Description theory, whose most important advocate has been Zenon Pylyshyn, began as an attempt to understand how the phenomena of imagery could be fitted into a computational theory of the mind. Pylyshyn holds that the way computers (and, by extension, brains) represent information is necessarily more like language than like pictures. The syntax, and, indeed, the vocabulary, of the hypothetical internal brain language (sometimes called ‘mentalese’) might be very different from that of any language that anyone speaks, but it is language-like in that it ultimately consists of symbolic tokens that represent things in the world in
454
Visual Imagery and Consciousness
much the same way as the words of a language do, not through resemblance but through some essentially arbitrary correspondence relation. In the case of ‘natural’ languages, like English or Chinese, this correspondence (which determines, for instance, that the word ‘dog’ refers to dogs) is established by social convention. In the case of humanly written computer programs, the external reference of symbolic tokens (when they have one) is set by the programmer. Matters are much less clear when we come to hypothetical symbolic tokens in the brain, and there is great controversy over how, or whether, the ‘symbol grounding problem’ (the problem of understanding how the ‘words’ of mentalese might be able to refer to things outside the brain) can be solved. Nevertheless, much work in cognitive science proceeds on the assumption that a solution is possible. If the symbolic tokens that comprise computational representational systems represent as words do, then it seems to follow that, by analogy with the way language represents visual scenes, the braincomputer represents visual scenes with descriptions. From this perspective, such mentalese descriptions are the end-product of perceptual information processing, and thus responsible for our perceptual experiences. Of course, we do not experience the descriptions as being descriptions, but, it is assumed, they nevertheless constitute our perceptual experience. If mentalese descriptions of visual scenes are retrieved from memory rather than being the result of present sensory input, or if they are constructed out of bits and pieces of various descriptions in memory, then we have the experience of mental imagery. Pylyshyn claims that, since descriptions can be partial, incomplete, and can leave out all sorts of information (not only details, but sometimes even facts about the global structure of a scene) description theory can explain the frequently indefinite and ambiguous nature of our imagery better than picture theory can. However, the theory’s main motivation is clearly the belief that it better respects fundamental facts about the nature of computational representation. If anything, however, description theorists have paid even less heed than quasi-pictorial theorists to the question of how images (or, come to that, percepts) come to be consciously experienced.
Clearly, we are not aware of mentalese descriptions as such, but the idea seems to be that we can, at least sometimes, be consciously aware of what they represent. Like the quasi pictorialists, however, description theorists appear to be implicitly committed to the view that mental representations, including the elementary symbolic tokens of mentalese, and the symbol complexes that make up visual descriptions (and, thus, mental images), are physically instantiated as patterns of brain excitation (just as representations in a computer are instantiated as electrical charge patterns in RAM chips and CPU registers). This brings us to exactly the same ‘hard problem’ as before: it seems impossible to conceive how patterns of brain excitation (which, themselves, are reducible to movements of ions, etc.) could be, or give rise to, conscious experience as we know it. Enactive Theory Quasi-pictorial and description theories are both attempts to explain how imagery, quasi-visual experience, might be explained within the context of an information processing theory of perception. They disagree over whether it should be identified with representations from an early or late stage of visual processing, and over the format of those representations. The enactive theory of imagery depends upon quite a different way of conceiving perception, one that was pioneered in the twentieth century by James Gibson, and is more recently exemplified in ‘active vision’ techniques used in robotics, and the ‘sensorimotor’ or ‘enactive’ theory of perception advocated by J. K. O’Regan and Alva Noe¨, amongst others. Instead of regarding vision as, at root, a matter of information flowing in through the eyes into the brain, and toward some internal center of consciousness, enactive theory regards it as a matter of the visual system actively seeking out and extracting (or ‘picking up’) desired information from the environment. Seeing is not like taking a photograph (even a digital photograph whose file gets passed on to a computer for further processing); it is more akin to performing a series of (many, rapid) scientific tests and measurements on the information-bearing ambient light that surrounds us (what Gibson called the ‘optic array’). Rather than being the
Visual Imagery and Consciousness
passive reception of information, vision (and perception in general) is a purposive process of, as it were, asking questions about our surroundings, and actively seeking out the answers. Visual consciousness is the experience of being engaged in this exploratory, questioning, information-seeking activity, involving not only the making of the tests and measurements themselves, but also the continual adjustment of our expectations, and thus our exploratory behavior itself, in response to their findings. Consider how we identify an object by touch. If something is simply pressed against our skin, we can tell very little about it. However, if we actively explore it, moving our fingers around to feel its shape and texture, seeking out corners and edges, squeezing it to assess its hardness, and so on, we can discover a great deal. What we learn derives not simply from the sensations we feel in our skin, but, crucially, from the way those sensations change in response to the purposeful, informationseeking movements of our fingers. Enactive visual theory holds that vision works in fundamentally the same sort of way, except that the exploratory, information-seeking behavior mostly happens so quickly and automatically that its details are unavailable to introspection. Eye movements are perhaps the most obvious (and experimentally accessible) aspect of this visual information-seeking behavior – our eyes constantly, and purposefully, but largely unconsciously, flit rapidly around to take in information from different points of interest – and the enactive theory of imagery finds support in a number of recent experiments showing that the stimulus-specific eye movement patterns produced when a subject examines a complex visual stimulus are (quite unconsciously) reenacted when the same subject later forms a memory image of that same stimulus. However, eye movements cannot be the whole story. There clearly are bottom-up, inner representations involved in human vision. For example, the pattern of illumination on the retina does indeed produce a corresponding pattern of excitation in the visual cortex, and this carries information about the scene momentarily before our eyes. From the enactivist perspective, however, such representations are not conscious or even mental in any meaningful sense. Quasi-pictorial theory seems to imply that, when we see, these representations
455
are what we are conscious of. Enactive perceptual theory denies this, and holds instead that we are conscious of the actual things out in front of our eyes. Like the optical image formed inside a camera, the excitation pattern of visual cortex carries information about our surroundings, but this only gives rise to visual knowledge and awareness inasmuch as this data source is purposefully explored, searched, and queried. It is this active, exploratory searching that turns the mere passive reception of visual information (something that a camera can do) into true perception. It is important to note, also, that the particular pattern of this exploratory questioning, and the particular tests and measurements brought into play (both external eye movements and internal data analysis processes), will be quite different when different sorts of things are being looked at. For example, in order to see a cat, we would need to go through a set of exploratory actions appropriate to cat-seeing. These might include moving our eyes to focus on the likely location of particular characteristic cat features (pointed ears, gently curving tail, etc.), but will also undoubtedly include purely internal processes whereby further questions are answered by querying the data flowing bottomup into the brain. The key idea is that there is a specific structured set of exploratory queries, what has been called a visual routine, appropriate to cat seeing (or even Tiddles-curled-up-asleep seeing), and a different set for each other type of thing that we are able to recognize. So, if visual experience arises from the active seeking for, and finding, of information, imagery, quasi-visual experience, arises when we actively seek certain information, and persist in going through the motions of looking for it, even though it is not there to be found. To visually recognize a cat is to run through a specific, cat-recognizing visual routine, and to imagine a cat is to perform (or partially perform) this same visual routine when there is no cat present. The experience is different from actual seeing because there is no perceptual feedback from the cat itself (and we must, as it were, force ourselves to go on looking for cat features even though it is becoming ever more clear that we will not find them) but it is similar to real cat-seeing inasmuch as the pattern of perceptual exploration that we perform is
456
Visual Imagery and Consciousness
similar. Arguably, this reflects the phenomenology of imagery, which is subjectively both like and unlike true seeing. In particular, it usually seems to take much more effort to sustain an image in consciousness for any length of time than it does to simply keep looking at something. From the enactive perspective, to have a conscious visual experience is not to have a representation in one’s brain, but to act, and to be interacting with the world, in a certain way. To have imagery is to act in the same sort of way, but to fail to interact. Actions (unlike brain states) have inherent intentionality, just as mental images do: actions are necessarily about, or directed at, something (possibly something imaginary, as when we search for the leprechaun’s gold), just as images are always images of something. When I enact my perceptual routine for looking at, and recognizing, a cat, my action is intended to enable me to see a cat, and in a real sense it remains so, it remains catdirected, even when I enact it knowing that no cat is actually there to be seen. However, to point out that actions have intentionality is not, thereby, to explain how they come to have that intentionality, and to say that mental images are really actions rather than entities (representational brain states) is not, in itself, to explain how they come to be conscious. Much more work needs to be done to make these matters clear. Although it has been argued that enactive (or ‘sensorimotor’) perceptual theory can explain perceptual consciousness, such claims remain very controversial. Nevertheless, the enactive theory of imagery clearly radically reframes the problem of explaining imaginal consciousness. No longer is it (as for quasi-pictorial and description theories) a problem of explaining how a brain state can be a conscious (and intentional) state. Rather the problem is to give a scientific account of the intentionality of action (a problem we face anyway), and then to explain how such action can somehow constitute consciousness. This is, no doubt, a difficult problem, but it is not the ‘hard problem.’
Conclusion At the time of writing, imagery is still a rather neglected topic within the broader, and growing,
field of consciousness studies. This is in rather stark contrast to the importance it was accorded by earlier, pre-behaviorist students of consciousness. Indeed, the behaviorist revolt against consciousness, which led to several decades of scientific neglect of the topic, seems to have been fueled, in considerable part, by frustration at the factious and irreconcilable ‘imageless thought’ controversy, and it clearly involved the rejection of imagery quite as much as the rejection of consciousness itself. Perhaps recent students of consciousness have been reluctant to fully engage the topic of imagery because it has already proven so controversial within cognitive science, and because the best-known theories that have been developed in that context do little to illuminate imagery’s conscious nature. However, this situation may now be changing, and a developing understanding of imagery may come, once again, to be seen as an integral and essential aspect of our developing understanding of consciousness itself. See also: Bistable Perception and Consciousness; Mental Representation and Consciousness.
Suggested Readings Baars BJ (ed.) (1996) Special Issue: Mental Imagery. Consciousness and Cognition 5(3). Berman D and Lyons W (2007) The first modern battle for consciousness: J. B. Watson’s rejection of mental images. Journal of Consciousness Studies 14(11): 5–26. Dean GM and Morris PE (2003) The relationship between self-reports of imagery and spatial ability. British Journal of Psychology 94: 245–273. Ellis RD (1995) Questioning Consciousness: The Interplay of Imagery, Cognition, and Emotion in the Human Brain. Amsterdam: John Benjamins. Kosslyn SM, Thompson WL, and Ganis G (2006) The Case for Mental Imagery. Oxford: Oxford University Press. Marks DF (1999) Consciousness, mental imagery and action. British Journal of Psychology 90: 567–585. Morris PE and Hampson PJ (1983) Imagery and Consciousness. London: Academic Press. O’Regan JK and Noe¨ A (2001) A sensorimotor account of vision and visual consciousness. Behavioral and Brain Sciences 24: 939–1031. Pylyshyn ZW (2002) Mental imagery: In search of a theory. Behavioral and Brain Sciences 25: 157–237. Richardson JTE (1999) Mental Imagery. Hove, UK: Psychology Press. Sartre J-P (1940) The Psychology of Imagination, Frechtman B (trans.). New York: Philosophical Library, 1948.
Visual Imagery and Consciousness Schwitzgebel E (2002) How well do we know our own conscious experience? The case of visual imagery. Journal of Consciousness Studies 9(v–vi): 35–53. Thomas NJT (1999) Are theories of imagery theories of imagination? An active perception approach to conscious mental content. Cognitive Science 23: 207–245.
457
Thomas NJT (2008) Mental imagery. In: Zalta EN (ed.) The Stanford Encyclopedia of Philosophy. Stanford, CA: CSLI . Thompson E (2007) Look again: Phenomenology and mental imagery. Phenomenology and the Cognitive Sciences 6: 137–170. Tye M (1991) The Imagery Debate. Cambridge, MA: MIT Press.
Biographical Sketch
Nigel J T Thomas was born in the United Kingdom where he acquired degrees in biochemistry and in humanities, and his doctorate in the history and philosophy of science from Leeds University (where he worked with U.T. Place, a pioneer of modern materialist theory of consciousness). He was a visiting scholar in psychology at the University of California, Davis (where he worked under two other pioneers of consciousness studies, Thomas Natsoulas and Charles Tart), and has studied with leading philosophers of mind and cognition Robert Cummins, John Heil, and Robert Gordon. He was a Mellon postdoctoral instructor in philosophy at the California Institute of Technology, and has taught philosophy, history of psychology, and history of science at various other California colleges. He is currently an adjunct professor at California State University, Los Angeles. He first became interested in the cognitive science of mental imagery in the early 1980s, and has published and presented many papers on the related topics of imagery, consciousness, and imagination. His Web site, ‘Imagination, Mental Imagery, Consciousness, and Cognition: Scientific, Philosophical and Historical Approaches’ (http://www.imagery-imagination.com/), gives access to most of his work on these topics, and provides many links to other relevant online resources.
William James on the Mind and Its Fringes B J Baars, The Neurosciences Institute, San Diego, CA, USA ã 2009 Elsevier Inc. All rights reserved.
Glossary Automaticity, habit – The tendency of practiced, predictable events to fade from consciousness after repetition, including voluntary actions, concepts, mental images, and attentional routines. Automatic processes tend to be specialized, take up little conscious capacity, and may resist voluntary control. Behaviorism – An influential physicalistic philosophy of psychology, some forms of which deny the existence or functional role of consciousness. Conscious experiences, focal – Experiences like sensory perception, visual imagery, inner speech, visualizable action plans, etc., which tend to have sensory features like color, texture, taste, object identity, and the like. Conscious experiences, fringe – Experiences that lack specific, sensory qualities, like the tip-of-the-tongue state (the intention to seek a missing word), feelings of knowing, familiarity, and plausibility, intuitive judgments, specific intentions to act, expectations, relational terms in grammar, logic and reasoning, abstract meanings, emotional connotations, and numerous other conscious or quasi-conscious events that can be reported with high accuracy but low sensory specificity. Conscious experiences, operational definition of – While there is no agreedupon theoretical definition of conscious experiences, in actual practice scientists have used ‘‘accurate verbal report’’ for several centuries with excellent reliability. This is a useful operational definition of a large range of conscious experiences, both sensory and endogenous (as in the case of conscious inner speech, visual imagery, and
episodic recall). All of sensory psychophysics is based on this straightforward method, going back to Newton’s discovery of the subjective color spectrum, which is highly reliable between subjects. Hypnosis – A state of surprisingly high compliance with external suggestions, especially after a perceived induction procedure, which may be arbitrary or symbolic. Hypnotic suggests can influence sensory perception, the normal sense of voluntary control, emotions, imagery, pain perception, analgesia, and their known brain correlates. A high percentage of the normal population is considered to be ‘‘highly hypnotizable’’ as measured by standardized tasks. Such subjects seem to treat the hypnotist’s suggestions as highly credible, although they do not act in violation of their ordinary social norms. Hypnosis may involve an absorbed state with minimal self-examination. Ideomotor theory – In William James and others, the notion that conscious goals are inherently impulsive, and tend to be carried out by default unless they are inhibited by other conscious thoughts or intentions. Introspective reports – Reports about conscious experiences, which can range from highly reliable ones (such as psychophysical reports) to unreliable ones (such as mental images evoked by abstract concepts). Introspectionism – A controversial term attributed by behavioristic historians to nineteenth century researchers on the topic of consciousness. Introspection was explicitly disavowed as a useful method by the most productive experimental psychologist of the nineteenth century, Wilhelm Wundt.Wundt criticized
459
460
William James on the Mind and Its Fringes
introspection as unreliable, and rarely used it. However, Wundt’s popular introduction to psychology used introspective examples, which may have misled many of his readers in English. Most of Wundt’s extensive body of writing was never translated into English, leading to decades of highly misleading interpretation of Wundt as an ‘‘introspectionist.’’ However, introspective method was utilized by the Wuertzburg School and by Edward Titchener, who were strongly criticized for that by Wundt. Stream of consciousness – The apparently unpredictable moment-to-moment contents and discontinuities of spontaneous conscious thinking. William James popularized the ‘‘stream of consciousness’’ metaphor, and also the bird flight analogy of ‘‘flights’’ and ‘‘perches’’ in the flow of conscious experiences.
‘‘Psychology,’’ William James wrote in 1890, ‘‘is the description and explanation of states of consciousness, such as sensations, desires, emotions, cognitions, reasonings, decisions, volitions, and the like.’’
That was the broad consensus during most of Western and Asian history, and during the century of scientific studies summarized in James’ Principles of Psychology (1890). The period from 1780 to 1910 was arguably the Golden Age of consciousness science, simply filled with discoveries about sensory experiences, mental and neurological disorders, hypnotic suggestion, visual imagery, voluntary control and its loss, and much more. James’ Principles offers 1400 pages of inspired writing on the major questions of human psychology. Building on mostly European studies, it has given us classic descriptions of the stream of conscious thought, selective attention, mental imagery, hypnosis, multiple personality, habit, effortful concentration, the basic arguments for and against unconscious processes, a theory of voluntary control and its disorders, the crucial distinction between self-as-subject and self-as-agent, and much more. On many of these topics James’ thinking is fully up to date, and it is embarrassing that much of the time he is still ahead of the scientific curve today.
One of the wonderful things about James’ Principles is his open embrace of the whole range of human experiences. He has the rare gift of letting go of his own biases for a while to let the phenomena shine. That, plus his extraordinary ability to describe and analyze subtle aspects of our experiences makes him more than a man obsessed with philosophical disputes. That is why scientists find themselves revisiting James on topics like selective attention, the stream of consciousness, the remarkable range of fringe experiences, and much more. James attributed his penchant for observation to Luis Agassiz, one of the great naturalists of the nineteenth century, with whom he traveled on an expedition to the Amazon. Like Charles Darwin, Agassiz was a passionate observer of flora and fauna, a great collector, endlessly seeking the patterns in living things. For example, Agassiz is credited with discovering the geological evidence for the Ice Ages, a remarkable feat of naturalistic observation. From Agassiz, James learned the vital habit of trying to free one’s mind from prior beliefs, for, in biology as in psychology, preconceptions undermine our ability to see. This habit was sometimes called ‘botanism,’ the passionate observation, collection, and classification of plants and animals for their own sake. James’ Principles ‘botanized’ the psychology of the nineteenth century. As a compendium of psychological facts and ideas the Principles has not been surpassed. James despised obstacles to free inquiry and pursued his interest in the mind everywhere, through hypnosis, nitrous oxide intoxication, accounts of religious experiences, his reluctance to get out of bed on a cold morning, the tip-of-thetongue state, psychic claims, and clinical accounts of multiple personality syndromes. He was rarely a true believer or a rigid skeptic, but always an explorer.
Fringe Consciousness By far the bulk of James’ Principles is devoted to the same empirical phenomena we study today. Only a few out of some thirty chapters of that work were philosophical, but those few chapters have received far more attention than
William James on the Mind and Its Fringes
James’ empirical side. Philosophers have done a better job in appreciating William James than have scientists, in general, though he was at least as much an empiricist as a metaphysician. Fringe-conscious experiences provide an excellent example of the empirical James. Intuitively we tend to think of conscious experiences as clear, percept-like, reportable events that stand out well as figure from ground. Those are the most commonly studied cases. But a large part of our mental life is occupied with ‘fringe’ events, which are experienced as fuzzy or vague, but which have properties suggesting that something very precise is going on. They include feelings of knowing, of familiarity, of beauty and goodness, of mismatch, incongruity, or surprise. As James points out, the fringe also includes a great variety of judgments, expectations, intentions, abstractions, intuitions, and logical or grammatical relations (like ‘if,’ ‘or,’ and ‘but’). Further, we seem to have accurate ‘feelings of knowing’ about potential conscious contents that are readily available, though they are not immediately conscious – our ability to retrieve words, our moods, potential actions, knowledge about others and ourselves, semantic memories, and much more. Feelings of knowing have now been studied experimentally in considerable detail, and the evidence indicates that (1) they are often accurate; (2) they enjoy high confidence ratings; but (3) they do not involve detailed, structured experiences – unlike the sight of a coffee cup, where we can talk about shape, color, shading, texture, figureground contrast, clear temporal boundaries, and much else. In addition, we now know of brain regions that seem to be activated by some fringe experiences. For example, the ‘sense of mental effort’ appears to evoke BOLD (fMRI) activity in the anterior cingulate and dorsolateral prefrontal cortex. These regions are known to be involved in expressed goals, and to be triggered by goal conflicts and barriers. These functional properties relate closely to the ‘sense of mental effort.’ The most famous example of a fringe experience is the tip-of-the-tongue state – the tantalizing feeling of searching for an elusive word. At first it seems like a curiosity, but then begins to reveal deeper and deeper implications. James writes,
461
Suppose we try to recall a forgotten name. The state of our consciousness is peculiar. There is a gap therein; but no mere gap. It is a gap that is intensely active. A sort of a wraith of the name is in it, beckoning us in a given direction, making us at moments tingle with the sense of our closeness, and then letting us sink back without the longed-for term. If wrong names are proposed to us, this singularly definite gap acts immediately so as to negate them. They do not fit into its mold.
The tip-of-the-tongue state is a delayed intention to find a missing word, a mental state that lacks qualities like color, sound, or taste; it has no clear boundaries in space and time, and no contrast between figure and ground. All expectations and intentions seem to be like this. To show the power of such states we need only interrupt some dense flow of predictable experience, for example, a printed ________ like this one. Spontaneously we want to fill in a word that fits. We can see the same effect by interrupting a joke just before the punch line; clever musical composers continuously play with our expectations about songs. Expectations and intentional states like this govern all our activities. They are not images or perceptions. Yet such colorless mental events compete against other sensory events for access to our conscious mental sphere. James thought that ‘fringe’ states comprise perhaps a third of our mental life. Some of us now believe that they shape all of our conscious experience, without exception. All human thought and action appear to be driven by expectations. The tip-of-the-tongue state provides a good case to study, because it draws out a colorless expectation over many seconds. We now have the first brain imaging studies of the tip-of-the-tongue state. It shows, as we might expect, that the state activates frontal cortex much more than the sensory regions of the back of the brain. That is presumably why it lacks sensory qualities. ‘The fringe’ is therefore a fine example of the way in which nineteenth century science went far beyond common sense in studying consciousness. Probably the most famous fringe event is the ‘tipof-the-tongue’ experience, which has proven to be a rich and productive domain of experimental study. But the range of fringe or ‘vague’ conscious phenomena, as James described it, is far broader than is generally recognized. It has only barely been touched in contemporary science.
462
William James on the Mind and Its Fringes
James Rejected the Unconscious Toward the end of the nineteenth century scientific thinkers like Pierre Janet and Sigmund Freud began to infer unconscious processes quite freely, based on hypnotic suggestion, conversion hysteria, slips of the tongue, self-serving forgetfulness, and the like. Freud’s ideas achieved unparalleled influence, so that the art and literature of the twentieth century is incomprehensible without them. Unlike Freud, William James fiercely resisted the psychological unconscious. In a remarkable section called ‘Refutation of alleged proofs of unconscious thoughts,’ he considers the possibility of unconscious intelligence. In his characteristically fair-minded way, he provided ten basic arguments ‘pro,’ followed by arguments ‘con.’ The ten arguments pro are still some of the best we have. The first real clash comes in the famous chapter on ‘Habit,’ where James considers how it is that, in learning a new skill like riding a bicycle, the conscious details of pedaling, steering, and balancing soon turn into the unconscious routines of expert cycling. We know today that the brain contains numerous unconscious networks that analyze and control such things as balance, eye movements, visual space, and the muscular control. Their existence can be inferred from vast amounts of psychological evidence, and today, with neuroimaging, we can actually see them at work in the brain. Unconscious habits seem to involve less cortical activity, and more subcortical mechanisms like the basal ganglia and cerebellum. James stated the case for the notion that habits are unconscious: . . .we do what originally required a chain of deliberately conscious perceptions and sensations. As the (habitual) actions still keep their intelligent character, intelligence must still preside over their execution. But since our consciousness seems all the while elsewhere engaged, such intelligence must consist of unconscious perceptions, inferences, and volitions.
But he could not tolerate unconscious intelligence. He wrote that, Reply: There is more than one alternate explanation. . . One is that the perceptions and volitions in habitual actions may be performed consciously, only so quickly and inattentively that no memory remains. Another
is that the consciousness of these actions exists, but is split-off from the rest of the consciousness of the hemispheres. . . .
Habits may therefore reflect fast, hard-to-remember, or split-off conscious contents. Rapid conscious ‘flashes’ may in fact exist, and there may indeed be dissociated conscious contents, as we know from studies of hypnotic dissociation. So James’ counterargument is by no means silly, but few scientists today rule out a major role for complex, unconscious intelligence. Take another of James’ arguments about the unconscious, drawn from biological instincts, like nest-building in birds. Instincts, as pursuit of ends by appropriate means, are manifestations of intelligence; but as the ends are not foreseen (consciously), the intelligence must be unconscious.
But again, Reply: . . .all the phenomena of instinct are actions of the nervous system, mechanically discharged by stimuli to the senses.
Unconscious processes are ‘merely physical.’ Crucially, James often resorts to a mind–body argument to rule out unconscious intelligence. In an age of computers, we no longer share James’ intuitions that all intelligent processes must be conscious. But he was too much a child of his times to accept the shocking consequences of unconscious thought. James and his contemporaries just could not imagine a high degree of unconscious intelligence. In their century, reason, purpose, and intelligence were believed to be the exclusive possession of consciousness. ‘Unconscious intelligence’ seemed a bizarre violation of common sense, as Helmholtz found out in the 1860s, when he suggested that the brain may come to unconscious conclusions about the visual world when direct information is missing. For example, each eye has a blind spot, about the size of a quarter at arm’s length. There are simply no light receptor cells in the blind spot. But we almost never see it, because we ‘fill in’ the gap, based on surrounding colors and textures. This idea is perfectly plausible and is generally believed today. In the nineteenth century it was heresy, leading to furious protests until Helmholtz
William James on the Mind and Its Fringes
felt compelled to withdraw the offending words ‘unconscious conclusion’ (unbewuesste Schluss). The existence of unconscious processes in the visual brain was not fully accepted until the 1970s! Today many scientists believe that the brain makes innumerable inferences every second. Thus in the nineteenth century unconscious events had to be attributed to the physical brain. But the brain was held to be incapable of logic; it was still a mechanical servant of lived experience. Ever since Aristotle, logic and rationality were believed to be the unique preserve of consciousness, the seat of reason. Sigmund Freud, who made the unconscious interesting to the public, never really believed that it could reason. Freud’s unconscious is bereft of logic or consistency, always in romantic turmoil. It is the irrational ‘cauldron of seething excitations,’ the dark dungeon of contending passions. The idea of an intelligent unconscious began to make sense only very recently, driven by the new tangible reality of the logic-crunching computer. When James tried to understand how unconscious habits could form from conscious origins he faced a forbidding paradox. On one side, accepting a brain basis for consciousness would go against his lifelong commitment to free will; on the other, he could not abandon his understanding of brain science. His solution was amazingly awkward: Novel actions had to originate in the nonphysical land of consciousness. As they became habits and faded from consciousness, they would somehow cross the great mind–body divide into the physical domain of the brain. Wrote James, An acquired habit, from the physiological point of view, is nothing but a new pathway of discharge formed in the brain, by which certain incoming currents ever after tend to escape. . . .the philosophy of habit is thus a chapter in physics rather than psychology. (italics added)
James was not happy with this awkward dualism; he twisted and turned through all the alternatives, but could not escape contradiction, hemmed in by his own incompatible assumptions.
James and the Mind–Body Problem This was the empirical James, who summed up the most important discoveries of the nineteenth century. But there is another James, the metaphysician.
463
This is the person who tried for a lifetime to solve the unending problem of free will. The metaphysical James came to the fore again in the last decades of his life. Unfortunately, James’ metaphysics undermined his own scientific writing, and may have destroyed his high reputation among psychologists, who were trying to create a stable academic profession. On metaphysics, James admitted that his thinking was not purely rational. We know from his history of suicidal despair and the need to believe in free will that there is a great undercurrent of emotion in his thinking. James never denied that. In an essay called The Will to Believe, he defended a belief in God even in the face of a lack of evidence. We have the right to believe at our own risk any hypothesis that is live enough to tempt our will. . . . So if I accept the religious hypothesis because doing so makes me more happy than I would otherwise be, then I am rationally justified in my decision.
And in Is Life Worth Living? he wrote, These, then, are my last words to you: Be not afraid of life. Believe that life is worth living, and your belief will help to create the fact.
Saving Free Will Human beings are all mind-brain philosophers, whether we know it or not. Are you freely responsible for your actions? If you say so, you are claiming free-will mentalism. If you do not believe in free will, but think that all human experience is only a fictional gloss on the firing of neurons, physicalism is your game. And if, like most of humanity, you find yourself switching between mind and body explanations in everyday life, you are adopting dualism. In any moment of the day we can slip subtly between two very different ways of thinking about ourselves. We appeal to a physical vocabulary to explain the effect of aspirin on headaches, but we switch to mind words whenever we want to claim credit or to assert our freedom from external control. Did I break my glasses? No, a book just fell on them. (Physical) But do I work hard to provide for my family? You bet, and I expect a little credit for it. (Mental) Children learn early on to excuse their
464
William James on the Mind and Its Fringes
actions as uncontrollable accidents when they might be blamed for the results, but to take personal credit when they do something praiseworthy; this childhood pattern hardly changes for many of us until the end of life. When we get a little more sophisticated we learn to import physical causation into psychological events: I did the wrong thing, yes, but it was because of sleepiness, distraction, something came over me. This is of course the key defense in the courts when a defendant claims extenuating circumstances – Prozac, the failures of society, a history of abuse, or the Twinkie defense, a murderous rage said to result from eating too much sugar. The language of the law is the language of free will, personal responsibility and just deserts. But the language of science is the language of simple physical causation. From this perspective a murderer has no more responsibility for killing his victim than a billiard ball has for a missed shot. The Mind–Body Vortex Tends to Swallow All Else The mind–body problem is still today the dominant obsession in the philosophy of mind. It asks how the physical world could possibly be reflected in our private experience; how our subjectively free intentions could emerge in physical action; and how all this could relate to the physical substrate of experience, the brain. Whenever scientists make significant advances – as in two centuries of findings about color perception – philosophers routinely tell us it is not good enough; we still do not know about real consciousness, which is now redefined to exclude the new discoveries. It looks an awful lot like a city-slicker trick that psychologists, simple country folk, fall for with astonishing regularity. Twentieth century science made a great commitment to physicalism. The most extreme versions of physicalism deny private experience completely, aiming to explain all things exclusively by public observables – neurons, or stimuli and responses, or molecules. Behaviorism is a psychological version of this philosophy, as B.F. Skinner often said. So is the neural reductionism that is widely held by neurobiologists today. Francis Crick’s hypothesis is that our experience of each
precious moment is fully explainable in terms of neurons in the brain. While this is a long-standing scientific hypothesis, it has an undeniable philosophical agenda. However, mentalism is alive. The physics Nobelist Roger Penrose claims that consciousness can only be understood by way of quantum mechanics. Penrose argues for a modern mentalism, that reality is to be found in the ‘quantum mind.’ Physics seems to show a division between quantum phenomena and the visible world of objects, but quantum explanations are thought to be more fundamental. Penrose defines consciousness in terms of the direct apprehension of mathematical truths – exactly Plato’s idea 24 centuries ago. The realm of consciousness – the quantum level – underlies visible reality. Those views make no contact with psychology or the brain, but they are sincerely held by some very intelligent scientists. Each classical position on mind and body seems plausible at certain times and perverse at others. Each is seductive, and each seems to lead to paradox. Intuitively we all swing back and forth between the three classical positions, sometimes in a single sentence. Taking a ‘physical’ aspirin for a ‘mental’ headache is intellectually perplexing; being a physicalist and yet taking personal credit for one’s own achievements – as if they were freely chosen – is equally inconsistent. Dualism avoids these contradictions at the price of its own unanswered puzzles: How could a mind relate to its brain? How do conscious intentions turn into the physical actions of the muscles? And how do physical sensations end up as conscious experiences? The mind–body puzzle is not some artifact of Western thought. Each classical position appears early in Asia as well as Europe, starting in India and China and later in Japan and South East Asia. In the West, mentalism was first stated in writing by Plato in fourth century BCE. Athens. A few centuries earlier, it was articulated with great power by Gautama Buddha and the Vedanta philosophers in India, and by the early Taoists in China. All mystical philosophies are mentalistic – they claim, like Roger Penrose, that a transcendent reality underlies our everyday world. Asian philosophies acknowledge the physical world, but suggest that it results from an imperfect realization of one’s own consciousness; at the bottom, reality is
William James on the Mind and Its Fringes
mental. Another strand in Indian philosophy is called dvaita, or dualism, from the same root as ‘dual.’ Dualism, physicalism, and mentalism can be found in many parts of the world. A lively philosophical cottage industry survives today on the mind–body problem. Philosophical thinking about consciousness is almost exclusively concerned with it, though thousands of other questions can be asked. Every novel has something to say about the varieties of human experience, but we dance around only one philosophical mulberry bush, and the dance never seems to change. So far, no one has found a settled solution. After more than two millennia of written debate on the subject, arguments are as persistent as ever. Arthur Schopenhauer called it the ‘World Knot’ – unsolvable but also unavoidable. It is interesting that Schopenhauer’s ideas were shaped by the Vedanta scriptures written more than 2000 years before, which had just been translated into Western languages in his time. Wilhelm Wundt, often called the founder of Western experimental psychology, was very much influenced by Schopenhauer, so that we can trace a direct line from the mind–body philosophers of the ancient Indian world to the beginnings of Western scientific psychology. The seductiveness of the World Knot is difficult to overstate. The Encyclopedia of Philosophy concludes that The mind–body problem remains a source of acute discomfort for philosophers. . . .It may well be that the relation between mind and body is an ultimate, unique, and unanalyzable one. If so, philosophical wisdom consists in . . . accepting it as the anomaly it is.
Science Usually Evades Unresolved Philosophical Puzzles Great philosophical controversies always arise with major scientific changes. When Copernicus and Newton argued that the sun could keep the earth in orbit, philosophers attacked them for proposing an obviously absurd idea: action at a distance. There were no giant rubber bands connecting the sun and the earth, or the earth and the moon. They just happened to stay connected because of an invisible thing called ‘gravity.’ But gravity could not be seen or touched. It was an imaginary theoretical idea. Newton had no answer, and we
465
still do not have one today. His response was to be purely pragmatic, saying ‘non fingo hypothesi ’ – I do not speculate – which was no answer at all. Charles Darwin’s evolutionary theory also evoked attacks from philosophical vitalists, like Henri Bergson, who argued for an invisible ‘vital essence’ in all living things. That may not seem a powerful argument in the age of biotechnology, but it convinced generations of philosophers that something was wrong with biological science. Biologists essentially ignored those attacks and went on studying the genetics of peas and fruit flies. Over many decades that paid off, and the philosophical arguments faded away. Successful science does not wait for all philosophical questions to be solved. It is very pragmatic. One of its practical moves is to sidestep questions that cannot be resolved, and simply find a straightforward way to gather evidence. After 25 centuries of debate about mind versus body there is little doubt that scientists should not try to solve the metaphysical problem first – that effort has a long record of failure. They should simply collect evidence about human consciousness and try to understand it. We need to follow our empirical noses unburdened by metaphysical baggage. That is happening today. But William James could not make that choice. He was too deeply committed to his own need to believe in free will, which protected him from suicidal despair. James Torn Mind–body debates were one front in the longrunning battle between science and religion in the nineteenth century. It took place in heated family arguments between parents and children around Victorian dinner tables, about ancient religious faiths and the new faith in Science and Progress. William James found himself squarely in the middle of this battle. He was a physician after all, trained in the physicalistic medicine of the nineteenth century; he was hired to teach brain anatomy at Harvard. Like other physicians of the time – Freud, Helmholtz, and Charcot – he learned to explain the mind’s evolutions in terms of brain processes first of all. But James was also a child of the transcendental tradition of Emerson and Thoreau, raised in a
466
William James on the Mind and Its Fringes
family where religious and philosophical debate flourished. Henry James Sr. was an enthusiastic Swedenborgian mystic and close friend of Ralph Waldo Emerson. Henry Jr. became one of the foremost psychological novelists. Alice, their brilliant sister, spent a life of suffering and illness that was thought to have a psychological element. It is not surprising that William was torn between physical and mind-centered science; the perfect person, in fact, to symbolize psychology at the end of the nineteenth century. James’ Quest for Personal Meaning James was a man blessed, and at times cursed, with an extravagance of talent. As his student wrote, ‘‘Brilliant, high-strung, dynamic, vivacious, resilient, unexpected, unconventional, picturesque – these are some of the terms that at once recur in recalling James.’’ It was a widely shared judgment. He was that rare thinker who was also admired as a human being. From his Swedenborgian father, he first learned about metaphysics and religion; with his brother Henry, the novelist, and his sister Alice, he talked for years about everything that three brilliant young people of the nineteenth century could imagine. The family traveled to Europe and the children learned to speak French and German. William trained as a medical doctor and painter, studied with some of the foremost scholars in Europe, accompanied the naturalist Luis Agassiz on an expedition to the Amazon; and finally found a profession teaching at Harvard. In psychology, James’ greatest achievement was The Principles of Psychology of 1890. It is still a bottomless well of ideas. A few years later he reduced it to a handier Briefer Psychology, which became the standard introductory text in America for the next 30 years.
James Sometimes Wandered Away from Standard Science In the very next chapter, William James jumps off the edge of science as we know it. He begins with a warning: The reader who found himself swamped with too much metaphysics in the last chapter will have a still worse time of it in this one, which is exclusively metaphysical.
It is an ominous sign, for here James is forced to adopt panpsychism, the idea that all matter must have some rudimentary consciousness. In a section titled ‘Evolutionary psychology demands a minddust’ we read, If evolution is to work smoothly, consciousness in some shape must have been present at the very origin of things. . . .we find that the more clear-sighted evolutionary philosophers are beginning to posit it there. Each atom of the nebula, they suppose, must have had an aboriginal atom of consciousness linked with it; and, just as the material atoms have formed bodies and brains by massing themselves together, so the mental atoms. . . have fused into those larger consciousnesses which we know in ourselves and suppose to exist in our fellow-animals.
The trouble is that one could say the same thing about living matter – just fill in ‘life’ for ‘consciousness.’ Arguments for a life essence were common in James’ day and well into the twentieth century, before biochemistry came of age. Almost no one believes them today. If we think of life as some indivisible essence it may seem right that it would exist to a tiny degree in every atom and molecule, but with a better understanding of carbon molecules this aura of reasonableness simply fades away. In just the same way, if we think of consciousness as some unanalyzable ‘essence’ we can make the beguiling leap of logic that everything must be conscious to some degree. But the more we learn about what neurons are doing to make conscious experience possible, the harder it is to believe in panpsychism. Consciousness is first of all a major biological adaptation. Panpsychism is not testable today, and remains extrascientific. Science thrives on testable questions. By mixing good science with hotly debated metaphysics, James cast doubt on the very foundations of the new psychology. By the end of the chapter James has rediscovered the soul. Many readers have certainly been saying to themselves for the last few pages: ‘Why on earth doesn’t the poor man say the soul and be done with it?’. . . all the arguments (made here) are also arguments for (the soul). . .. I confess, therefore, that to posit a soul influenced in some mysterious way by the brain-states and responding to them by conscious affections of its own, seems to me the line of least logical resistance. . .
William James on the Mind and Its Fringes
And with a flourish he concluded: nature in her unfathomable designs has mixed us of clay and flame, of brain and mind. . . the two things hang indubitably together and determine each other’s being, but how or why, no mortal may ever know.
Summary and Conclusions William James summarized an extraordinary century of discovery in the 1400 pages of his Principles of 1890. Almost all his empirical phenomena are still well-validated today. But he became ensnared in mind–body issues, which were not testable or solvable in his time. Because he was deeply involved in nineteenth century debates about free will and personal meaning, it was not possible for James to fully separate his role as an empiricist from his other identity as a metaphysician. As the most famous advocate of philosophical pragmatism, James is still well known in philosophy today. But in the sciences, his contributions were lost during the behaviorist era, and are poorly understood even today, because of retrospective misunderstandings of his many empirical observations about conscious functions. The role of fringe consciousness is one example of a first-rate
467
empirical discovery that has been well-validated by modern research, but which is still not widely understood as an entire separate category of mental life. Thus we still have a great deal to learn from James’ Principles of 1890. See also: Neuroscience of Volition and Action; Philosophical Accounts of Self-Awareness and Introspection; Self: Personal Identity; Self: The Unity of Self, Self-Consistency.
Suggested Readings Baars BJ (1988) A Cognitive Theory of Consciousness. New York, NY: Cambridge University Press. Bargh JA and Morsella E (2008) The unconscious mind. Psychological Science 3(1): 73–79. Cohen MS, Kosslyn SM, Breiten HC, et al. (1996) Changes in cortical activity during mental rotation: A mapping study using functional MRI. Brain 119: 89–100. Kikyo H, Kenichi O, and Miyashita Y (2002) Neural correlates for feeling-of-knowing: An fMRI parametric analysis. Neuron 36: 177–186. Maril A, Wagner AD, and Schacter Daniel L (2001) On the tip of the tongue: An event-related fMRI study of semantic retrieval failure and cognitive conflict. Neuron 31: 653–660. Wagner J, Stephan T, Kalla R, et al. (2008) Mind the bend: Cerebral activations associated with mental imagery of walking along a curved path. Experimental Brain Research 191: 247–255.
Biographical Sketch
Bernard J. Baars is former senior research fellow in theoretical neurobiology at the Neurosciences Institute in San Diego (www.nsi. edu). His PhD is in cognitive psychology from UCLA. He is interested in human language, the brain basis of consciousness, volition, and a variety of related topics including the history of scientific studies of consciousness, and neuroethics. Baars pioneered a cognitive theory of consciousness called Global Workspace Theory, which is widely cited in scientific and philosophical sources. Together with William P. Banks, Baars has edited the journal Consciousness and Cognition for more than
468
William James on the Mind and Its Fringes
fifteen years (from Academic Press/Elsevier). With Nicole M. Gage, Baars has written an introductory text for cognitive neuroscience, called Cognition, Brain and Consciousness: An Introduction to Cognitive Neuroscience. (Baars and Gage, Eds. San Diego,: Elsevier/Academic Press, 2007). Baars was founding president of the Association for the Scientific Study of Consciousness and has an ongoing research collaboration for large-scale cognitive modeling with professor Stan Franklin (University of Memphis, Institute for Intelligent Systems). Consciousness: An Introduction to Cognitive Neuroscience. (Baars and Gage, Eds. San Diego, Calif.: Elsevier/Academic Press, 2007). Baars was founding President of the Association for the Scientific Study of Consciousness and has an ongoing research collaboration for large-scale cognitive modeling with professor Stan Franklin (University of Memphis, Institute for Intelligent Systems).
Subject Index Notes Cross-reference terms in italics are general cross-references, or refer to subentry terms within the main entry (the main entry is not repeated to save space). Readers are also advised to refer to the end of each article for additional cross-references - not all of these cross-references have been included in the index cross-references. The index is arranged in set-out style with a maximum of three levels of heading. Major discussion of a subject is indicated by bold page numbers. Page numbers suffixed by T and F refer to Tables and Figures respectively. vs. indicates a comparison. This index is in letter-by-letter order, whereby hyphens and spaces within index headings are ignored in the alphabetization. For example, acid rain is alphabetized after acidity, not after acid(s). Prefixes and terms in parentheses are excluded from the initial alphabetization. Where index subentries and sub-subentries pertaining to a subject have the same page number, they have been listed to indicate the comprehensiveness of the text. A abnormal states of mind See psychopathology aboulia 2:117–118 absent tasks 2:66 Absolute Unitary Being 2:273–274 absorption (hypnotic suggestibility) 1:356 access consciousness basic concepts 1:144, 1:158 characteristics 1:85, 1:86t, 2:161 conscious content 1:285 neurobiological theories 2:95 accurate reports 1:27–28 acetylcholine 1:305f, 1:308, 2:103, 2:107, 2:277f, 2:282 acetylene 1:296 action slips 1:85, 2:65 activation spreading 1:435 acute dystonia 2:267 acute stress reactions 2:253t Adaptive Control of Thought (ACT) theory 1:137 adenosine 2:104 adjustment reaction disorders 2:253t Adler, Alfred 2:241 adrenergic system 1:297–298, 2:108 adrenocorticotropic hormone (ACTH) 2:281 aesthetics 1:1–7 aesthetic perception 1:2 artist’s responses 1:2 consciousness factors 1:3 face perception 1:2 historical background 1:1–2 neuroaesthetics 1:2 neuroimaging research 1:5 theoretical perspectives 1:1 affective blindsight 1:233 affective responses affective attributions feelings-as-information model 1:237 implicit explicit affective attributions 1:238–239 metaconsciousness 1:239, 2:37 mood manipulation 1:237, 2:431 binding problem 1:235 conscious versus unconscious feelings 1:236 general discussion 1:232 neuroimaging research 1:234, 2:382 subliminal affective priming 1:233, 2:427–428 unconscious stimuli 1:233, 2:418
African grey parrots 1:27 Agassiz, Luis 2:460, 2:466 agnosia characteristics 2:258 cognitive disorders 2:264–265 sensorimotor pathways 1:176 visual agnosia 2:92, 2:124–125, 2:124f, 2:125f agoraphobia 2:251–252, 2:252t agraphaesthesia 2:264–265 AIM model of altered states of consciousness 2:218 akathisia 2:255, 2:266 akinetopsia 1:63, 2:94 alcohol classifications and characteristics 2:221, 2:224t hallucinations 2:260 intoxicated behaviors 1:86–87, 2:262t involuntary movements 2:267 Korsakoff ’s syndrome 2:2 psychiatric diagnoses 2:252t time perception disorders 2:262 alertness 2:103 See also selective attention alien control delusions 1:286, 1:364–365, 2:258t, 2:260t Alkire, Michael 1:310 allocentric neglect 2:71–72 allocentric representations 1:190, 2:9 Allport, DA 2:332 alogia 2:256t alpha power 1:439 altered states of consciousness 1:9–21 basic concepts change in overall pattern of experience 1:10 misrepresentations of experiences 1:11, 2:4 normal state of consciousness (NSC) 1:9 recognizable change in overall pattern of experience 1:11 workable definition 1:12 brain-injured patients 2:367, 2:368f coma 2:367, 2:368f exceptional/higher states of consciousness characteristics 1:15 cosmic consciousness 1:19 enlightenment 1:19–20 flow experiences 1:16 lucid dreaming 1:16 meditation 1:15 mystical experiences 1:19 near-death experiences (NDEs) 1:18 optimal experiences 1:16
469
470
Subject Index
altered states of consciousness (continued) out-of-body experiences (OBEs) 1:17 psychoactive drugs 2:220–221, 2:225 runner’s high 1:16 general discussion 1:9, 1:20 hypnosis 1:14, 1:353 locked-in syndrome (LIS) 2:370 pain assessment behavioral scales 1:244, 1:244t, 1:245t ethical considerations 1:247 general discussion 1:243, 1:246, 1:246f, 1:248 psychoactive drugs 2:217–229 AIM model 2:218 color perception 2:226, 2:227f definitions 2:217 drug classifications 2:221 form constants 2:226 general discussion 2:228 historical background 2:217 phenomenological characteristics 2:223 research concerns 2:217 transient hypofrontality theory 2:219, 2:220f recovered states 2:369 religious/spiritual experiences 2:273 sleep disturbances 2:369 sleep-related experiences 1:12, 2:248 trances 1:353 vegetative states 2:369 Alzheimer’s disease 2:1, 2:247, 2:262t, 2:263–264, 2:414 ambiguous motion stimuli 1:94f, 1:96 ambitendence 2:266 amelia 2:290 American Sign Language (ASL) 1:455 alpha(a)-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) 1:308–309 amnesia altered states of consciousness 1:12 amnesia confabulation correlations 2:8 anesthetic effects 1:299–300 anterior communicating artery aneurysms 2:3 anterograde amnesia 1:12, 1:79, 1:189, 2:3, 2:14, 2:262, 2:414 autobiographical memory 1:79 bllindness amnesia comparisons 1:57 conscious awareness 1:186 dissociative amnesia 2:414 explicit memory 1:184, 1:189, 2:414 hypnosis 1:352, 2:414 implicit memory 1:184, 1:370 Korsakoff ’s syndrome 2:264 retrograde amnesia 1:12, 1:79, 1:189, 2:3–4, 2:15, 2:262, 2:414 skill acquisition and learning 2:14 time perception disorders 2:262 See also sleep amphetamines classifications and characteristics 2:221–222, 2:224t color perception 2:227f psychiatric diagnoses 2:252, 2:252t self-harm risks 2:255 amphibians 2:122 amygdala affective blindsight 1:233 affective response analysis 1:234, 2:382 dreaming 2:363f emotional evaluation processes 2:363–364, 2:381 mammalian avian comparison research 1:28–29 memory functions 1:185f, 2:7 pain perception 1:246 placebo response analysis 2:213 religious/spiritual experiences 2:277f, 2:280, 2:284 sleep regulation 2:359f, 2:363–364 visual systems 2:123 voluntary action 1:124f anagrams 1:443 analgesia pain treatment 1:247 placebo treatments 2:205, 2:212 analysis-based problem solving 1:433 anarchic hand syndrome 2:267
anatomical location theories 1:95 Anaxagoras 1:340 anemia 2:262t anesthesia 1:295–313 anesthetic agents characteristics 1:296 coma 2:101 neuroimaging research 1:307f psychoactive drugs 2:221 arousal system anesthetic effects 1:301, 1:302f cortical regulation 1:304, 1:305f global versus local theories 1:304, 1:306f historical research 1:302 cerebral metabolism 1:152f dose-related effects amnesia effects 1:299–300 early ether vaporizer 1:298, 1:299f historical research 1:298 immobility 1:300 passage of time 1:300 unconsciousness 1:299 electrical activity 1:301, 1:302f general discussion 1:295 historical research 1:298 hypnosis 1:353–354 hysterical symptoms 2:233 illustrative diagram 1:148f inhaled agents 1:296, 1:307f molecular mechanisms 1:296 neuroimaging research 1:306, 1:307f region-specific interactions 1:307, 1:309f thalamocortical switch hypothesis 1:306, 1:307f, 1:310 theoretical perspectives Anesthetic Cascade theory 1:311 cognitive unbinding theory 1:311 endogenous London forces 1:311 information disintegration theory 1:311 information integration theory (IIT) 1:311 information processing theory 1:310 thalamocortical switch hypothesis 1:310 aneurysms 2:3 angel dust (PCP) classifications and characteristics 2:222, 2:224t phenomenological characteristics 2:225, 2:226 animal consciousness 1:23–36 introspective awareness 2:194–195 mirror test 1:24 philosophical theories 1:349 phylogenetic definitions general discussion 1:24 natural selection theory versus behaviorism 1:25 sensory higher-order distinctions 1:24 research methodology birds 1:27 cephalopods 1:29 commentary key paradigm 1:26 consciousness correlates 1:25 evidentiary support 1:26 homologous brain structures 1:27 human consciousness-based benchmarks 1:25 universal properties 1:33 visual recognition studies 1:24 animal magnetism 1:353, 2:235 Anna O. 2:236 anoetic consciousness 1:186 anomalous monism theory 1:348, 2:46 anorexia nervosa 2:258t anosognosia 2:72, 2:264–265 anterior cingulate cortex (ACC) automatic controlled processing research 1:90, 1:90f creativity 1:438 fringe consciousness 2:461 habit formation processes 1:320 intergroup processing 2:384 memory functions 1:184, 1:185f, 1:186f mind wandering research 2:67 motor-skill learning 2:15
Subject Index
pain perception 1:246, 1:246f placebo response analysis 2:206, 2:211, 2:213 religious/spiritual experiences 2:277f, 2:278, 2:284 REM sleep 2:360, 2:363f selective memory retrieval 1:209, 1:212f anterior communicating artery aneurysms 2:3 anterior insular cortex (AIC) 1:90, 1:90f anterior superior temporal gyrus (aSTG) 1:441 anterograde amnesia 1:12, 1:79, 1:189, 2:3, 2:14, 2:262, 2:414 antidepressants 2:222 antihistamines 2:221 antipsychotic medications 2:222 antisocial personality disorder 2:269t anxiety states 2:254 apparent mental causation theory 1:141 apparent motion quartet 1:93, 1:94f, 1:96 appetite suppressants 2:221–222 apraxia 2:265 Archimedes 1:431 arginine vasopressin 2:276t, 2:281 Aristotle aesthetics 1:1 consciousness studies 1:330 mental imagery 2:447 mind body problem 1:341 sleep studies 2:358 art and aesthetics 1:2, 1:4 articulation suppression 1:391–392 artificial general intelligence (AGI) 1:38 artificial grammar 1:372, 1:372f, 1:374f artificial-grammar learning 2:415 artificial intelligence (AI) 1:37–46 behavioral tests 1:43 ethical and legal issues 1:44 general discussion 1:37, 1:44 internal architecture tests 1:43 machine consciousness behavioral tests 1:43 criticisms algorithmic factors 1:40 Chinese Room experiment 1:40 hard problem of consciousness 1:40 limitations of artificial intelligence 1:41 current research associated cognitive characteristics 1:41 associated external behaviors 1:41 consciousness correlates 1:42 future developments 1:42 global workspace theory 1:42 Internet connectivity 1:42 phenomenally conscious machines 1:42 internal architecture tests 1:43 positive theories cognitive theories 1:39 higher-order thought theory 1:40 information integration theory (IIT) 1:39 inner speech 1:40 natural selection theory 1:39 research background 1:37 theories of consciousness global workspace theory 1:38 model approaches 1:38 ascending reticular activating system (ARAS) 1:302, 1:308, 2:357 Assessment of Discomfort in Dementia Protocol (ADD) 1:244t associationism basic concepts 2:232 sensory-motor reflex physiology 2:232, 2:233 astereognosia 2:264–265 athetoid movements 2:267 athymhormia 2:117–118 atomism 1:340 atropine 2:222, 2:224t attention affective responses 1:233–234 attentional manipulation 1:50, 1:55 attention consciousness relationship 2:96, 2:144 automatic attention response 1:87 change blindness 1:47–59
bllindness amnesia comparisons 1:57 change blindness inattentional blindness comparisons 1:56 change difference concepts 1:49 change motion concepts 1:49 detection 1:49 experimental approaches attentional manipulation 1:50 change detection task 2:440–441 general discussion 1:49 perceptual set types 1:50 response types 1:50 task types 1:49 general discussion 1:47, 1:58 identification 1:49 localization 1:49 research background 1:47, 1:48f research implications first-order representational theory 1:162 general discussion 1:51 scene perception 1:51 visual attention short-term memory relationship 1:51, 2:441 selective attention 1:67 sensory-motor theory of consciousness 1:141 spatiotemporal distinctions 1:48 subjective experiences 2:395, 2:398–399 visual attention visual experience comparisons 1:58 diffuse attention 1:442–443 inattentional blindness 1:47–59 bllindness amnesia comparisons 1:57 change blindness inattentional blindness comparisons 1:56 experimental approaches attentional manipulation 1:55 general discussion 1:54 perceptual set types 1:55 response types 1:54 task types 1:54 expression suppression relationship 1:53 general discussion 1:47, 1:58 implicit perception 2:417 research background 1:52, 1:53f research implications general discussion 1:55 scene perception 1:56 visual attention 1:56 restricted unrestricted effects comparisons 1:53 selective attention 1:67 visual attention visual experience comparisons 1:58 mind wandering 2:57–69 basic concepts 2:33, 2:57 causal theories attentional control 2:62f, 2:64 current concerns 2:62f, 2:63 general discussion 2:61 task environment influences 2:62, 2:62f consequences action slips 2:65 general discussion 2:64 mindless reading 2:63, 2:64 future research directions 2:68 information flow attentional transitions 2:59f, 2:60 general discussion 2:58 mental states 2:58, 2:59f schematic diagram 2:59f mechanisms ironic processing theory 1:216, 2:60 low-level versus higher-level cognitive processes 2:60, 2:329 metaconsciousness/meta-awareness 2:33, 2:35, 2:61 neuroimaging research absent tasks 2:66 control mechanisms 2:67 default mode hypothesis 2:67 general discussion 2:66 selective attention 1:61–75 attention awareness relationship attentional blink 1:67, 1:234 awareness of attention 1:71 awareness of awareness 1:70, 1:71f
471
472
Subject Index
attention (continued) blindsight 1:69, 1:69f change blindness 1:47–59, 1:67, 1:86, 1:233–234, 2:85, 2:96, 2:167 general discussion 1:66 gist situations 1:68 inattentional blindness 1:47–59, 1:67, 1:233–234, 2:85, 2:167, 2:335 neural mechanisms 2:84 neurochemical mechanisms 2:103 otherwise-engaged attention 1:68 postattentive awareness 1:70 basic concepts 1:61 binding problem 1:62, 1:235 creativity 1:442–443 general discussion 1:71 neurochemical mechanisms 2:103 object recognition feature characteristics 1:65, 1:65f feature integration theory 1:63 feedforward models 1:65 reverse hierarchy theory (RHT) 1:66 unconscious conscious perception 2:144 unconscious information processing 2:171 attentional blink attention awareness relationship 1:67, 1:234 implicit perception 2:417 unconscious information processing 2:167, 2:171 auditory hallucinations 1:397, 2:249, 2:259 auditory stream segregation 2:332–333, 2:333f Austen, Jane 2:189 autism 1:398 autobiographical memory 1:77–82 characteristics 2:1 cognitive feelings familiarity 1:77–78 knowing 1:78 remembering 1:77 conscious feelings 1:79 definition 1:77 egocentric representations 2:9 imagined feelings 1:79 inner speech 1:396 Korsakoff ’s syndrome 2:264 malfunctions amnesia 1:79 confabulations 1:80, 2:3, 2:264 de´ja` vu 1:80–81 false memories 1:80, 2:4 neuropsychology 2:2 autoimmune diseases 2:262t automatic attention response 1:87 automatic thoughts and actions 1:83–92 automatic controlled processing comparisons (dual processing theory) automatic attention response 1:87 basic concepts 1:84, 2:413 brain systems 1:90, 1:90f conscious monitoring 1:89 consciousness dual processing relationship 1:85, 1:86t functional role 1:87 general discussion 1:91 information flow 1:89 neural/psychological correlates 2:420 theories of consciousness 1:137 theory of mind research 2:405, 2:406t, 2:407t top-down conscious processing 1:89–90, 2:96 general discussion 1:83 habits attentional control changes 1:320 dual-task interference 1:320 explicit learning research 1:323 general discussion 1:319 goal-directed skills 1:322, 2:426 implicit learning research 1:322 intention behavior prediction connection 1:321 neuroimaging research 1:320 implicit social cognition 1:383 perceptual awareness 2:420 autonoetic consciousness 1:77, 1:188 autonomic arousal 2:253t, 2:254
autonomic nervous system (ANS) 2:281, 2:282, 2:284 autopilot situations 1:86–87 autoscopic hallucinations 2:259 autotopagnosia 2:290 avian brain research 1:27 avoidant personalities 2:252t, 2:269t awareness affective responses 1:233–234 blindsight 1:119, 2:141 contents of awareness 2:61 definition 1:147, 1:148f folk theory of mind concept 1:253, 1:253f linguistic processes 1:449 metaconsciousness/meta-awareness 2:33–41 affective attributions 1:239, 2:37 automatic thoughts and actions 1:88–89 basic concepts 2:33 dissociation theories general discussion 2:35 temporal dissociations 2:35 translation dissociations 2:38 general discussion 2:40, 2:375 historical research 2:34 mindfulness meditation 2:39 monitoring processes 2:34 temporal dissociations flow experiences 2:37 general discussion 2:35 mind wandering 2:33, 2:35, 2:61 recovered memories 2:37 unwanted thoughts 2:36, 2:67 verbal overshadowing 2:38 neural mechanisms 2:82 neurochemical mechanisms 2:103 perception awareness research 1:206 research challenges 1:375 response override hypothesis 1:205–219 basic concepts 1:205, 1:206f general discussion 1:217 perception awareness research 1:206 retrieval stopping processes Freudian suppression model 1:214 individual differences 1:214 intrusive memory awareness 1:215 neurobiological theories 1:212 Think/No-Think paradigm (TNT) 1:210, 1:211f, 1:212f, 1:213f thought suppression 1:216 working memory control mechanisms 1:216 selective retrieval inhibition mechanisms 1:209 inhibitory control processes 1:207 neurobiological theories 1:209 retrieval-induced forgetting (RIF) 1:206f, 1:208 retrieval stopping processes 1:210 selective attention 1:61–75 attention awareness relationship attentional blink 1:67, 1:234 awareness of attention 1:71 awareness of awareness 1:70, 1:71f blindsight 1:69, 1:69f change blindness 1:47–59, 1:67, 1:86, 1:233–234, 2:85, 2:96, 2:167 general discussion 1:66 gist situations 1:68 inattentional blindness 1:47–59, 1:67, 1:233–234, 2:85, 2:167, 2:335 neural mechanisms 2:84 neurochemical mechanisms 2:103 otherwise-engaged attention 1:68 postattentive awareness 1:70 basic concepts 1:61 binding problem 1:62, 1:235 general discussion 1:71 neurochemical mechanisms 2:103 object recognition feature characteristics 1:65, 1:65f feature integration theory 1:63 feedforward models 1:65 reverse hierarchy theory (RHT) 1:66
Subject Index
self-awareness 2:187–199 consciousness development processes 1:223, 1:224 dream content analysis 2:343 inner speech 1:398 phenomenal consciousness 2:161 situatedness 2:187 voluntary action 2:118 sleep states 2:365 visual experience 2:435–443 acetylcholine 2:107 binding problem 2:157 brain processing paradox 2:441, 2:442f change blindness 2:441 general discussion 2:435, 2:442 neural mechanisms 2:75–86 attention awareness relationship 2:84 binocular rivalry 2:78 conscious content 2:79 cortical regions 2:82 discrimination tasks 2:76–77 encoding processes 2:79 general discussion 2:75, 2:85 perceptual threshold 2:75 reversible figures 2:78 stimuli processing 2:75 visual masking 2:77, 2:92–93, 2:166f, 2:167 perception action frames of reference 2:127 rich experience impoverished input paradox color perception 2:438, 2:438f edge alignment 2:437, 2:437f edge enhancement 2:436, 2:436f edge-filling abilities 2:437, 2:438f face recognition 2:439–440 feature exaggeration 2:439 gravity and light constraints 2:438, 2:439f Hermann grid 2:436–437, 2:436f heuristic processes 2:438, 2:439f prior knowledge 2:439 psychophysical research 2:435 rich experience limited report paradox 2:440 serotonin 2:104 virtual reality processes 2:106
B Baars, Bernard automatic thoughts and actions 1:88 conscious control action planning connection 1:180–181 global workspace theory 1:38, 1:137, 1:158–159, 1:288, 2:58–59 hard problem of consciousness 2:88 unconscious conscious processing 1:379 backward masking 2:77, 2:170 bacteria 2:122, 2:262t bad dreams 1:14, 2:345 Bailenson, Jeremy 2:376 Balint Holmes syndrome binding problem 2:150, 2:150f illusory conjunctions 1:64, 2:73 neglect manifestations 2:73 selective attention 1:64 sensorimotor pathways 1:176 barbiturates 2:221 Bargh, John 2:376 basal forebrain arousal system 1:304, 1:305f dreaming 2:360, 2:363f memory functions 2:7 sleep regulation 1:308 basal ganglia dementia 2:264 implicit memory 1:188 memory capacity 2:337 neglect 2:71 religious/spiritual experiences 2:279 visual systems 2:123, 2:127 voluntary action 1:124, 1:124f, 1:125f
Bastian, Charlton 2:111 Baumgarten, Alexander 1:1–2 Bayesian decision theory 1:143 beauty See aesthetics bee dance 1:425 beginning and ending of life 2:308 behaviorism action control 1:173 animal consciousness 1:25 basic concepts 1:333, 1:344 consciousness studies 1:135, 2:412 criticisms 1:334 free will theory 2:464 opposing viewpoints 1:335 Skinner’s theories 1:335 behavior priming methods automatic social behavior 1:385 evaluative priming 1:386 habit-formation processes 1:323, 1:325 subliminal affective priming 1:233, 2:427–428 Behcet’s syndrome 2:262t ‘‘being for’’ concept 2:390 beliefs and desires basic concepts 1:417 theory of mind research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t event-related potentials (ERPs) 2:407 executive control 2:403 false beliefs 2:401, 2:402f folk theories 1:252, 1:253f, 1:256, 2:188–189 general discussion 2:408 mirror neurons 2:408 neuroimaging research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 language 2:401, 2:402f personality traits 2:405 social cognition 2:404 personality traits 2:405 social cognition 2:404 Benjamin Franklin Commission 1:353 benzodiazepines 2:221 Bergson, Henri 2:106, 2:465 Berinmo language 1:457 Berkeley, George 1:343, 2:448 Bernheim, Hippolyte 1:354, 2:231 beta(b)-endorphins 2:276t, 2:281 betel nut 2:221–222 bicameral mind 1:399 Bigelow, Henry Jacob 1:298 Bilgrami, Akeel 2:194 binding problem 2:147–158 action control 2:152 Balint Holmes syndrome 2:73 basic concepts 2:147, 2:148f crossmodal binding 2:151 dream content analysis 2:345 general discussion 2:157 multisensory integration 2:151 visual awareness 2:157 visual systems affective response analysis 1:235 Balint Holmes syndrome 2:150, 2:150f Livingstone Hubel model 2:151 neural encoding mechanisms 2:83–84 psychophysical research 2:148, 2:148f selective attention 1:62 spatiotemporal distinctions 2:149 synesthesia 2:150 theoretical solutions convergence coding 2:153, 2:153t feature integration theory 2:155, 2:155f general discussion 2:152
473
474
Subject Index
binding problem (continued) population coding 2:153, 2:153t recurrent processing model 2:156 synchrony model 2:153t, 2:154 Binet, Alfred 2:231 binge eating/drinking 2:266–267 binocular rivalry bistable perception functional magnetic resonance imaging (fMRI) 1:101 neural correlates 1:100, 1:101 theoretical models 1:99 visual stimuli 1:94f, 1:97 human consciousness studies 1:25 selective attention 1:63 unconscious information processing basic concepts 2:165, 2:166f postretinal processing 2:169 research background 2:92–93 visual competition studies 2:78, 2:128–129 visual processing 2:128–129 Binswanger’s disease 2:262t biographical memory 2:9 biosemantics (Millikan) 1:425 bipolar disorder 2:108, 2:253t birds 1:27 bistable perception 1:93–106 basic concepts 1:93, 1:94f first-person skepticism 2:397 general discussion 1:104 neural correlates binocular rivalry 1:100, 1:101 cortical/subcortical processing 1:101 extrastriate cortex 1:101 general discussion 1:99 measurement techniques 1:99 motion perception 1:101, 1:102f neural synchronization 1:104 parietal and prefrontal cortices 1:98f, 1:103 reversal-related activations 1:101, 1:102f theoretical perspectives 1:95 theoretical models behavioral evidence 1:97 binocular rivalry 1:99 high-level theories 1:98, 1:98f low-level theories 1:97, 1:98f theoretical perspectives anatomical location theories 1:95 consciousness correlates 1:94 neural correlates 1:95 state-change theories 1:95 visual stimuli ambiguous motion 1:94f, 1:96 binocular rivalry 1:97 gamma distribution 1:94f, 1:96 reversible figures 1:96 bizarreness 2:344 blackboard architecture 1:38 blackouts 2:262 blepharospasms 2:267 blindness functional (hysterical) blindness 1:363, 2:417 motion-induced blindness 2:166, 2:170 blindsight 1:107–122 access consciousness 2:161 affective blindsight 1:233 attention awareness relationship 1:69, 1:69f basic concepts 1:107, 2:127 conscious awareness 1:119, 2:390 implicit perception 2:417 implicit processing 1:114, 1:115f occurrences 1:115 pupillometry 1:114, 1:115f rehabilitation strategies 1:120 research background early human evidence 1:108, 1:109f early nonhuman primate evidence 1:109 follow-up research 1:113
recent human evidence 1:112, 1:112f, 1:113f recent nonhuman primate evidence 1:111 research implications controversial issues 1:116, 1:161 decision criteria/response bias 1:117 experimental approaches 1:116 normal, degraded vision comparisons 1:118 residual islands/tags 1:117 sensorimotor pathways 1:176 stimulus parameters 1:116 subjective threshold approach 2:141 unconscious information processing 1:284, 2:168 visual pathways 1:119, 1:119f voluntary action 1:287 blind spot 2:107, 2:462–463 blink-contingent research techniques 1:50 Block, Ned consciousness intentionality relationship 1:419 Inverted Earth concept 2:26–27 phenomenal consciousness basic concepts 1:158, 2:95 characteristics 1:85 conscious content 1:285 subjective properties 1:144, 2:390 philosophical theories 1:346 voluntary action 1:287 blood-oxygen level dependent (BOLD) signal 1:100, 1:129 Blumer, Herbert 2:376 blunting disorders 2:253t, 2:254 bodily hallucinations 2:260t body awareness 2:289–300 affective awareness 2:290 conceptual awareness 2:290 direct awareness 2:293 general discussion 2:289, 2:299 high-level body awareness 2:290 low-level body awareness basic concepts 2:290 body-internal spatial coding 2:291 moment-to-moment representation 2:291 neurological mechanisms 2:292 object-relative spatial coding 2:291 orientational coding 2:292 proprioception 2:292 proportional awareness 2:293 self-specifying information 2:292, 2:295–296 self-world dualism 2:296 semantic awareness 2:290 somatic proprioception 2:295 spatial perceptual frames of reference 2:296 visual perception 2:292 body-parts knowledge 2:290 body-relative information 2:290 bonobos 1:27 borderline personality disorders 2:254, 2:269t Boring, Edwin G 1:335 Bornstein, Robert 2:376 botanism 2:460 bowerbirds 1:1 Braid, James 1:353–354 brain aneurysms 2:3 brain death cerebral metabolism 1:152f clinical definitions 1:148 historical research 1:148 pain perception 1:246f brain imaging techniques See neuroimaging techniques brain lesions blindsight 1:108, 1:109f, 1:112, 1:176, 2:417 brain death 1:148 cognitive disorders 2:264 encoding processes 2:79 intergroup processing studies 2:380 involuntary movements 2:267 memory functions 2:7 microconsciousness theory 2:94
Subject Index
neglect 1:58, 2:71 pathological disorders 2:117 perceptual awareness studies 2:82 skill acquisition and learning studies 2:14 theory of mind research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 false beliefs 2:401, 2:402f language 2:401, 2:402f personality traits 2:405 social cognition 2:404 brain processing paradox 2:441, 2:442f brainstem brainstem auditory-evoked potentials (BAEPs) 1:151 visual systems 2:127 brain teasers 1:440 Bregman, Albert 2:332–333 Brentano, Franz intentionality basic concepts 1:419 intentional inexistence analytical approaches 1:421 biosemantics (Millikan) 1:425 Brentano’s thesis 1:423 higher-order thought theory 1:426 indicator semantics (Dretske) 1:424 language structure analysis 1:421 mental states 1:420 Representation Theory of Mind (RTM) 1:423 mental states 1:158 mind body problem 1:343 Breuer, J 2:234, 2:236 Broadbent, Donald 1:136, 2:65, 2:328, 2:413 Broad, CD 1:197 broad content 2:27 Broca’s area See left inferior frontal gyrus (LIFG) Bufo bufo 1:62 Burge, Tyler 1:347 Burt, Cyril 1:334 buspirone 2:221
C caffeine classifications and characteristics 2:221–222 phenomenological characteristics 2:220f psychiatric diagnoses 2:252, 2:252t candle problem 1:433, 1:434f, 1:444f Cannabis sativa 2:221 capacity (attentional control) 2:413 Capgras syndrome 1:223, 2:188, 2:257 capture control 1:55 carbon monoxide poisoning 2:262t cardinal cells 2:153 card sorting tests 2:6 Cartesian dualism 1:342, 2:45, 2:181, 2:452 castration anxiety 2:241 catastrophic reaction disorders 2:253t catatonia 2:266 catecholamines 2:102 categorical perception 1:454 cathartic therapy 2:237 cathexis 2:234 central executive system of consciousness 1:136 central medial thalamus 1:308, 1:309f central nervous system (CNS) 2:248–249 central pattern generators (CPGs) 1:32 central tegmentum 2:108–109 centrencephalic integrating system 1:305, 1:306f cephalopod consciousness 1:29 cerebellum creativity 1:438 implicit memory 1:188 mammalian avian comparison research 1:28 motor-skill learning 2:15
proprioception 2:292 voluntary action 1:124, 1:124f, 1:125f cerebral cortex aesthetic responses 1:5 alpha power 1:439 anesthetic effects 1:302 cortical arousal 1:436, 1:439 habit formation processes 1:320 mammalian avian comparison research 1:28 memory functions 1:184 neglect 2:71 NREM (non-REM) sleep 2:359 pain perception 1:246 placebo response analysis 2:206, 2:213 proprioception 2:292 REM sleep 2:362 visual systems 2:122, 2:123f cerebrum fission factor 2:307 psychological continuity 2:306 Chalmers, David hard problem of consciousness 1:144, 1:159, 2:375 naturalistic dualism 2:54 change blindness 1:47–59 attention awareness relationship 1:67, 1:86, 1:233–234, 2:85, 2:96 bllindness amnesia comparisons 1:57 change blindness inattentional blindness comparisons 1:56 change difference concepts 1:49 change motion concepts 1:49 detection 1:49 experimental approaches attentional manipulation 1:50 change detection task 2:440–441 general discussion 1:49 perceptual set types 1:50 response types 1:50 task types 1:49 general discussion 1:47, 1:58 historical research 1:48f identification 1:49 localization 1:49 research background 1:47, 1:48f research implications first-order representational theory 1:162 general discussion 1:51 scene perception 1:51 visual attention short-term memory relationship 1:51, 2:441 sensory-motor theory of consciousness 1:141 spatiotemporal distinctions 1:48 subjective experiences 2:395, 2:398–399 unconscious information processing 2:167 visual attention visual experience comparisons 1:58 Charcot, Jean-Martin 1:354, 2:231, 2:233 Charles Bonnet syndrome 2:260 Chase and Sanborn problem 1:3 Checklist of Nonverbal Pain Indicators (CNPI) 1:244, 1:244t childhood sexual instinctual drives 2:239 Children’s Hospital of Eastern Ontario Pain Scale (CHEOPS) 1:244t chimpanzees visual recognition studies 1:24 vocabulary acquisition 1:27 Chinese Room experiment 1:40, 1:453 Chisholm, Roderick 1:420, 1:423 chloroform 1:296 chlorpromanzine 2:222 choice blindness paradigm 1:287–288 choice probability 2:76–77 cholinergic psychedelics 2:222 cholinergic system 1:308, 2:102, 2:360 Chomsky, Noam 1:450, 2:413–414 choreiform movements 2:267 chronic fatigue syndrome (CFS) 2:117–118 chronic goals 2:424–425, 2:429 chronic respiratory failure 2:262t cingulate cortex 1:124, 1:124f, 1:125f cingulate gyrus 2:278, 2:284 Circle of Willis 2:3 clang associations 2:256t
475
476
Subject Index
classical conditioning 2:1 Cleeremans, Axel 1:142 Clifford, William 1:343 clinical delusions 1:363 clinical unconsciousness 2:235 closed-loop process 2:13 clozapine 2:222 cocaine classifications and characteristics 2:221–222, 2:224t physical hallucinations 2:261 self-harm risks 2:255 cocktail-party effect 2:429 coconscious 2:235 codeine 2:222 cognitive contents 2:160–161 cognitive processes altered states of consciousness 2:219 autobiographical memory familiarity 1:77–78 knowing 1:78 remembering 1:77 automatic thoughts and actions 1:83–92 automatic controlled processing comparisons (dual processing theory) automatic attention response 1:87 basic concepts 1:84, 2:413 brain systems 1:90, 1:90f conscious monitoring 1:89 consciousness dual processing relationship 1:85, 1:86t functional role 1:87 general discussion 1:91 information flow 1:89 neural/psychological correlates 2:420 spatial neglect 1:87 theories of consciousness 1:137 theory of mind research 2:405, 2:406t, 2:407t top-down conscious processing 1:89–90, 2:96 general discussion 1:83 habits attentional control changes 1:320 dual-task interference 1:320 explicit learning research 1:323 general discussion 1:319 goal-directed skills 1:322 implicit learning research 1:322 intention behavior prediction connection 1:321 neuroimaging research 1:320 implicit social cognition 1:383 cephalopods 1:29 Cognitive Revolution 1:135, 1:335, 2:389, 2:413 hypnotic phenomena 1:354–355, 1:355–356, 1:358 implicit social cognition 1:383–388 automatic thoughts and actions 1:383 general discussion 1:387 social constructs attitudes 1:384 automatic social behavior 1:385 construct formation and change processes 1:385 general discussion 1:384 goal pursuit 1:385 measurement techniques 1:386 self-concept 1:385 stereotypes 1:384 intentionality 1:417–429 basic concepts 1:419 behavior/reason explanations 1:254 causal history factors 1:254–255 consciousness definitions 1:419 consciousness intentionality relationship 1:417, 2:22, 2:179 embodied intentionality 1:428 folk theories basic concepts 1:253, 1:253f behavior/reason explanations 1:254 language of mental activities 1:255 observability 1:256f Husserl’s theory 1:428 intentional inexistence analytical approaches 1:421 biosemantics (Millikan) 1:425
Brentano’s thesis 1:423 higher-order thought theory 1:426 indicator semantics (Dretske) 1:424 language structure analysis 1:421 mental states 1:256, 1:420 Representation Theory of Mind (RTM) 1:423 Merleau-Ponty’s theory 1:428, 2:181 phenomenal consciousness 1:427 phenomenological characteristics 2:179 voluntary action 1:253, 1:253f, 2:115 linguistic processes 1:450, 2:413–414 mental imagery 2:445–457 background information 2:446 definitions 2:446 general discussion 2:456 historical research 2:447 subjective individual differences 2:449 theoretical perspectives description (mentalese) theory 2:453 enactive theory 2:454 general discussion 2:450 picture theory 2:451, 2:451f mind wandering 2:57–69 basic concepts 2:33, 2:57 causal theories attentional control 2:62f, 2:64 current concerns 2:62f, 2:63 general discussion 2:61 task environment influences 2:62, 2:62f consequences action slips 2:65 general discussion 2:64 mindless reading 2:63, 2:64 future research directions 2:68 information flow attentional transitions 2:59f, 2:60 general discussion 2:58 mental states 2:58, 2:59f schematic diagram 2:59f mechanisms ironic processing theory 1:216, 2:60 low-level versus higher-level cognitive processes 2:60, 2:329 metaconsciousness/meta-awareness 2:33, 2:35, 2:61 neuroimaging research absent tasks 2:66 control mechanisms 2:67 default mode hypothesis 2:67 general discussion 2:66 neurobiological theories 2:87 neurocognitive panpsychism 2:98 number perception 1:457–458, 1:458f psychoactive drugs 2:225 psychopathology basic concepts 2:261 delirium 2:261, 2:262t dementia 2:261, 2:262t identity perception 2:263 insight 2:265 memory 2:263, 2:263t place perception 2:262–263 self-perception 2:264 time perception 2:262 visuospatial information 2:264 response override hypothesis 1:205–219 basic concepts 1:205, 1:206f general discussion 1:217 perception awareness research 1:206 retrieval stopping processes Freudian suppression model 1:214 individual differences 1:214 intrusive memory awareness 1:215 neurobiological theories 1:212 Think/No-Think paradigm (TNT) 1:210, 1:211f, 1:212f, 1:213f thought suppression 1:216 working memory control mechanisms 1:216 selective retrieval inhibition mechanisms 1:209 inhibitory control processes 1:207
Subject Index
neurobiological theories 1:209 retrieval-induced forgetting (RIF) 1:206f, 1:208 retrieval stopping processes 1:210 social cognitive neuroscience 2:379–388 behavior modification studies 2:386 categorization 2:380 categorization evaluation relationship 2:382 complex situations 2:385 evaluation processes 2:381 general discussion 2:386 intergroup processing 2:379 motivational effects 2:383 neuroimaging research 2:380 same-race memory bias 2:381 self-categorization 2:383 theoretical perspectives 2:380 subjective property 2:389–400 Cognitive Revolution 2:389 first-person knowledge 2:394 first-person skepticism 2:397 historical research 2:389 phenomenal consciousness 1:144, 2:390 scientific research 2:392 testimonies of consciousness 2:399 theories of consciousness artificial intelligence (AI) global workspace theory 1:38 machine consciousness 1:39 model approaches 1:38 cognitive processes 1:135–146 hard problem of consciousness 1:144 higher-order thought theory 1:426 historical research 1:135 illusory theories apparent mental causation theory 1:141 general discussion 1:140 multiple drafts model 1:140 influential theories general discussion 1:137 global workspace theory 1:137 information integration theory (IIT) 1:139 intermediate level theory 1:138 learning process theories general discussion 1:141 higher-order Bayesian decision theory 1:143 radical plasticity thesis 1:142 sensory-motor theory 1:141 precursor theories automatic controlled processing 1:137 central executive system 1:136 modularity 1:137, 1:451 supervisory attentional system 1:136 top-down/bottom-up cognitive processing 2:259 unconscious cognition 2:411–421 affective states 2:418 attentional control 2:413, 2:419 capacity limits 2:413 conscious shyness 2:419 historical research 2:412 implicit learning 2:415 implicit memory 2:414 implicit perception 2:416 implicit thought 2:418 neural/psychological correlates 2:420 variable/qualitative-based differences 1:379 visual system research 2:123 See also dreaming Cognitive Revolution 1:135, 1:335, 2:389, 2:413 color perception color-digit synesthesia 2:150 color phi phenomenon 1:198 externalist theories 2:27–28 Inverted Earth concept 2:26–27 linguistic processes 1:457 microconsciousness theory 2:94 neural encoding mechanisms 2:80 nineteenth-century studies 1:330
477
psychoactive drugs 2:226, 2:227f visual experience 2:438, 2:438f coma 1:147–156 anesthetics 2:101 behavioral coma scales 1:150, 1:151, 1:245t cholinergic system 2:102 clinical definitions 1:149 desferrioxamine/prochlorperazine doses 2:102 diagnostic criteria 1:150, 1:244 Glasgow Coma Scale (GCS) 1:150, 1:245t illustrative diagram 1:148f induction mechanisms 2:101 pain assessment analgesic use 1:247 behavioral coma scales 1:150, 1:151, 1:245t ethical considerations 1:247 general discussion 1:243, 1:248 neural correlates 1:246 pain perception 1:246, 1:246f sleep states 2:367, 2:368f Coma/Near Coma Scale 1:151, 1:245t Coma Pain Scale (CPS) 1:245–246 Coma Recovery Scale-Revised 1:245t coma recovery scale-revised (CRS-R) 1:151 command hallucinations 2:255 commentary key paradigm 1:26 commissurotomy 1:448, 2:302 common sense knowledge 1:37–38 common toad (Bufo bufo) 1:62 compatibilism definition 1:265 theoretical perspectives 1:266 compound remote associate (CRA) problems 1:438, 1:441 Comprehensive Drug Abuse and Prevention and Control Act (1970) 2:222–223 compulsive behaviors 2:118, 2:258t, 2:266–267 Comte, Auguste 2:399 conceptual implicit memory 1:186, 2:9 Condillac, E´tienne Bonnot de 1:61–62, 1:71 conditioning affective response analysis 1:238–239 placebo treatments brain mechanisms 2:209 endogenous opioid responses 2:206, 2:210 research controversies 2:211 stimulus response conditions 2:209 confabulations altered states of consciousness 1:12 amnesia confabulation correlations 2:8 anterior communicating artery aneurysms 2:3 basic concepts 2:5 domain specificity 2:8 dream content analysis 2:343 executive theories 2:5–6 false memories 2:4 hypnosis 2:4–5 irrelevant memory suppression 2:7 memory malfunctions 1:80, 2:3, 2:264 reality monitoring 2:6 retrieval theories 2:5–6 conscious content 1:94, 1:136, 1:284, 1:290, 2:79 conscious deliberation 1:287 conscious inessentialism 1:280 conscious level 1:284 consciousness access consciousness basic concepts 1:144, 1:158 characteristics 1:85, 1:86t, 2:161 conscious content 1:285 neurobiological theories 2:95 action control action-effect codes 1:178f afferent versus efferent action 2:113 basic concepts 1:177, 1:178f binding problem 2:152 conscious awareness 2:15, 2:111 conscious control action planning connection 1:179, 2:15 conscious will 1:173, 1:286, 1:286f, 2:111, 2:116 folk theory of mind concept 1:253–254, 1:253f
478
Subject Index
consciousness (continued) intentionality 1:253–254, 1:253f, 2:115 motor programs computer-programming analogy 1:173 sensorimotor processing 1:174, 2:111 sensory effects 1:172 pathological disorders 2:117 self-awareness theories 2:118 sensorimotor processing feedforward feedback connections 1:174 multiple-pathway models 1:175, 1:175f, 1:178f single-pathway models 1:175, 1:175f sequence learning 2:14 somatic markers 1:172–173 voluntary action 1:171 aesthetics 1:3 altered states of consciousness 1:9–21 basic concepts change in overall pattern of experience 1:10 misrepresentations of experiences 1:11, 2:4 normal state of consciousness (NSC) 1:9 recognizable change in overall pattern of experience 1:11 workable definition 1:12 exceptional/higher states of consciousness characteristics 1:15 cosmic consciousness 1:19 enlightenment 1:19–20 flow experiences 1:16 lucid dreaming 1:16 meditation 1:15 mystical experiences 1:19 near-death experiences (NDEs) 1:18 optimal experiences 1:16 out-of-body experiences (OBEs) 1:17 psychoactive drugs 2:220–221, 2:225 runner’s high 1:16 general discussion 1:9, 1:20 hypnosis 1:14, 1:353 pain assessment behavioral scales 1:244, 1:244t, 1:245t ethical considerations 1:247 general discussion 1:243, 1:246f, 1:248 pain perception 1:246 psychoactive drugs 2:217–229 AIM model 2:218 color perception 2:226, 2:227f definitions 2:217 drug classifications 2:221 form constants 2:226 general discussion 2:228 historical background 2:217 phenomenological characteristics 2:223 research concerns 2:217 transient hypofrontality theory 2:219, 2:220f religious/spiritual experiences 2:273 sleep-related experiences 1:12, 2:248 trances 1:353 animal consciousness 1:23–36 mirror test 1:24 philosophical theories 1:349 phylogenetic definitions general discussion 1:24 natural selection theory versus behaviorism 1:25 sensory higher-order distinctions 1:24 research methodology birds 1:27 cephalopods 1:29 commentary key paradigm 1:26 consciousness correlates 1:25 evidentiary support 1:26 homologous brain structures 1:27 human consciousness-based benchmarks 1:25 universal properties 1:33 visual recognition studies 1:24 artificial intelligence (AI) 1:37–46 behavioral tests 1:43 ethical and legal issues 1:44 general discussion 1:37, 1:44
internal architecture tests 1:43 machine consciousness algorithmic factors 1:40 behavioral tests 1:43 Chinese Room experiment 1:40 cognitive theories 1:39 criticisms 1:40 current research 1:41 hard problem of consciousness 1:40 higher-order thought theory 1:40 information integration theory (IIT) 1:39 inner speech 1:40 internal architecture tests 1:43 limitations of artificial intelligence 1:41 natural selection theory 1:39 positive theories 1:39 Turing test 1:43 research background 1:37 theories of consciousness global workspace theory 1:38 model approaches 1:38 autobiographical memory 1:77–82 characteristics 2:1 cognitive feelings familiarity 1:77–78 knowing 1:78 remembering 1:77 conscious feelings 1:79 definition 1:77 imagined feelings 1:79 malfunctions amnesia 1:79 confabulations 1:80, 2:3, 2:264 de´ja` vu 1:80–81 false memories 1:80, 2:4 neuropsychology 2:2 basic concepts 1:157–170 access consciousness 1:158 awareness 1:157 general discussion 1:157, 1:403 higher-order theories dispositionalist theory 1:167 higher-order thought theory 1:165 inner sense concepts 1:163 intrinsicalism 1:166 reportability tests 1:167 intentional consciousness 1:158 mental states 1:157–158 phenomenal consciousness 1:158 philosophical theories 1:339 qualitative consciousness 1:158 sensory stimulation 1:157 state consciousness 1:157 theoretical perspectives controversial issues 1:160 first-order representational theory 1:161 higher-order theories 1:163 transitive consciousness 1:157, 1:160 bistable perception 1:93–106 basic concepts 1:93, 1:94f first-person skepticism 2:397 general discussion 1:104 neural correlates binocular rivalry 1:100, 1:101 cortical/subcortical processing 1:101 extrastriate cortex 1:101 general discussion 1:99 measurement techniques 1:99 motion perception 1:101, 1:102f neural synchronization 1:104 parietal and prefrontal cortices 1:98f, 1:103 reversal-related activations 1:101, 1:102f theoretical perspectives 1:95 theoretical models behavioral evidence 1:97 binocular rivalry 1:99 high-level theories 1:98, 1:98f low-level theories 1:97, 1:98f
Subject Index
theoretical perspectives anatomical location theories 1:95 consciousness correlates 1:94 neural correlates 1:95 state-change theories 1:95 visual stimuli ambiguous motion 1:94f, 1:96 binocular rivalry 1:97 gamma distribution 1:94f, 1:96 reversible figures 1:96 blindsight 1:119, 2:141 cerebral metabolism 1:152f conscious content 1:94, 1:136, 1:284, 1:290, 2:79 conscious deliberation 1:287 conscious level conscious content distinction 1:284 consciousness dual processing relationship 1:85, 1:86t consciousness intentionality relationship 1:417, 2:179 conscious shyness 2:419 conscious will brain activation 1:123 folk theories 1:257 habit conscious will connection 1:316 historical research 1:123 pathological disorders 2:117 voluntary action action initiation 1:132 dual causation pathways 1:286f historical research 2:111 legal implications 1:132 long-term intentions 1:132 readiness potential study 1:173, 1:280, 2:116 theoretical perspectives 1:286 cosmic consciousness 1:19 definition 1:24, 1:147, 1:148f, 1:419, 2:232 developmental processes 1:221–229 age-related constraints 1:227 Capgras syndrome 1:223, 2:188 general discussion 1:221, 1:227 levels of consciousness 1:94, 1:221, 1:222, 1:225f minimal consciousness 1:222, 1:225f phenomenological onset 1:222 prefrontal cortex growth 1:223 preschooler prefrontal patient study 1:225 reflective (recursive) consciousness 1:223, 1:225f self-awareness theories 1:223, 1:224 subjective perspective labels 1:226 disorders of consciousness (DOC) 1:147–156 clinical definitions brain death 1:148 coma 1:149 locked-in syndrome (LIS) 1:150 minimally conscious state (MCS) 1:149 vegetative states 1:149 diagnostic criteria behavioral evaluations 1:150 coma recovery scale-revised (CRS-R) 1:151 electrophysiology 1:151 external stimulation 1:153, 1:153f full outline of unresponsiveness (FOUR) 1:151 general discussion 1:150 Glasgow coma scale (GCS) 1:150 objective evaluations 1:151 positron emission tomography (PET) analysis 1:152f resting cerebral metabolism 1:151, 1:152f general discussion 1:147 historical research 1:148 treatment strategies 1:153 emotion 1:231–241 affective attributions feelings-as-information model 1:237 implicit explicit affective attributions 1:238–239 metaconsciousness 1:239, 2:37 mood manipulation 1:237, 2:431 binding problem 1:235 conscious versus unconscious feelings 1:236 general discussion 1:232 neuroimaging research 1:234, 2:382 REM sleep 2:363–364
subliminal affective priming 1:233 unconscious stimuli 1:233, 2:418 executive ignorance 1:171 first-person knowledge 2:394 folk theories 1:251–263 behavior/reason explanations 1:254 causal history factors 1:254–255 folk theory of mind concept 1:252, 1:253f, 2:188–189 free will 1:260 functional role 1:251 general discussion 1:261 incommunicability general discussion 1:258 incorrigibility 1:260 privileged access 1:258 language of mental activities 1:255 observability 1:256f personhood criteria 1:257 phenomenal consciousness 1:256 fringe consciousness feelings of knowing 2:448, 2:461 general discussion 2:460 immediate memory 2:331–332 spatial situatedness 2:187 tip-of-the-tongue state 2:461 functional role 1:279–293 conscious inessentialism 1:280 conscious level conscious content distinction 1:284 Cummins-function factor 1:283 dreaming 1:292 dual processing theory 1:87 epiphenomenalism 1:280 error correction 1:292 evolutionary development theories 1:282 flexible behaviors 1:288 function functionalism controversy 1:283 general discussion 1:279, 1:292 integration consensus boundary factors 1:290 differentiation processes 1:290 general discussion 1:288 global workspace theory 1:288 skill acquisition and learning 1:289, 2:13 theory of neuronal group selection (TNGS) 1:291 prediction abilities 1:292 primary consciousness higher-order consciousness distinctions 1:284 rational action 1:287 readiness potential study 1:173, 1:280, 1:281f social interactions 1:292 volition 1:280, 1:286, 1:286f habits 1:315–328 basic concepts 1:315 formation processes general discussion 1:316 goal-directed skills 1:318, 2:426 stimulus response conditions 1:317 general discussion 1:326 habit conscious will connection 1:316 nonconscious (automatic) processes attentional control changes 1:320 dual-task interference 1:320 explicit learning research 1:323 general discussion 1:319 goal-directed skills 1:322 implicit learning research 1:322 intention behavior prediction connection 1:321 neuroimaging research 1:320 sensemaking factors authorship experiences 1:325 experienced ease of retrieval 1:325 general discussion 1:324 self-perception 1:324 stimulus response conditions 1:316, 1:317 higher-order thought theory 1:40, 1:285, 1:348, 1:426 historical research 2:159 illusory experiences 1:141, 1:286, 1:326, 1:360, 2:397 implicit social cognition 1:383–388
479
480
Subject Index
consciousness (continued) automatic thoughts and actions 1:383 general discussion 1:387 social constructs attitudes 1:384 automatic social behavior 1:385 construct formation and change processes 1:385 general discussion 1:384 goal pursuit 1:385 measurement techniques 1:386 self-concept 1:385 stereotypes 1:384 inner speech 1:389–402 definitions 1:389 developmental characteristics 1:392 dysfunctional self-talk autism 1:398 general discussion 1:396 hyperactivity 1:398 schizophrenia 1:397 functions language skills 1:395 memory 1:396 miscellaneous functions 1:396 self-regulation 1:394 task-switching performance 1:396 general discussion 1:400 measurement techniques 1:391 neuroanatomy 1:393 self-awareness theories 1:398 theoretical perspectives 1:389 intentionality 1:417–429 basic concepts 1:419 behavior/reason explanations 1:254 causal history factors 1:254–255 consciousness definitions 1:419 consciousness intentionality relationship 1:417, 2:22 , 2:179 embodied intentionality 1:428 folk theories basic concepts 1:253, 1:253f behavior/reason explanations 1:254 language of mental activities 1:255 observability 1:256f Husserl’s theory 1:428 intentional inexistence analytical approaches 1:421 biosemantics (Millikan) 1:425 Brentano’s thesis 1:423 higher-order thought theory 1:426 indicator semantics (Dretske) 1:424 language structure analysis 1:421 mental states 1:256, 1:420 Representation Theory of Mind (RTM) 1:423 Merleau-Ponty’s theory 1:428, 2:181 phenomenal consciousness 1:427 phenomenological characteristics 2:179 voluntary action 1:253, 1:253f, 2:115 levels of consciousness 1:94, 1:221, 1:225f linguistic background 1:252 linguistic processes 1:447–459 categorical perception 1:454 characteristics 1:447 conscious perception 1:453 general discussion 1:447, 1:458 meaningfulness 1:452 modularity 1:451 structural characteristics color perception 1:457 number perception 1:457–458, 1:458f sign language 1:454 thought language connection 1:456 symbolic representations 1:452 unconscious processing 1:449, 2:413–414 mental imagery 2:445–457 background information 2:446 definitions 2:446 general discussion 2:456
historical research 2:447 subjective individual differences 2:449 theoretical perspectives description (mentalese) theory 2:453 enactive theory 2:454 general discussion 2:450 picture theory 2:451, 2:451f mental representations 2:19–32 basic concepts 2:19 general discussion 2:19 misrepresentations of experiences 2:20, 2:21 modes of presentation 2:20–21 nonrepresentational theories 2:31 phenomenal consciousness 2:22 psychosemantic theories 2:21 representational theories basic concepts 2:23 criticisms 2:26 enactive theory 2:31 externalism 2:27, 2:58, 2:59f internalism 2:29, 2:58, 2:59f narrow content 2:29 subjectivity 2:30 weak versus strong representationalism 2:31 teleosemantic theories 2:22 narrative consciousness 2:161 neurobiological theories 2:87–100 access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 cognitive influences 2:87 Duplex Vision Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:91 neurocognitive panpsychism 2:98 general discussion 2:99 Global Neuronal Workspace Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:90 neurocognitive panpsychism 2:98 global versus local theories 2:89 hard problem of consciousness 2:87 Local Recurrence Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:93 neurocognitive panpsychism 2:98 microconsciousness theory 2:94 neural correlates 1:95, 2:88, 2:420 neural indexes 2:98 neurocognitive panpsychism 2:98 Reentrant Dynamic Core Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:89 neurocognitive panpsychism 2:98 neurochemical mechanisms 2:101–110 awareness alertness controls 2:103 clinical consciousness anesthetics 2:101 cholinergic system 2:102 coma 2:101 desferrioxamine/prochlorperazine doses 2:102 coma anesthetics 2:101 cholinergic system 2:102 desferrioxamine/prochlorperazine doses 2:102 induction mechanisms 2:101 general discussion 2:101 phenomenal consciousness acetylcholine 2:107 adrenergic system 2:108 bipolar disorder 2:108 mood manipulation 2:108 schizophrenia 2:109 serotonin 2:104 virtual reality processes 2:106
Subject Index
normal state of consciousness (NSC) 1:9 phenomenal consciousness basic concepts 1:158, 2:22 characteristics 1:85, 1:86t, 2:161, 2:390 conscious content 1:285 folk theories 1:256 intentionality 1:427 neurobiological theories 2:95 personhood criteria 1:257 representational theories basic concepts 2:23 criticisms 2:26 enactive theory 2:31 externalism 2:27 internalism 2:29 subjectivity 2:30 self-awareness theories 2:161 serotonin 2:104 subjective properties 1:144, 2:390 phenomenological characteristics 2:175–186 embodiment 2:181 general discussion 2:175 intentionality 2:179 intersubjectivity 2:183 methodological approach 2:176 self-consciousness 2:177 phenomenological subtraction and addition 1:403, 2:219–220 philosophical theories 1:339–350 Aristotilean philosophy 1:341 behaviorism 1:344 modern approaches animal consciousness 1:349 anomalous monism theory 1:348, 2:46 externalism 1:347, 2:27 first-order representational theory 1:348–349 functionalism 1:283, 1:346 general discussion 1:345 higher-order thought theory 1:285, 1:348 identity theory 1:345–346 internalism 2:29, 2:180–181 materialism 1:345–346 mental causation 1:347, 2:46 nineteenth-century idealism theory Berkeley 1:343 Brentano 1:343, 1:423, 2:179 Clifford 1:343 Darwin 1:343 general discussion 1:343 Husserl 1:343, 1:428, 2:176, 2:179 Kantian theory 1:343 pre-Socratic philosophers 1:340 radical emergentism 1:343–344 scientific realism 1:345 scientific revolution era Cartesian dualism 1:342, 2:181 Galilean philosophy 1:341–342 Gassendi 1:343 general discussion 1:341 Hobbes 1:343 Leibniz 1:342 Malebranche 1:342–343 Spinoza 1:342 verificationism 1:344 vitalism 1:344–345 phylogenetic definitions general discussion 1:24 natural selection theory versus behaviorism 1:25 sensory higher-order distinctions 1:24 primary consciousness higher-order consciousness distinctions 1:284 psychopathology 2:245–271 basic concepts 2:246 cognitive disorders basic concepts 2:261 delirium 2:261, 2:262t dementia 2:261, 2:262t identity perception 2:263 insight 2:265
memory 2:263, 2:263t place perception 2:262–263 self-perception 2:264 time perception 2:262 visuospatial information 2:264 diagnostic concerns 2:268 disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257 misidentification syndromes 2:257 self-harm risks 2:255 emotional disorders 2:252 general discussion 2:245 mood and affect disorders 2:252, 2:253t movement disorders characteristics 2:265 involuntary movements 2:267 voluntary movements 2:266 perception disorders first rank symptoms 2:260t general discussion 2:258 hallucinations 2:259 top-down/bottom-up cognitive processing 2:259 personality disorders 2:267, 2:269t phenomenological characteristics delirium versus dementia 2:249–250 form versus content 2:251 functional versus organic disorders 2:248–249 general discussion 2:247 levels of explanation 2:251, 2:252t psychosis versus neurosis 2:248 reactive versus instrumental behavior 2:251 terminology 2:250t visual versus auditory hallucinations 2:249 psychiatric diagnoses 2:247 reflective consciousness 1:223, 1:225f, 2:161, 2:177, 2:344 relational binding 1:189 research challenges 1:375 response override hypothesis 1:205–219 basic concepts 1:205, 1:206f general discussion 1:217 perception awareness research 1:206 retrieval stopping processes Freudian suppression model 1:214 individual differences 1:214 intrusive memory awareness 1:215 neurobiological theories 1:212 Think/No-Think paradigm (TNT) 1:210, 1:211f, 1:212f, 1:213f thought suppression 1:216 working memory control mechanisms 1:216 selective retrieval inhibition mechanisms 1:209 inhibitory control processes 1:207 neurobiological theories 1:209 retrieval-induced forgetting (RIF) 1:206f, 1:208 retrieval stopping processes 1:210 restructuring experiences 1:435 scientific research 1:329–338 background information 1:329 behaviorism basic concepts 1:333 criticisms 1:334 opposing viewpoints 1:335 Skinner’s theories 1:335 cognitive theories basic concepts 1:335 euphemisms 1:336 neuroscience research 1:336 subjective experiences 1:336, 2:389–400 early explorations 1:330 research concerns 1:337 systematic studies imageless thought controversy 1:332 introspectionism 1:331 mind body problem 1:333 nineteenth-century studies 1:330 unconscious information processing 1:331
481
482
Subject Index
consciousness (continued) selective attention 1:61–75 attention awareness relationship attentional blink 1:67, 1:234 awareness of attention 1:71 awareness of awareness 1:70, 1:71f blindsight 1:69, 1:69f change blindness 1:47–59, 1:67, 1:86, 1:233–234, 2:85, 2:96, 2:167 general discussion 1:66 gist situations 1:68 inattentional blindness 1:47–59, 1:67, 1:233–234, 2:85, 2:167, 2:335 neural mechanisms 2:84 neurochemical mechanisms 2:103 otherwise-engaged attention 1:68 postattentive awareness 1:70 basic concepts 1:61 binding problem 1:62, 1:235 general discussion 1:71 neurochemical mechanisms 2:103 object recognition feature characteristics 1:65, 1:65f feature integration theory 1:63 feedforward models 1:65 reverse hierarchy theory (RHT) 1:66 self-consciousness 2:177 semantic concepts 1:252 sleep altered states of consciousness brain-injured patients 2:367, 2:368f characteristics 1:12 coma 2:367, 2:368f locked-in syndrome (LIS) 2:370 recovered states 2:369 sleep disturbances 2:369 vegetative states 2:369 arousal awareness states 2:365 social foundations 2:375–377 conscious preconscious unconscious systems 2:375 digital representations 2:376 historical research 2:376 language studies 2:375 mere exposure effects 2:376 mimicry 2:376 self-identity 2:375 theory of symbolic interactionism 2:375 state consciousness 1:157, 1:284 state trait relationship 2:160 streams of consciousness 1:140, 1:354, 2:58, 2:412 temporal structure 1:193–204 A-series B-series theory 1:195 passage of time 1:193, 1:300 psychological characteristics 1:195 specious present doctrine 1:196, 1:197f time consciousness models act content model 1:199 color phi phenomenon 1:198 diachronic coconsciousness 1:199 flash-lag illusion 1:198 general discussion 1:198 Husserl’s phenomenological model 1:200, 1:201f hybrid model 1:202 overlap model 1:199 protention retention theory 1:200 synchronous coconsciousness 1:199 testimonies of consciousness 2:399 theories of consciousness artificial intelligence (AI) global workspace theory 1:38 machine consciousness 1:39 model approaches 1:38 cognitive processes 1:135–146 hard problem of consciousness 1:144 higher-order thought theory 1:426 historical research 1:135 illusory theories apparent mental causation theory 1:141 general discussion 1:140 multiple drafts model 1:140
influential theories general discussion 1:137 global workspace theory 1:137 information integration theory (IIT) 1:139 intermediate level theory 1:138 learning process theories general discussion 1:141 higher-order Bayesian decision theory 1:143 radical plasticity thesis 1:142 sensory-motor theory 1:141 precursor theories automatic controlled processing 1:137 central executive system 1:136 modularity 1:137, 1:451 supervisory attentional system 1:136 transitive consciousness 1:157, 1:160, 1:284 unconscious conscious perception 2:135–146 action control 1:171–181 general discussion 2:135, 2:145 goal pursuit 2:423–433 automatic thoughts and actions 2:426 decision-making tasks 2:428 general discussion 2:423, 2:427, 2:432 goal contagion 2:427 goal establishment 2:424 goal intentions 2:425 goal-related words 2:430 goal striving 2:427 habitual behaviors 2:426 implementation intentions 2:425 monitoring processes 2:429 planning and implementation stage 2:425 revision stage 2:430 subliminal affective priming 2:427–428 historical research 2:412 linguistic processes 1:453 Local Recurrence Theory 2:93 model approaches null hypothesis model 2:138, 2:139f objective threshold/nonmonotonic approach 2:139f, 2:142 objective threshold/rapid decay approach 2:139f, 2:141 subjective threshold approach 2:139, 2:139f model validity 2:138 neural/psychological correlates 2:420 objective threshold approach 2:137, 2:139f, 2:141 subjective threshold approach 2:136, 2:139, 2:139f unconscious information processing 2:159–173 access consciousness 2:161 attention-related factors 2:171 binocluar rivalry 1:94f, 1:97, 1:99, 2:92–93, 2:165, 2:166f cognitive contents 2:160–161 criterion content 2:162 decision-making tasks 2:428 definitions 2:160 exhaustive exclusion 2:164 feedforward feedback connections 2:171 general discussion 2:171 habit-formation processes 1:323 implicit versus explicit learning 1:369, 1:378 indirect measurement strategies 2:164 invisible stimuli 2:75, 2:165, 2:166f methodological issues 2:162 narrative consciousness 2:161 nineteenth-century studies 1:331 normal versus neurologically impaired observers 2:165 phenomenal consciousness 2:161 postretinal processing 2:168 reflective consciousness 2:161 research challenges 1:375 research concerns 2:160 results and analysis 2:168 signal detection theory (SDT) 2:163 subliminal affective priming 1:233, 2:427–428 threshold measurements 2:163 unresolved issues attention consciousness relationship 2:96, 2:144 brain activation 2:144 direct indirect task relationship 2:143
Subject Index
implicit perception 2:144 masking techniques 2:144 subliminal perception 2:144 visual processing 2:127 unconscious information processing 2:159–173 attention-related factors 2:171 conscious level conscious content distinction 1:284 feedforward feedback connections 2:171 general discussion 2:171 implicit versus explicit learning 1:369, 1:378 invisible stimuli methods attentional limits 2:167 binocluar rivalry 1:94f, 1:97, 1:99, 2:92–93, 2:165, 2:166f flash suppression 2:166 general discussion 2:165 motion-induced blindness 2:166 semistabilized images 2:165, 2:166f transcranial magnetic stimulation (TMS) 2:82, 2:167 visual crowding 2:166f, 2:167 visual (two-transient) masking 2:77, 2:166f, 2:167 methodological issues criterion content 2:162 exhaustive exclusion 2:164 general discussion 2:162 indirect measurement strategies 2:164 invisible stimuli 2:75, 2:165, 2:166f normal versus neurologically impaired observers 2:165 signal detection theory (SDT) 2:163 threshold measurements 2:163 postretinal processing attentional limits 2:171 backward masking 2:170 binocular rivalry 2:169 blindsight 2:168 continuous flash suppression 2:170 feature inheritance 2:170, 2:170f form blindness 2:169 general discussion 2:168 motion-induced blindness 2:170 neurological patients 2:168 normal observers 2:169 transcranial magnetic stimulation (TMS) 2:170 research challenges 1:375 research concerns access consciousness 2:161 cognitive contents 2:160–161 definitions 2:160 narrative consciousness 2:161 phenomenal consciousness 2:161 reflective consciousness 2:161 results and analysis general discussion 2:168 postretinal processing 2:168 voluntary action general discussion 1:126 movement control 1:126 movement initiation 1:127 readiness potential study 1:127, 1:128f conscious shyness 2:419 conscious will folk theories 1:257 habit conscious will connection 1:316 pathological disorders 2:117 voluntary action action initiation 1:132 brain activation 1:123 dual causation pathways 1:286f historical research 2:111 legal implications 1:132 long-term intentions 1:132 readiness potential study 1:173, 1:280, 2:116 theoretical perspectives 1:286 continuous flash suppression 2:170 continuous moment hypothesis 2:332, 2:332f controlled processes automatic controlled processing comparisons (dual processing theory) basic concepts 1:84, 2:413 brain systems 1:90, 1:90f
conscious monitoring 1:89 consciousness dual processing relationship 1:85, 1:86t functional role 1:87 general discussion 1:91 information flow 1:89 neural/psychological correlates 2:420 theories of consciousness 1:137 theory of mind research 2:405, 2:406t, 2:407t top-down conscious processing 1:89–90, 2:96 general discussion 1:83 convergence coding 2:153, 2:153t convergent thought 1:438 conversion disorder paralysis 1:363–364, 2:248–249, 2:417 convexity bias 2:438 Cooley, Charles Horton 2:375 Copernicus, N 2:465 coprolalia 2:266–267 copropraxia 2:266–267 corpus collosum 1:185f, 1:448 cortical arousal 1:436, 1:439 cortical blindness 1:107, 2:127 cortical dementia 2:264 corticotropin-releasing hormone (CRH) 2:281 cortisol 2:276t cosmic consciousness 1:19 costly signaling theory (CST) 2:349 Cotard’s delusion 2:254, 2:258t counterfactual thought 1:444, 2:21–22 counting cultures 1:457–458 crack cocaine 2:255 cranial arteritis 2:262t creativity 1:431–446 active unconscious 1:437 background information 1:431 definitions 1:432 Dyads of Triads Task 1:437, 1:437f future research directions 1:443 implicit thought 2:418 incubation periods 1:436 insight problem-solving experiences 1:433, 1:434f, 1:444f, 2:68, 2:418 neuroimaging research compound remote associate (CRA) problems 1:438, 1:441 divergent thought 1:438 moment of insight 1:439, 1:441f right-hemisphere priming 1:437 restructuring experiences 1:433 spontaneous discoveries 1:431 theoretical perspectives anagrams 1:443 attention differences 1:442 conscious versus unconscious thought 1:441 unconscious problem-solving experiences 1:436 Creutzfeldt Jacob disease 2:262t, 2:263–264 crib speech 1:389 Crick, Francis 1:206, 1:337, 2:88, 2:464 CRIES (pain scale) 1:244t Critchley, Macdonald 2:105 criterion content 2:162 crossmodal binding 2:151 cross-race disadvantage 2:439–440 crowding, visual 2:166f, 2:167 Cummins-function factor 1:283 Cummins, Robert 1:283 current concerns 2:62f, 2:63 cyclopropane 1:296 cyclothymia 2:253t cytochrome oxidase (CO) 2:151
D Dani language 1:457 Darwin, Charles 1:25, 1:236, 1:343, 2:465 date rape drug 2:226–227 Davidson, Donald 1:348, 2:46, 2:194 daydreaming default mode hypothesis 2:67 mental states 2:59–60 neural mechanisms 2:219–220
483
484
Subject Index
DB (research patient) 1:108, 1:112, 1:112f, 1:113, 1:113f deafferentiation 2:291 deafness inner speech 1:395–396 linguistic processes 1:454 death instinctual drive 2:239 decision-making tasks 1:268, 2:76–77, 2:428 declarative memory 2:15, 2:263t deep brain stimulation 1:153 default mode hypothesis 2:67 deficit states 2:253t, 2:254 degenerative diseases 2:262t Dehaene, Stanislas 1:158–159, 2:58–59, 2:90, 2:96 de´ja` ecoute 2:262–263 de´ja` vu 1:80–81, 2:262–263 Delboeuf, Josef 2:231, 2:236 delirium causal theories 2:261, 2:262t delirium tremens 2:249, 2:262–263 phenomenological characteristics 2:249–250 place perception 2:262–263 visual hallucinations 2:260 delta-(9)-tetrahydrocannabinol (THC) 2:227–228, 2:227f delusions alien control delusions 1:286, 1:364–365, 2:258t, 2:260t characteristics 2:258t clinical delusions 1:363 definition 1:12 delusional infestations 2:261 delusions of control 2:258t delusions of grandeur 2:253t, 2:258t delusions of guilt 2:258t delusions of reference 2:258t erotic delusions 2:258t jealousy 2:258t primary delusions 2:260t psychoactive drugs altered states of consciousness 2:220 phenomenological characteristics 2:223 schizophrenia 2:109 self-harm risks 2:255 shared delusions 2:258t dementia causal theories 2:261, 2:262t dementia pugilistica 2:262t identity perception 2:263 implicit memory 2:414 phenomenological characteristics 2:249–250 place perception 2:262–263 psychiatric diagnoses 2:247 visual hallucinations 2:260 Demerol 2:222 Democritus 1:340 Dennett, Daniel 1:140, 1:158–159, 1:163, 1:423, 2:399 dentate gyrus 1:185f dependent personalities 2:269t depressants 2:221 depression characteristics 2:253t, 2:254 delusions of guilt 2:258t hallucinations 2:260 mind wandering 2:63–64 placebo treatments 2:202, 2:213 psychiatric diagnoses 2:247, 2:252t depth psychology 2:231 derailment 2:256t Descartes, Rene´ action control 1:175 consciousness intentionality relationship 1:417–418, 2:181 mental imagery 2:447–448, 2:452 mental states 1:157–158, 2:23 mind body problem 1:342, 2:45, 2:452 personal identity studies 2:303 privileged access concept 2:189–190 unconscious conscious processing 1:378 description (mentalese) theory 2:453 descriptive psychopathology 2:246 desferrioxamine/prochlorperazine doses 2:102
desflurane 1:296 designer drugs 2:222 desires See beliefs and desires detection tasks 1:49, 1:54 determining quality 2:238 determinism 1:265 D.F. (patient) 2:124–125, 2:124f diabetes mellitus 2:262t diachronic coconsciousness 1:199 Diagnostic and Statistical Manual (DSM) 2:247 dichoptic presentations 1:55 dichotic listening 2:65, 2:417 diencephalon 1:184, 1:185f, 1:186f diethyl ether 1:296 diffuse attention 1:442–443 Dimensional Change Card Sort 1:226, 1:227 dimethyl tryptamine (DMT) 2:284 diminished consciousness See minimally conscious state (MCS) direct awareness 2:293 discrepancy detection 2:431 discrete moment hypothesis 2:332, 2:332f discrimination tasks 1:143, 2:76–77 disease and placebo treatments 2:202 disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257 misidentification syndromes 2:257 self-harm risks 2:255 disorders of consciousness (DOC) 1:147–156 clinical definitions brain death 1:148 coma 1:149 locked-in syndrome (LIS) 1:150 minimally conscious state (MCS) 1:149 vegetative states 1:149 diagnostic criteria behavioral evaluations 1:150 coma recovery scale-revised (CRS-R) 1:151 full outline of unresponsiveness (FOUR) 1:151 general discussion 1:150 Glasgow coma scale (GCS) 1:150 objective evaluations electrophysiology 1:151 external stimulation 1:153, 1:153f positron emission tomography (PET) analysis 1:152f resting cerebral metabolism 1:151, 1:152f general discussion 1:147 historical research 1:148 treatment strategies 1:153 dispositionalist theory 1:167 dissociation theories access phenomenal consciousness relationship 2:95 deafferentiation 2:291 hypnosis 1:354 perception action frames of reference 2:129 sensory-motor reflex physiology 2:234 theoretical perspectives 1:359, 1:359f dissociative amnesia 2:414 dissociative identity disorder 2:302, 2:414 distal decisions 1:268 distractibility 2:62f, 2:64 divergent thought 1:438 doctrine of Forms 1:340–341 DOLOPLUS 2 (pain scale) 1:244t Donders, Franciscus Cornelis 1:175 dopamine arousal system 1:305f neurochemical mechanisms 2:102, 2:104 placebo response effects 2:212, 2:214 religious/spiritual experiences 2:275, 2:277f, 2:279 dorsal premotor cortex (dPMC) automatic controlled processing research 1:90, 1:90f voluntary action 2:115–116 dorsolateral prefrontal cortex (DLPFC) automatic controlled processing research 1:90, 1:90f fringe consciousness 2:461
Subject Index
placebo response analysis 2:211, 2:213 religious/spiritual experiences 2:284 REM sleep 2:345 retrieval stopping processes 1:212f, 1:213f voluntary action 1:124, 1:124f, 1:125f dorsolateral thalamus 1:28–29 dreaming 2:341–355 background information 2:341 bad dreams 1:14, 2:345 characteristics 1:13, 2:361 consciousness functions 1:292 cross-cultural studies 2:346 dream content analysis bizarreness 2:344 dream characters 2:343 dream self 2:343 emotional experiences 2:344 general discussion 2:342 misfortunes 2:343–344 reflective consciousness 2:344 sensory modalities 2:342 social interactions 2:343, 2:351 threatening events 2:344 dream themes 2:342 evolutionary functions costly signaling theory (CST) 2:349 general discussion 2:348 psychological problem-solving and creativity 2:350 sentinel theory 2:349 simulation functions 2:350 false awakening 1:10–11 general discussion 2:354, 2:370 historical research 2:357 hypnosis 1:352 illustrative diagram 1:148f lucid dreaming characteristics 1:14, 1:16, 2:364 illustrative diagram 1:148f neural mechanisms 2:219–220 neurobiological correlates 2:362 recurrent dreams 2:345 simulation functions general discussion 2:350 play behaviors 2:351 social simulation hypothesis 2:351 threat simulation theory (TST) 2:352 theoretical perspectives continuity hypothesis (CH) 2:347 evolutionary functions 2:348 psychoanalytic theories 2:346 psychological healing theories 2:348 random activation theories (RATs) 2:347 Dretske, Fred 1:162, 1:423, 1:424, 2:195 Driesch, Hans 1:344–345 drowsiness 1:148f drug abuse 2:255, 2:262t drug classifications See psychoactive drugs dualism Aristotilean philosophy 1:341 Cartesian dualism 1:342, 2:45, 2:181 criticisms 2:45 naturalistic dualism 2:54 self-world dualism 2:296 Spinoza 1:342 substance dualism 2:45 token mental events 2:45 dual processing theory automatic attention response 1:87 basic concepts 1:84, 2:413 brain systems 1:90, 1:90f conscious monitoring 1:89 consciousness dual processing relationship 1:85, 1:86t functional role 1:87 general discussion 1:91 information flow 1:89 neural/psychological correlates 2:420 theories of consciousness 1:137
theory of mind research 2:405, 2:406t, 2:407t top-down conscious processing 1:89–90, 2:96 duck rabbit image 2:397, 2:451f Duplex Vision Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:91 neurocognitive panpsychism 2:98 Dyads of Triads Task 1:437, 1:437f dying brain hypothesis 1:18 dynamic core hypothesis (DCH) 1:290–291, 2:89 dynamic unconsciousness 1:378 dysarthria 2:256t dyscalculia 2:265 dysfunctional self-talk autism 1:398 general discussion 1:396 hyperactivity 1:398 schizophrenia 1:397 dysgraphia 2:265 dyskinesias 2:267 dysthymia 2:253t dystonias 2:267
E Ebbinghaus Illusion 2:129, 2:130f echolalia 1:389, 2:256t ecstasy (MDMA) classifications and characteristics 2:222, 2:224t phenomenological characteristics 2:225, 2:226–227 Edelman, Gerald 1:139, 1:288, 2:89 ego 2:240 ego boundaries 2:257 egocentric neglect 2:71 egocentric representations 2:9 egocentric speech 1:389 ego-dystonic 2:250t ego instinctual drive 2:239 ego-syntonic 2:250t Einstein, Albert 1:345 Ekbom’s syndrome 2:261 electroencephalography (EEG) anesthetic effects 1:301, 1:302f bistable perception analysis 1:99 cephalopods 1:33 creativity research 1:438 disorders of consciousness (DOC) 1:151 intergroup processing 2:380 neural encoding mechanisms 2:81 placebo response analysis 2:207 religious/spiritual experiences 2:275–276, 2:283 REM sleep 2:359f, 2:360 selective memory retrieval 1:210 voluntary action studies 1:129 electrolyte disturbances 2:262t electromyography 1:391–392, 2:359f, 2:360, 2:380 electrooculograms (EOGs) 2:359f, 2:360 Ellenberger, Henri 2:233, 2:235 embedded private speech 1:389 embodied intentionality 1:428, 2:181 emergentism 1:340, 1:343–344 emotion 1:231–241 affective attributions feelings-as-information model 1:237 implicit explicit affective attributions 1:238–239 metaconsciousness 1:239, 2:37 mood manipulation 1:237, 2:431 binding problem 1:235 conscious versus unconscious feelings 1:236 general discussion 1:232 neuroimaging research 1:234, 2:382 REM sleep 2:363–364 subliminal affective priming 1:233 unconscious stimuli 1:233, 2:418 emotional disorders 2:252 emotional incontinence 2:253t
485
486
Subject Index
empathy 2:184, 2:255, 2:269t empty speech 2:256t enactive theory 2:31, 2:454 encephalitis lethargica 2:248–249, 2:262t endocrine diseases 2:262t endogenous opioid responses 2:206, 2:210, 2:212 endorphins 2:276t, 2:281 enflurane 1:296 enlightenment characteristics 1:19–20 entactogens 2:226–227 entorhinal cortex 1:184, 1:186f environmental dependency syndrome 2:118 ephedra 2:221–222 epileptic seizures attention awareness relationship 1:86–87 binocular rivalry 2:129 hyperemotionalism 2:254–255 place perception 2:262–263 visual hallucinations 2:260 epiphenomenalism free will theory background information 1:273 facilitated communication 1:273 scientific epiphenomenalism 1:274 utilization behavior 1:273–274 perceptual learning 2:82 theories of consciousness 1:141, 1:280 episodic memory amnesia 1:189 characteristics 1:188, 1:369, 2:1, 2:263t definition 1:77 dreaming 2:347 egocentric representations 2:9 neuropsychology 2:2 Think/No-Think paradigm (TNT) 1:213–214 working memory control mechanisms 1:216 Eros 2:240 erotic delusions 2:258t error correction 1:292 error-related negativity 2:384 Esdaile, James 1:353–354 ether 1:296, 1:298, 2:222 ethylene 1:296 etomidate 1:296 Eureka moment 1:431 evaluative priming 1:386 event-related potentials (ERPs) creativity research 1:440 disorders of consciousness (DOC) 1:151 free will theory 1:272 intergroup processing 2:381 memory capacity 2:337 mind wandering research 2:66 selective memory retrieval 1:210 theory of mind research 2:407 executive control/executive monitoring model confabulations 2:5–6 folk theory of intentionality 1:254 habit formation processes 1:320 hypnosis research 1:359, 1:359f neuroimaging research 2:403 selective memory retrieval 1:210, 1:211f sensory memory 2:329, 2:330f theories of consciousness 1:136 theory of mind research 2:403 executive ignorance 1:171 exhaustive exclusion 2:164 experimental psychopathology 2:246 explanatory gap 2:50, 2:88, 2:394 explicit learning characteristics 1:369 habits 1:323 explicit memory amnesia 1:184, 1:189, 2:414 characteristics 1:188, 1:369, 2:1, 2:263t definition 1:184 implicit explicit interactions 1:190, 2:414 neuroanatomy 1:189
research developments 1:190 research methodology 1:370 explicit perception 1:50, 1:54, 2:416 explicit social cognition 1:384 extended reticulo-thalamic activating system (ERTAS) 1:303, 1:308 externalism 1:347, 2:27, 2:58, 2:59f, 2:180–181 extinction attention issues 1:58 neglect manifestations 2:72 visual systems 2:128 extracampine visual hallucinations 2:259 extrastriate cortex bistable perception 1:101 implicit memory 1:187 perceptual learning 2:83 REM sleep 2:362, 2:363f
F face perception 1:2 face recognition 2:439–440 Faces, Legs, Activity, Cry, Consolability (FLACC) Pain Assessment Tool 1:244, 1:244t facilitated communication 1:273 faked conditions 1:365 Fallon, James 1:310 false awakening 1:10–11 false beliefs event-related potentials (ERPs) 2:407 general discussion 2:401, 2:408 mirror neurons 2:408 neuroimaging research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 language 2:401, 2:402f personality traits 2:405 social cognition 2:404 false memories 1:80, 2:4 familiarity amnesia 1:189 definition 1:77–78 feature conjunctions 2:155–156 feature exaggeration 2:439 feature inheritance 2:170, 2:170f feature integration theory 1:63, 1:138, 2:155, 2:155f feature maps 1:63, 2:155, 2:155f Fechner, Gustav 2:436 feedforward feedback connections 1:174, 2:156, 2:171 feelings-as-information model 1:237 feelings of knowing 2:448, 2:461 Festinger, Leon 2:376 figure-ground reversing figures 1:94f, 1:96 filtering theories 1:136, 2:328, 2:413 finger agnosia 2:264–265 finite-state grammar 1:372, 1:372f first-order representational theory basic concepts 1:161 controversial issues 1:160 general discussion 1:348–349 representational content 2:30 subjectivity 2:30 weak versus strong representationalism 2:31 first-person perspective cognition research 2:389 dream content analysis 2:343 first-person skepticism 2:397 psychopathology 2:245–271 basic concepts 2:246 cognitive disorders basic concepts 2:261 delirium 2:261, 2:262t dementia 2:261, 2:262t identity perception 2:263 insight 2:265 memory 2:263, 2:263t place perception 2:262–263
Subject Index
self-perception 2:264 time perception 2:262 visuospatial information 2:264 diagnostic concerns 2:268 disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257 misidentification syndromes 2:257 self-harm risks 2:255 emotional disorders 2:252 general discussion 2:245 mood and affect disorders 2:252, 2:253t movement disorders characteristics 2:265 involuntary movements 2:267 voluntary movements 2:266 perception disorders first rank symptoms 2:260t general discussion 2:258 hallucinations 2:259 top-down/bottom-up cognitive processing 2:259 personality disorders 2:267, 2:269t phenomenological characteristics delirium versus dementia 2:249–250 form versus content 2:251 functional versus organic disorders 2:248–249 general discussion 2:247 levels of explanation 2:251, 2:252t psychosis versus neurosis 2:248 reactive versus instrumental behavior 2:251 terminology 2:250t visual versus auditory hallucinations 2:249 psychiatric diagnoses 2:247 subjective experiences 2:394 fission factor 2:307 5-HT receptors 1:305f, 2:104 flash suppression 2:166, 2:170 Flavell, John 2:34 flexible behaviors 1:288 flight of ideas 2:256t Flohr, Hans 1:310 flow experiences 1:16, 2:37 fluent dysphasis 2:256t focus of attention 2:335, 2:336f See also inattentional blindness Fodor, Jerry 1:137, 1:451 folic acid 2:262t folk theories 1:251–263 behavior/reason explanations 1:254 causal history factors 1:254–255 folk theory of mind concept 1:252, 1:253f, 2:188–189 free will 1:257 functional role 1:251 general discussion 1:261 incommunicability general discussion 1:258 incorrigibility 1:260 privileged access 1:258 introspection 2:188–189 language of mental activities 1:255 observability 1:256f personhood criteria 1:257 phenomenal consciousness 1:256 picture theory 2:451 forbidden thoughts 2:429–430 forgetting processes 1:206f, 1:208 formal thought disorder 2:250t form blindness 2:169 formication 2:261 Forms, doctrine of 1:340–341 fornix 1:184, 1:185f, 1:186f frame problem 1:37–38 free association 2:238 free will theory 1:265–277 conscious intentions decision-making distinctions 1:268 empirical approaches 1:268 epiphenomenalism
background information 1:273 facilitated communication 1:273 scientific epiphenomenalism 1:274 utilization behavior 1:273–274 proximal decisions causal theories 1:270 event-related potentials (ERPs) 1:272 Libet’s studies 1:127, 1:128f, 1:269 readiness potentials 1:127, 1:128f, 1:269 unconscious and conscious proximal decisions 1:271 veto actions 1:272 window of opportunity 1:270 folk theories 1:260 general discussion 1:276 mind body problem 2:463 neuroscience research 1:130 theoretical background causal theories 1:267 compatibilist theories 1:266 controversial issues 1:267 definitions 1:265 libertarian theories 1:266 moral responsibility 1:268 Fregean sense 2:20–21 Fregoli’s syndrome 2:258 Freud, Sigmund associationism 2:232 dream theories 2:346 dynamic unconsciousness 1:378 hypnosis 1:354 implicit measurement techniques 1:386 interpretive psychopathology 2:246 psychodynamic theories of the unconscious clinical unconsciousness 2:235 conscious preconscious unconscious systems 2:234, 2:240, 2:419 rational therapies 2:237 research background 2:231 sexual repression castration anxiety 2:241 causal theories 2:237 instinctual drives 2:239 mental structural theory 2:240 Oedipus complex 2:241 seduction hypothesis 2:238 sexual instinctual drive 2:239 two-stage repression 2:239 selective memory retrieval 1:214 unconscious conscious processing 1:331, 1:378, 2:412, 2:462 fringe consciousness feelings of knowing 2:448, 2:461 general discussion 2:460 immediate memory 2:331–332 spatial situatedness 2:187 tip-of-the-tongue state 2:461 frontal cortex creativity 1:438 mental disorders 2:248–249 neglect 2:71 visual systems 2:123 frontal eye fields 1:124, 1:124f frontotemporal dementias (FTDs) 2:262t, 2:263–264 Full Outline of Unresponsiveness 1:245t full outline of unresponsiveness (FOUR) 1:151 functional disorders 2:250t functional hallucinations 2:259 functional (hysterical) blindness 1:363, 2:417 functionalism 1:179–180, 1:283, 1:346, 2:48–49 functional magnetic resonance imaging (fMRI) aesthetic responses 1:5 affective response analysis 1:234, 2:382 automatic controlled processing research 1:90 binocular rivalry 2:79, 2:129 bistable perception anatomical-location versus state-change theories 1:95 binocular rivalry 1:101 general discussion 1:93, 1:99 motion perception 1:101, 1:102f
487
488
Subject Index
functional magnetic resonance imaging (fMRI) (continued) neural synchronization 1:104 parietal and prefrontal cortices 1:98f, 1:103 reversal-related activations 1:101, 1:102f visual processing 1:101 blindsight 1:117 creativity research 1:439, 1:441f habit formation processes 1:320 hypnosis research general discussion 1:361 hypnotic analogue research alien control delusions 1:364–365 clinical delusions 1:363 conversion disorder paralysis 1:363–364 disrupted perception 1:363 functional (hysterical) blindness 1:363 general discussion 1:363 hallucinations 1:364 malingering/faked conditions 1:365 instrumental research 1:363 intrinsic research 1:361 inner speech 1:393 intergroup processing 2:380, 2:384–385 mind wandering research 2:36, 2:67 neural encoding mechanisms 2:81 pain perception 1:246, 1:246f placebo response analysis 2:206, 2:212 religious/spiritual experiences 2:275 REM sleep 2:357–358, 2:362, 2:363f selective attention 1:62 selective memory retrieval 1:209, 1:212f skill acquisition and learning 2:14 theory of mind research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 language 2:401, 2:402f personality traits 2:405 social cognition 2:404 thought suppression 2:67 voluntary action 1:129, 2:115–116 fusiform gyrus intergroup processing 2:381 working memory control mechanisms 1:216
G Galilei, Galileo 1:341 Gallup, Gordeon 1:24 Galton, Francis 2:449 gamma-aminobutyric acid (GABA) anesthetic agents 1:296, 1:305f, 2:101 psychoactive drugs 2:221 religious/spiritual experiences 2:275, 2:276t, 2:277f, 2:279 sleep regulation anesthetic agents 1:307 NREM (non-REM) sleep 2:359, 2:359f REM sleep 2:359f, 2:360 gamma-hydrooxybutyric acid (GHB) 2:102 ganglion cells 2:436f, 2:437 gap-contingent research techniques 1:50 gaseous anesthetic agents 1:296 Gassendi, Pierre 1:343 gender concepts in language 1:458 general anesthesia See anesthesia general anxiety disorders 2:248, 2:252t, 2:253t general paralysis of the insane (GPI) 2:248–249 generative grammar 1:450, 2:413–414 Gerstmann’s syndrome 2:265 Gestalt psychology 1:2, 1:433, 2:412 Gibson, James 2:454 Gibson, JJ 2:292 gist situations attention awareness relationship 1:68 brain processing paradox 2:442 Glasgow Coma Scale (GCS) 1:150, 1:245t
Global Neuronal Workspace Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:90 neurocognitive panpsychism 2:98 global self-evaluation 2:317 global workspace theory anatomical location theories 1:95 artificial intelligence (AI) 1:38, 1:42 basic concepts 1:137, 1:158–159 first-order representational theory 1:163 information flow 2:58–59 integration consensus 1:288 neurobiological theories 2:90 perceptual awareness studies 2:84 glucose metabolism 1:152, 1:152f glutamate receptors anesthetics 2:101–102 mammalian avian comparison research 1:28–29 religious/spiritual experiences 2:275, 2:277f, 2:281 sleep regulation 2:359f glycine 2:101–102 goal pursuit 2:423–433 automatic thoughts and actions 2:426 conscious control action planning connection 1:179, 2:15 decision-making tasks 2:428 folk theory of mind concept 1:252, 1:253f general discussion 2:423, 2:427, 2:432 goal contagion 2:427 goal establishment 2:424 goal intentions 2:425 goal-related words 2:430 goal striving 2:427 habit-formation processes 1:318, 1:322, 2:426 habitual behaviors 2:426 implementation intentions 2:425 implicit social constructs 1:385 monitoring processes 2:429 planning and implementation stage 2:425 revision stage 2:430 subliminal affective priming 2:427–428 voluntary action 1:171 willful causation 1:325 Goodale, Melvyn 2:91 gradual-change research techniques 1:50 grammar 1:450, 2:413–414, 2:415 grandiose delusions 2:253t, 2:258t grasping abilities 2:125, 2:125f, 2:129, 2:130f gravity and light constraints 2:438, 2:439f Greenwald, Anthony 2:141 grouping problems 2:149 group relations 2:379–388 behavior modification studies 2:386 categorization 2:380 categorization evaluation relationship 2:382 complex situations 2:385 evaluation processes 2:381 general discussion 2:386 group formation 2:379 intergroup processing 2:380 motivational effects 2:383 same-race memory bias 2:381 self-categorization 2:383 growth-associated-protein-43 (GAP-43) 2:109–110 gTum-mo meditation 2:283 Gurney, Edmund 1:354 gustatory hallucinations 2:260–261 GY (research patient) 1:113, 1:115f
H habits 1:315–328 basic concepts 1:315 formation processes general discussion 1:316 goal-directed skills 1:318, 2:426 skill acquisition and learning 2:15 stimulus response conditions 1:317
Subject Index
general discussion 1:326 habit conscious will connection 1:316 James’ theories 2:462 nonconscious (automatic) processes attentional control changes 1:320 dual-task interference 1:320 explicit learning research 1:323 general discussion 1:319 goal-directed skills 1:322 implicit learning research 1:322 intention behavior prediction connection 1:321 neuroimaging research 1:320 sensemaking factors authorship experiences 1:325 experienced ease of retrieval 1:325 general discussion 1:324 self-perception 1:324 stimulus response conditions 1:316, 1:317 Haier, Richard 1:310 hair pulling 2:266–267 Hall and Van de Castle system 2:342 hallucinations auditory hallucinations 1:397, 2:249, 2:259 command hallucinations 2:255 control mechanisms 2:259 cultural and spiritual significance 2:223 dysfunctional self-talk 1:397 5-HT receptors 2:105 gustatory hallucinations 2:260–261 hallucinogens classifications and characteristics 2:222, 2:224t phenomenological characteristics 2:223 visual hallucinations 2:260 hypnosis 1:352 mania disorders 2:253t neuroimaging/hypnotic analogue research 1:364 olfactory hallucinations 2:237, 2:260–261 perception disorders 2:259 physical hallucinations 2:261 psychoactive drugs 2:220, 2:222 psychodynamic theories of the unconscious 2:235 religious/spiritual experiences 2:284 schizophrenia 2:109 sleep-related experiences 1:12 visual hallucinations 2:249, 2:260 haloperidol 2:222 halothane 1:296, 1:307f Hammeroff, Stuart 1:311 hard problem of consciousness general discussion 1:144, 2:375 machine consciousness 1:40 mind body problem 1:333 neurobiological theories 2:87 Hartmann, Eduard von 2:232 Hartmann, Ernest 2:350 Hawthorne effects 2:204 heavy metal poisoning 2:262t Heidbreder, Edna 1:334 Heidegger, Martin 2:181 Helmholtz, Hermann 2:412, 2:436, 2:462–463 hemianopia 1:112f hemispatial neglect 2:71, 2:149, 2:264 hemispherectomy 2:307 hepatic failure 2:262t Hermann grid 2:436–437, 2:436f heroin 2:222, 2:224t higher-order Bayesian decision theory of consciousness 1:143 higher-order consciousness 1:24, 1:284 higher-order thought theory basic concepts 1:165 conscious content 1:285 general discussion 1:348, 2:375 intentional inexistence 1:426 introspection theories 2:195–196 machine consciousness 1:40 high-level body awareness 2:290 high stress situations 1:86–87 high workload situations 1:86–87
hippocampus autobiographical memory 2:2 cognitive disorders 2:265 conscious awareness 1:189, 2:420 functional role 1:189 mammalian avian comparison research 1:28–29 memory functions 1:184, 1:185f, 1:186f, 2:15, 2:414 pain perception 1:246 recollective experiences 1:189 relational binding 1:189 religious/spiritual experiences 2:277f, 2:280 REM sleep 2:360 selective memory retrieval 1:212–213 visual systems 2:123 histamine 1:305f, 2:104 histrionic behaviors 2:254, 2:269t HIV 2:262t Hobbes, Thomas 1:343 hollow face illusions 2:131–132, 2:131f homunculus fallacy 1:136–137, 1:138, 2:452 Hudetz, Anthony 1:311–312 Hull, Clark 1:335 human organisms See personal identity Hume, David behaviorism 1:333 introspective awareness 2:448 personal identity studies 2:301, 2:303 self-awareness theories 2:294 hum speech 1:447 Huntington’s disease 2:262t, 2:264, 2:267 Husserl, Edmund 1:200, 1:343, 1:428, 2:176, 2:179 Huxley, Aldous 2:104 Huxley, Thomas 1:123, 1:334, 1:343–344, 2:87–88 hyperactivity 1:398 hyperemesis gravidarum 2:262t hyperemotionalism 2:253t, 2:254–255 hyperglycemia 2:262t hyperthyroidism 2:251–252 hypnagogic state 1:12, 2:248 hypnopompic state 1:12, 2:248 hypnosis 1:351–368 altered states of consciousness 1:14, 1:353 amnesia 1:352, 2:414 basic concepts 1:351 Charcot’s theories 2:233 confabulations 2:4–5 future research 1:365 historical background 1:353 hypnotic induction basic concepts 1:14, 1:351 executive control/executive monitoring model 1:359 hypnotizability measurements 1:355 intrinsic research 1:361 hypnotic suggestibility general discussion 1:14–15 historical research 1:354, 1:355 hypnotic nonhypnotic suggestibility comparisons 1:357 hypnotizability measurements 1:355 individual aptitude 1:356 Janet’s theories 2:234 neural mechanisms 2:219–220 neuroimaging research general discussion 1:361 hypnotic analogue research alien control delusions 1:364–365 clinical delusions 1:363 conversion disorder paralysis 1:363–364 disrupted perception 1:363 functional (hysterical) blindness 1:363 general discussion 1:363 hallucinations 1:364 malingering/faked conditions 1:365 instrumental research 1:363 intrinsic research 1:361 posthypnotic suggestions 1:351–352 as rational therapy
489
490
Subject Index
hypnosis (continued) Breuer’s therapies 2:236 Delboeuf ’s therapies 2:236 Janet’s therapies 2:235 spontaneous amnesia 1:12 theoretical perspectives dissociation theories 1:359, 1:359f executive control/executive monitoring model 1:359, 1:359f explicit implicit effects 1:358 general discussion 1:357 illusion of control concept 1:360 primary impacts 1:358 trances 1:353 hypnotics (drugs) 2:221, 2:224t hypnotic virtuosos 1:14, 1:355 hypochondriasis 2:253t, 2:258t hypocretins 2:104 hypodermic syringe 1:296 hypoglycemia 2:252t, 2:262t hypomania 2:253–254, 2:253t hypothalamus arousal system 1:304, 1:305f dreaming 2:363f mammalian avian comparison research 1:28 neurochemical mechanisms 2:108–109 pain perception 1:246 religious/spiritual experiences 2:275, 2:277f, 2:280, 2:281, 2:284 sleep regulation 1:307, 2:359f hypoxia 2:125–126 hypthyroidism 2:262t hysteria Breuer’s theories 2:234 Charcot’s theories 2:233 Freud’s theories 2:234 Janet’s theories 2:234 psychopathology 2:248–249 research background 2:233 hysterical blindness 1:363, 2:417
I iconic gesture systems 1:455 id 2:240 identification tasks 1:49, 1:54 identity perception 2:263 identity theory 1:345–346 ideomotor theory 1:172, 1:286, 1:352, 1:355–356 illegal substances 2:255, 2:262t illogicality 2:256t illusory conjunctions Balint Holmes syndrome 2:73 binding problem 2:149 selective attention 1:64 imageless thought controversy 1:332, 2:448 imagination 2:57, 2:447 See also mental imagery imagined feelings 1:79 imitation behavior disorders 2:118 immediate memory 2:327–339 attentional control 2:328 capacity limits to focus of attention 2:335, 2:336f, 2:413 definitions 2:327 general discussion 2:337 theoretical models 2:329, 2:330f time-dependent activated memory auditory stream segregation 2:333f basic concepts 2:331 discrete moment versus continuous moment hypothesis 2:332f partial report procedure 2:333f schematic diagram 2:330f tone memory experiment 2:335f impasse 1:435 implicit association test (IAT) 1:386–387 implicit learning 1:369–381 characteristics 1:369 general discussion 1:379 habits 1:322
process dissociation procedure 1:377 research challenges 1:375 research methodology artificial grammar 1:372, 1:372f, 1:374f dynamic system control 1:372–373 finite-state grammar 1:372, 1:372f sequence learning 1:372f, 1:373, 2:14 skill acquisition and learning 2:14 unconscious cognition 2:415 unconscious information processing 1:378 variable/qualitative-based differences 1:379 implicit memory 1:369–381 amnesia 1:184, 1:370 characteristics 1:186, 1:369, 2:1, 2:263t definition 1:184 functional role 1:188 general discussion 1:379 implicit explicit interactions 1:190, 2:414 neuroanatomy 1:187 procedural memory 1:186, 2:1, 2:15, 2:263t process dissociation procedure 1:377 research challenges 1:375 research developments 1:190 research methodology 1:370 unconscious cognition 2:414 unconscious information processing 1:378 variable/qualitative-based differences 1:379 implicit perception change blindness 1:50 inattentional blindness 1:54 unconscious cognition 2:416 unconscious conscious perception 2:144 implicit social cognition 1:383–388 automatic thoughts and actions 1:383 general discussion 1:387 social constructs attitudes 1:384 automatic social behavior 1:385 construct formation and change processes 1:385 general discussion 1:384 goal pursuit 1:385 measurement techniques 1:386 self-concept 1:385 stereotypes 1:384 implicit thought 2:418 inarticulateness 2:256t inattentional blindness 1:47–59 attention awareness relationship 1:67, 1:233–234, 2:85 basic concepts 2:335 bllindness amnesia comparisons 1:57 change blindness inattentional blindness comparisons 1:56 experimental approaches attentional manipulation 1:55 general discussion 1:54 perceptual set types 1:55 response types 1:54 task types 1:54 expression suppression relationship 1:53 general discussion 1:47, 1:58 implicit perception 2:417 research background 1:52, 1:53f research implications general discussion 1:55 scene perception 1:56 visual attention 1:56 restricted unrestricted effects comparisons 1:53 unconscious information processing 2:167 visual attention visual experience comparisons 1:58 incidental perceptual sets 1:50 incoherent speech 2:256t incompatibilism 1:265 incongruous affect disorder 2:253t, 2:254–255 incorrigibility 1:260 increased rate of speech 2:256t incubation 1:436 indicator semantics (Dretske) 1:424 infections 2:262t inferior frontal junction (IFJ) 1:90, 1:90f
Subject Index
inferior parietal lobule neglect 2:71 religious/spiritual experiences 2:280 inferior temporal (IT) cortex implicit memory 1:187 religious/spiritual experiences 2:284 visual systems basic concepts 1:62–63 binocular rivalry 2:128–129 bistable perception 1:101 dorsal ventral streams 2:123 schematic diagram 1:62f, 2:123f information integration theory (IIT) 1:403–416 basic concepts 1:139 computer simulation studies 1:407 digital camera analogy 1:405 general discussion 1:415 integration consensus 1:291 machine consciousness 1:39 measurement techniques complex determinations 1:407 general discussion 1:405 information measures 1:405 integration measures 1:406 neurobiological theories 1:407, 2:90 phenomenological subtraction and addition 1:403 phi (F) function 1:139, 1:291, 1:406 photodiode analogy 1:403 point-of-view implications 1:405 quality of consciousness generation mechanisms 1:409 qualia space (Q-space) analyses experience properties 1:411 experience shape 1:410 geometrical analyses 1:412 inhaled anesthetic agents 1:296, 1:307f inhibitory control 1:207, 2:403 See also executive control/executive monitoring model inner sense concepts 1:163 inner speech 1:389–402 definitions 1:389 developmental characteristics 1:392 dysfunctional self-talk autism 1:398 general discussion 1:396 hyperactivity 1:398 schizophrenia 1:397 functions language skills 1:395 memory 1:396 miscellaneous functions 1:396 self-regulation 1:394 task-switching performance 1:396 general discussion 1:400 inner speech writings 1:389 measurement techniques 1:391 neuroanatomy 1:393 self-awareness theories 1:398 theoretical perspectives 1:389 theories of consciousness 1:40 insects under skin 2:261 insight characteristics 2:250t cognitive disorders 2:265 problem-solving experiences 1:433, 1:434f, 1:444f, 2:68 , 2:418 See also creativity instinctual drives 2:239 insula pain perception 1:246 placebo response analysis 2:206–207, 2:213 integration consensus boundary factors 1:290 differentiation processes 1:290 general discussion 1:288 global workspace theory 1:288 skill acquisition and learning 1:289, 2:13 theory of neuronal group selection (TNGS) 1:291
intentionality 1:417–429 basic concepts 1:419 behavior/reason explanations 1:254 causal history factors 1:254–255 consciousness definitions 1:419 consciousness development processes 1:222 consciousness intentionality relationship 1:417, 2:22, 2:179 embodied intentionality 1:428 folk theories basic concepts 1:253, 1:253f behavior/reason explanations 1:254 language of mental activities 1:255 observability 1:256f Husserl’s theory 1:428 intentional inexistence analytical approaches 1:421 biosemantics (Millikan) 1:425 Brentano’s thesis 1:423 higher-order thought theory 1:426 indicator semantics (Dretske) 1:424 language structure analysis 1:421 mental states 1:256, 1:420 Representation Theory of Mind (RTM) 1:423 longitudinal intentionality 1:202 mental imagery 2:445–457 background information 2:446 definitions 2:446 general discussion 2:456 historical research 2:447 subjective individual differences 2:449 theoretical perspectives description (mentalese) theory 2:453 enactive theory 2:454 general discussion 2:450 picture theory 2:451, 2:451f mental representations 2:19–32 basic concepts 2:19 general discussion 2:19 misrepresentations of experiences 2:20, 2:21 modes of presentation 2:20–21 nonrepresentational theories 2:31 phenomenal consciousness 2:22 psychosemantic theories 2:21 representational theories basic concepts 2:23 criticisms 2:26 enactive theory 2:31 externalism 2:27, 2:58, 2:59f internalism 2:29, 2:58, 2:59f narrow content 2:29 subjectivity 2:30 weak versus strong representationalism 2:31 teleosemantic theories 2:22 Merleau-Ponty’s theory 1:428, 2:181 phenomenal consciousness 1:427 phenomenological characteristics 2:179 sensorimotor intentions 1:125–126, 1:125f transverse intentionality 1:202 voluntary action 1:125, 1:125f, 1:253, 1:253f, 2:115 intentional perceptual sets 1:50 intentional states 1:40 intention tremors 2:267 intermediate level theory of consciousness 1:138 internalism 2:29, 2:58, 2:59f, 2:180–181 International Classification of Diseases (ICD) 2:247 interpretive psychopathology 2:246 interspersed stimuli 1:55 intersubjectivity 2:183 intonational gestures 1:447–448 intralaminar thalamus 1:308, 1:309f intravenous anesthetic agents 1:296, 1:307f intrinsicalism 1:166 introspection 2:187–199 characteristics general discussion 2:188 immediacy 2:190 infallibility 2:190 mental states 2:188
491
492
Subject Index
introspection (continued) pain awareness 2:190 privileged access 2:189 transparency 2:190 conscious inessentialism 1:280 nineteenth-century studies 1:331, 2:448 privileged access concept 1:258 rational action 1:287 self-consciousness 2:177 self-reporting issues 1:331 theoretical perspectives displaced perception theories 2:195 evidence theories 2:192 general discussion 2:191 perceptual theories 2:191 self-knowledge concept 2:194 transcendental theories 2:194 intrusive memories 1:210, 1:215, 2:258t intuition 2:50, 2:418 See also creativity intuitive knowledge 1:369 invariance 2:83–84 invertebrates 1:29 Inverted Earth concept 2:26–27 involuntary movements 2:267 ironic processing theory 1:216, 2:60 isocortex 1:28 isoflurane 1:296, 1:307f Iyengar, Shanto 2:376
J jamais ecoute 2:262–263 jamais vu 2:262–263 James, William 2:459–468 automaticity 1:83 biographical background 2:466 depth psychology 2:231 fringe consciousness feelings of knowing 2:461 general discussion 2:460 immediate memory 2:331–332 spatial situatedness 2:187 tip-of-the-tongue state 2:461 general discussion 2:467 on habit formation processes 2:462 ideomotor theory 1:286 introspective awareness 2:448 mental states 2:412 mind body problem free will theory 2:463 general discussion 2:463 philosophical controversies 1:333, 2:465 on nest-building 2:462 panpsychism 2:466 research background 1:330, 2:460 self-identity studies 2:375 specious present doctrine 1:196 streams of consciousness 1:354–355, 2:58, 2:412 unconscious conscious processing 1:378 unconscious intelligence 2:462 voluntary action studies 1:123, 1:172, 2:111 Janet, Pierre hypnosis 1:354, 2:235 unconscious hysterical symptoms psychodynamic theories 2:234 rational therapies 2:235 research background 2:231 unconscious processing 1:331, 2:462 jealousy 2:258t John, E. Roy 1:311 Jung, Carl Gustav 2:241, 2:346
K Kant, Immanuel 1:1–2, 1:343, 2:302 ketamine 1:296, 2:102, 2:222, 2:224t khat 2:221–222
Knight’s move thinking 2:256t knowing 1:78, 1:188–189, 2:448, 2:461 knowledge (autobiographical memory) 1:77 Koch, Christoph 1:206, 2:88 Korsakoff ’s syndrome 2:2, 2:264 Ku¨lpe, Oswald 2:390 Kundalini Yoga meditation 2:284
L Lamme, Viktor 2:93 language artificial-grammar learning 2:415 conscious processes 1:447–459 categorical perception 1:454 characteristics 1:447 conscious perception 1:453 general discussion 1:447, 1:458 meaningfulness 1:452 modularity 1:451 structural characteristics color perception 1:457 number perception 1:457–458, 1:458f sign language 1:454 thought language connection 1:456 symbolic representations 1:452 disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257 misidentification syndromes 2:257 self-harm risks 2:255 goal-related words 2:430 higher-order thought theory 1:167 inner speech 1:395 language of mental activities 1:255 language structure analysis 1:421 mental representations 2:21 neuroimaging research 2:401, 2:402f religious/spiritual experiences 2:280 sign language 1:395–396, 1:454 symbolic interactionism 2:375 theory of mind research 2:401, 2:402f thought language connection 1:389, 1:456 unconscious processing 1:449, 2:413–414 lateral dorsal tegmental nucleus (LDTN) 1:304, 1:305f, 2:102, 2:108–109 lateral geniculate nucleus (LGN) religious/spiritual experiences 2:277f, 2:279 visual systems basic concepts 1:62–63 cortical/subcortical processing 1:101 schematic diagram 1:62f, 2:123f visual pathways 1:119, 1:119f lateral hypothalamus 1:305f, 2:282 lateral inhibition 2:436 lateral occipital complex (LOC) 1:101, 1:306, 1:307f, 2:77–78, 2:83 lateral occipitotemporal cortex 1:97 lateral prefrontal cortex 2:364 laterodorsal tegmentum (LDT) 2:360 Lau, Hakwan 1:143 Lazarus, Richard 2:376 left inferior frontal gyrus (LIFG) 1:393, 1:397 Leibniz, Gottfried Wilhelm 1:342, 2:303, 2:412 levels of consciousness 1:94, 1:221, 1:225f Levine, J 1:159 Lewy-body dementia 2:262t lexical decision tasks 2:430 Lhermitte, F 2:118 libertarianism definition 1:265 theoretical perspectives 1:266 Libet, Benjamin 1:127, 1:128f, 1:173, 1:269, 1:280, 2:115 libido 2:242 Liepmann, Hugo 2:111 life instinctual drive 2:239 light-sensitive organisms 2:122 limbic system memory functions 1:184 neurochemical mechanisms 2:108–109
Subject Index
religious/spiritual experiences 2:275, 2:280 voluntary action 1:124f, 1:125f limen theory 2:412 Livingstone Hubel model 2:151 local field potentials (LFPs) 1:99–100 localization tasks 1:49, 1:54 Local Recurrence Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:93 neurocognitive panpsychism 2:98 locked-in syndrome (LIS) clinical definitions 1:150 historical research 1:148 illustrative diagram 1:148f sleep states 2:370 Locke, John 1:159, 2:301, 2:448 locus coeruleus 1:246, 1:305f, 1:307, 2:108–109, 2:281 Loeb, Jacques 1:334 Logical Positivists 1:344 logic-based artificial intelligence 1:37 Loligo 1:31–32 longitudinal intentionality 1:202 long-term memory autobiographical memory 1:77 consciousness development processes 1:223 Lotze, Hermann 1:172 love drug 2:226–227 low-level body awareness basic concepts 2:290 body-internal spatial coding 2:291 moment-to-moment representation 2:291 neurological mechanisms 2:292 object-relative spatial coding 2:291 orientational coding 2:292 proprioception 2:292 LSD (lysergic acid diethylamide) classifications and characteristics 2:222, 2:224t color perception 2:227f phenomenological characteristics 2:104, 2:220f, 2:225 religious/spiritual experiences 2:284 time perception disorders 2:262 visual hallucinations 2:260 lucid dreaming characteristics 1:14, 1:16, 2:364 illustrative diagram 1:148f
M Mabbott, JD 1:197 macaque monkeys visual system research binding problem 1:64–65 bistable perception 1:101 blindsight 1:26 neuronal synchronization 2:154 selective attention 1:63 See also primates Mach, Ernst 1:345 machine consciousness behavioral tests 1:43 criticisms algorithmic factors 1:40 Chinese Room experiment 1:40 hard problem of consciousness 1:40 limitations of artificial intelligence 1:41 current research associated cognitive characteristics 1:41 associated external behaviors 1:41 consciousness correlates 1:42 future developments 1:42 global workspace theory 1:42 Internet connectivity 1:42 phenomenally conscious machines 1:42 definition 1:38 internal architecture tests 1:43 personhood criteria 1:257 positive theories
493
cognitive theories 1:39 higher-order thought theory 1:40 information integration theory (IIT) 1:39 inner speech 1:40 natural selection theory 1:39 madness 2:248 magnetic diseases 2:235 magnetic resonance imaging functional magnetic resonance imaging (fMRI) aesthetic responses 1:5 affective response analysis 1:234, 2:382 automatic controlled processing research 1:90 binocular rivalry 2:79, 2:129 bistable perception anatomical-location versus state-change theories 1:95 binocular rivalry 1:101 general discussion 1:93, 1:99 motion perception 1:101, 1:102f neural synchronization 1:104 parietal and prefrontal cortices 1:98f, 1:103 reversal-related activations 1:101, 1:102f visual processing 1:101 blindsight 1:117 creativity research 1:439, 1:441f habit formation processes 1:320 hypnosis research general discussion 1:361 hypnotic analogue research 1:363 instrumental research 1:363 intrinsic research 1:361 inner speech 1:393 intergroup processing 2:380, 2:384–385 mind wandering research 2:36, 2:67 neural encoding mechanisms 2:81 pain perception 1:246, 1:246f placebo response analysis 2:206, 2:212 religious/spiritual experiences 2:275 REM sleep 2:357–358, 2:362, 2:363f selective attention 1:62 selective memory retrieval 1:209, 1:212f skill acquisition and learning 2:14 theory of mind research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 language 2:401, 2:402f personality traits 2:405 social cognition 2:404 thought suppression 2:67 voluntary action 1:129, 2:115–116 magnetoencephalography (MEG) bistable perception analysis 1:99, 1:104 human consciousness studies 1:25 neural encoding mechanisms 2:81 placebo response analysis 2:207 voluntary action studies 1:129 Malebranche, Nicholas 1:342–343 malingering/faked conditions 1:365, 2:248–249 mammals brain research 1:27 visual system research neuronal synchronization 2:154 visually-guided behaviors 2:122–123 mammillary bodies 1:185f, 1:186f mania 2:253–254, 2:253t, 2:258t, 2:260 manic-depressive psychosis 2:248–249 mannerisms 2:266 manual gestures 1:448 marijuana (Cannabis sativa) classifications and characteristics 2:221, 2:224t color perception 2:227f phenomenological characteristics 2:227–228 Mashour, George 1:311 masking techniques backward masking 2:77, 2:170 feature inheritance 2:170, 2:170f memory capacity 2:336, 2:336f metacontrast masks 2:78
494
Subject Index
masking techniques (continued) pattern masks 2:77 unconscious conscious perception 2:144 visual crowding 2:166f, 2:167 visual masking 2:77, 2:92–93, 2:166f, 2:167 materialism theory 1:345–346 McGinn, Colin 2:50 McTaggart, J 1:195 Mead, Herbert 2:375 medial prefrontal cortex (MPFC) automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 general discussion 2:401–402 language processing 2:402–403 personality traits 2:405 social cognition 2:404 medial superior temporal (MST) cortex 1:62–63, 1:62f medial temporal lobes amnesia 2:414 anterior communicating artery aneurysms 2:3 autobiographical memory 2:2 conscious awareness 1:189, 2:420 implicit memory 1:187 mental disorders 2:248–249 visual processing 2:129 meditation altered states of consciousness 2:273 electroencephalography (EEG) 2:283 exceptional/higher states of consciousness 1:15 Kundalini Yoga meditation 2:284 mindfulness meditation 2:39, 2:274, 2:283 neural mechanisms 2:219–220 neuroimaging techniques 2:275 psychological research 2:285 samadhi states 1:15–16 Tibetan Buddhism 2:274, 2:283, 2:284 transcendental meditation 2:220–221, 2:274, 2:283 Yoga meditation 2:283, 2:284 Yoga Nidra guided meditation 2:284 Zen meditation 2:283 medulla 2:359f melanin-concentrating hormone (MCH) 1:305f melatonin 2:276t, 2:282 memory 1:183–192, 2:1–11 amnesia 1:184, 1:370 amnesia confabulation correlations 2:8 anterior communicating artery aneurysms 2:3 autobiographical memory 1:77–82 characteristics 2:1 cognitive feelings familiarity 1:77–78 knowing 1:78 remembering 1:77 conscious feelings 1:79 definition 1:77 egocentric representations 2:9 imagined feelings 1:79 inner speech 1:396 Korsakoff ’s syndrome 2:264 malfunctions amnesia 1:79 confabulations 1:80, 2:3 de´ja` vu 1:80–81 false memories 1:80, 2:4 neuropsychology 2:2 biographical memory 2:9 cognitive disorders 2:263, 2:263t confabulations altered states of consciousness 1:12 amnesia confabulation correlations 2:8 anterior communicating artery aneurysms 2:3 basic concepts 2:5 domain specificity 2:8 dream content analysis 2:343 executive theories 2:5–6 false memories 2:4 hypnosis 2:4–5
irrelevant memory suppression 2:7 memory malfunctions 1:80 reality monitoring 2:6 retrieval theories 2:5–6 declarative memory 2:15, 2:263t definition 1:184 domain specificity 2:8 episodic memory characteristics 1:369, 2:263t definition 1:77 dreaming 2:347 egocentric representations 2:9 neuropsychology 2:2 Think/No-Think paradigm (TNT) 1:213–214 working memory control mechanisms 1:216 explicit memory amnesia 1:184, 1:189, 2:414 characteristics 1:188, 1:369, 2:1, 2:263t definition 1:184 implicit explicit interactions 1:190, 2:414 neuroanatomy 1:189 research developments 1:190 research methodology 1:370 false memories 1:80, 2:4 general discussion 2:10 iconic memory 2:161 immediate memory 2:327–339 attentional control 2:328 capacity limits to focus of attention 2:335, 2:336f, 2:413 definitions 2:327 general discussion 2:337 theoretical models 2:329, 2:330f time-dependent activated memory auditory stream segregation 2:333f basic concepts 2:331 discrete moment versus continuous moment hypothesis 2:332f partial report procedure 2:333f schematic diagram 2:330f tone memory experiment 2:335f implicit memory 1:369–381 amnesia 1:184, 1:370 characteristics 1:186, 1:369, 2:1, 2:263t definition 1:184 functional role 1:188 general discussion 1:379 implicit explicit interactions 1:190, 2:414 neuroanatomy 1:187 procedural memory 1:186, 2:1, 2:15, 2:263t process dissociation procedure 1:377 research challenges 1:375 research developments 1:190 research methodology 1:370 unconscious cognition 2:414 unconscious information processing 1:378 variable/qualitative-based differences 1:379 inner speech 1:396 irrelevant memory suppression 2:7 Korsakoff ’s syndrome 2:2 long-term memory autobiographical memory 1:77 consciousness development processes 1:223 restructuring experiences 1:435 memory cues 2:7 neuroanatomy 1:184, 1:185f, 1:186f nondeclarative memory 2:263t priming 1:190 prospective memory 2:263t psychological continuity 2:306 reality monitoring 2:6 recollective experiences amnesia 1:189 autobiographical memory 1:77 functional role 1:191 recovered memories 2:37 relational binding 1:189 response override hypothesis 1:205–219 basic concepts 1:205, 1:206f general discussion 1:217
Subject Index
perception awareness research 1:206 retrieval stopping processes Freudian suppression model 1:214 individual differences 1:214 intrusive memory awareness 1:215 neurobiological theories 1:212 Think/No-Think paradigm (TNT) 1:210, 1:211f, 1:212f, 1:213f thought suppression 1:216 working memory control mechanisms 1:216 selective retrieval inhibition mechanisms 1:209 inhibitory control processes 1:207 neurobiological theories 1:209 retrieval-induced forgetting (RIF) 1:206f, 1:208 retrieval stopping processes 1:210 restructuring experiences 1:435 semantic memory 1:188, 1:448–449, 2:2, 2:9, 2:263t sensory memory 2:327–339 capacity limits to focus of attention 2:335, 2:336f, 2:413 definitions 2:327 general discussion 2:337 theoretical models 2:329, 2:330f time-dependent activated memory auditory stream segregation 2:333f basic concepts 2:331 discrete moment versus continuous moment hypothesis 2:332f partial report procedure 2:333f schematic diagram 2:330f tone memory experiment 2:335f short-term memory change blindness 1:51, 2:441 insight problem-solving experiences 1:435–436 theories of consciousness 1:136 visual short-term memory (vSTM) 1:48, 1:51, 2:440 skill acquisition and learning 2:13–18 action pattern organization 2:14 general discussion 2:13, 2:17 neuroimaging research 2:14 perceptual-motor learning 2:13 sequence learning 2:14 source memory 2:1–2 systematic processing 2:7 time-dependent activated memory auditory stream segregation 2:333f basic concepts 2:331 discrete moment versus continuous moment hypothesis 2:332f partial report procedure 2:333f schematic diagram 2:330f tone memory experiment 2:335f transsaccadic memory 1:48 visual experience 2:435–443 brain processing paradox 2:441, 2:442f change blindness 2:441 general discussion 2:435, 2:442 rich experience impoverished input paradox color perception 2:438, 2:438f edge alignment 2:437, 2:437f edge enhancement 2:436, 2:436f edge-filling abilities 2:437, 2:438f face recognition 2:439–440 feature exaggeration 2:439 gravity and light constraints 2:438, 2:439f Hermann grid 2:436–437, 2:436f heuristic processes 2:438, 2:439f prior knowledge 2:439 psychophysical research 2:435 rich experience limited report paradox 2:440 subjective experiences 2:390 visual short-term memory (vSTM) 1:48, 1:51, 2:440 working memory altered states of consciousness 2:219 attentional control 2:62f, 2:64 automatic thoughts and actions 1:87 capacity limits 2:335, 2:336f, 2:413 characteristics 2:263t consciousness development processes 1:223 control mechanisms 1:216 general discussion 2:337
habit formation processes 1:320 information flow 2:58, 2:59f inner speech 1:396 insight problem-solving experiences 1:435–436, 2:68 theories of consciousness 1:136 meningitis 2:262t mental causation 1:141, 1:347, 2:46 mental disorders 2:245–271 cognitive disorders basic concepts 2:261 delirium 2:261, 2:262t dementia 2:261, 2:262t identity perception 2:263 insight 2:265 memory 2:263, 2:263t place perception 2:262–263 self-perception 2:264 time perception 2:262 visuospatial information 2:264 diagnostic concerns 2:268 disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257 misidentification syndromes 2:257 self-harm risks 2:255 emotional disorders 2:252 general discussion 2:245 mood and affect disorders 2:252, 2:253t movement disorders characteristics 2:265 involuntary movements 2:267 voluntary movements 2:266 perception disorders first rank symptoms 2:260t general discussion 2:258 hallucinations 2:259 top-down/bottom-up cognitive processing 2:259 phenomenological characteristics delirium versus dementia 2:249–250 form versus content 2:251 functional versus organic disorders 2:248–249 general discussion 2:247 levels of explanation 2:251, 2:252t psychosis versus neurosis 2:248 reactive versus instrumental behavior 2:251 terminology 2:250t visual versus auditory hallucinations 2:249 psychiatric diagnoses 2:247 psychopathology approaches 2:246 mentalese theory 2:453 mental imagery 2:445–457 background information 2:446 definitions 2:446 general discussion 2:456 historical research 2:447 subjective individual differences 2:449 theoretical perspectives description (mentalese) theory 2:453 enactive theory 2:454 general discussion 2:450 picture theory 2:451, 2:451f top-down/bottom-up cognitive processing 2:259 See also mind wandering mentalism 2:464 mental representations 2:19–32 basic concepts 2:19 general discussion 2:19 misrepresentations of experiences 2:20, 2:21 modes of presentation 2:20–21 nonrepresentational theories 2:31 phenomenal consciousness 2:22 psychosemantic theories 2:21 representational theories basic concepts 2:23 criticisms 2:26 enactive theory 2:31 externalism 2:27, 2:58, 2:59f internalism 2:29, 2:58, 2:59f
495
496
Subject Index
mental representations (continued) narrow content 2:29 subjectivity 2:30 weak versus strong representationalism 2:31 teleosemantic theories 2:22 mental time travel 1:191 mental tokens 2:45 mental types 2:47 mere exposure effects 2:376 Merikle, Philip 2:139 Merleau-Ponty, Maurice 1:428, 2:181 mescaline classifications and characteristics 2:222, 2:224t color perception 2:227f phenomenological characteristics 2:104, 2:220f, 2:225 mesencephalic reticular formation (MRF) 1:304 Mesmer, Franz Anton 1:353 mesmerism 1:353 mesopontine tegmental anesthesia area (MPTA) 1:308 mesopontine tegmentum 1:304 metabolic diseases 2:262t metacognition animal consciousness 1:24 insight problem-solving experiences 1:433, 1:434f, 2:68, 2:418 metaconsciousness/meta-awareness 2:33–41 affective attributions 1:239, 2:37 automatic thoughts and actions 1:88–89 basic concepts 2:33 dissociation theories general discussion 2:35 temporal dissociations 2:35 translation dissociations 2:38 general discussion 2:40, 2:375 historical research 2:34 mindfulness meditation 2:39 monitoring processes 2:34 temporal dissociations flow experiences 2:37 general discussion 2:35 mind wandering 2:33, 2:35, 2:61 recovered memories 2:37 unwanted thoughts 2:36, 2:67 verbal overshadowing 2:38 visual system research 2:123 metaconsciousness/meta-awareness 2:33–41 affective attributions 1:239, 2:37 automatic thoughts and actions 1:88–89 basic concepts 2:33 dissociation theories general discussion 2:35 temporal dissociations 2:35 translation dissociations 2:38 general discussion 2:40, 2:375 historical research 2:34 mindfulness meditation 2:39 monitoring processes 2:34 temporal dissociations flow experiences 2:37 general discussion 2:35 mind wandering 2:33, 2:35, 2:61 recovered memories 2:37 unwanted thoughts 2:36, 2:67 verbal overshadowing 2:38 metacontrast masks 2:78 metalinguistic knowledge 1:449 metamemory functions knowing 1:78 noetic consciousness 1:78 methadone 2:222 methaqualone 2:221 methodological solipsism 2:29 methylxanthines 2:221–222 Meynert, Theodor 2:102, 2:232 microconsciousness theory 2:94 middle temporal (MT) cortex 1:62–63, 1:62f migraines 2:260, 2:262–263, 2:262t Miller, George 1:136, 1:334, 2:337 Millikan, R 1:425 Mill, John Stewart 1:343–344, 2:399 Milner, David 2:91
mimicry 2:376 mind-altering drugs 2:221 mind body problem 2:43–55 Aristotilean philosophy 1:341 basic concepts 2:43 epistemological interpretation 2:50 essence problem 2:49 explanatory gap 2:50 free will theory 2:463 general discussion 2:54 intuition criticisms 2:51 explanatory gap 2:50 general discussion 2:50 item problem functionalism theory 2:48–49 mental tokens 2:45 mental types 2:47 supervenience thesis 2:48 token-identity theory 2:45 type-identity theory 2:47–48 James’ theories 1:333, 2:463 mental states 2:43 nineteenth-century idealism theory Berkeley 1:343 Brentano 1:343, 1:423 Clifford 1:343 Darwin 1:343 general discussion 1:343 Husserl 1:343, 1:428 Kantian theory 1:343 philosophical barriers 1:333 ontological interpretation 2:53 philosophical perspectives 1:333, 2:464 placebo treatment 2:201 pre-Socratic philosophers 1:340 quantum phenomena 2:464 scientific revolution era Cartesian dualism 1:342, 2:45, 2:181, 2:452 Galilean philosophy 1:341–342 Gassendi 1:343 general discussion 1:341 Hobbes 1:343 Leibniz 1:342 Malebranche 1:342–343 Spinoza 1:342 mindfulness meditation 2:39, 2:274, 2:283 mindless reading 2:63, 2:64 mindset 1:444 mind’s eye function 2:259, 2:451 See also mental imagery mind wandering 2:57–69 basic concepts 2:33, 2:57 causal theories attentional control 2:62f, 2:64 current concerns 2:62f, 2:63 general discussion 2:61 task environment influences 2:62, 2:62f consequences action slips 2:65 general discussion 2:64 mindless reading 2:63, 2:64 future research directions 2:68 information flow attentional transitions 2:59f, 2:60 general discussion 2:58 mental states 2:58, 2:59f schematic diagram 2:59f mechanisms ironic processing theory 1:216, 2:60 low-level versus higher-level cognitive processes 2:60, 2:329 metaconsciousness/meta-awareness 2:33, 2:35, 2:61 neuroimaging research absent tasks 2:66 control mechanisms 2:67 default mode hypothesis 2:67 general discussion 2:66 minimal attachment 1:451–452 minimal consciousness 1:222, 1:225f minimally conscious state (MCS)
Subject Index
clinical definitions 1:149 diagnostic criteria 1:150, 1:244 external stimulation 1:153f historical research 1:148 illustrative diagram 1:148f pain assessment ethical considerations 1:247 general discussion 1:243, 1:248 pain perception 1:247 treatment strategies 1:153 mirror neurons 2:408 misfortunes 2:343–344 misidentification syndromes 2:257 misinformation effect 1:208 mismatched gestures 1:448 mitgehen 2:266 modularity theory 1:137, 1:451 moment of insight 1:439, 1:441f monkeys visual system research binding problem 1:64–65 bistable perception 1:101 blindsight 1:26, 2:128 early nonhuman primate evidence 1:109 neuronal synchronization 2:154 recent nonhuman primate evidence 1:111 selective attention 1:63 visually-guided behaviors 2:123 See also primates mood and affect disorders 2:252, 2:253t mood manipulation 1:237, 2:108, 2:431 Moore’s law 1:42–43 moral responsibility 1:268 Morgan, C Lloyd 1:343–344 morning glory 2:222 morphine 2:222, 2:224t Morton, William Thomas Green 1:298 motion-induced blindness 2:166, 2:170 motivation 2:324, 2:383, 2:430–431 motor cortex aesthetic responses 1:5 automatic controlled processing research 1:90, 1:90f REM sleep 2:362–363, 2:363f voluntary action 1:124, 1:124f, 1:125f, 2:111 motor responses 1:54 motor restlessness 2:255, 2:266 motor-skill learning 2:15 movement disorders characteristics 2:265 involuntary movements 2:267 voluntary movements 2:266 Muller Lyer illusion 1:451, 1:451f multi-infarct dementia 2:262t, 2:263–264 multiple drafts model of consciousness 1:140 multiple occupancy 2:307 multiple personalities 2:250t, 2:414 multiple sclerosis 2:262t mu(m)-opioid receptor (MOR) 2:213 muscular atonia 1:13 mushrooms 2:224t musical hallucinations 2:260 mutism 2:256t Myers, Charles Samuel 2:231 Myers, Frederic William Henry 1:354–355, 2:231 mystical experiences characteristics 1:19 psychedelics 2:225 psychotropic drugs 2:284
N Naccache, Lionel 1:158–159, 2:96 N-acetylaspartylglutamate (NAAG) 2:281 nail biting 2:266–267 naloxone 2:206, 2:222 narcissistic personality disorders 2:269t narcolepsy 1:308, 2:262t narcotics 2:222
narrative consciousness 2:161 narrow content 2:29 naturalistic dualism 2:54 naturalistic representations of intentionality 1:423 natural selection theory 1:25, 1:39 near-death experiences (NDEs) characteristics 1:18 psychedelics 2:225 Necker cube 1:96, 2:451f negative inner speech See also dysfunctional self-talk negativism 2:266 neglect 2:71–74 affective response analysis 1:235 attention issues 1:58 Balint Holmes syndrome 2:73 characteristics 2:71, 2:92–93 general discussion 2:73 stimuli processing 2:72 symptoms 2:72 visual hemineglect 2:85 neologisms 2:256t Neonatal Infant Pain Scale (NIPS) 1:244, 1:244t neoplastic diseases 2:262t nervous sleep 1:353–354 nervous system animal consciousness 1:25, 1:33 cephalopods 1:29 vertebrate development 1:28 neural Darwinism 1:291 Neural Darwinism 2:89 neural networks cephalopods 1:32 global workspace theory 1:42 machine consciousness research 1:41 neural synchronization 1:104 neuroaesthetics 1:2 neurobiological theories 2:87–100 access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 cognitive influences 2:87 dreaming 2:362 Duplex Vision Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:91 neurocognitive panpsychism 2:98 general discussion 2:99 Global Neuronal Workspace Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:90 neurocognitive panpsychism 2:98 global versus local theories 2:89 hard problem of consciousness 2:87 information integration theory (IIT) 1:407 Local Recurrence Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:93 neurocognitive panpsychism 2:98 microconsciousness theory 2:94 neural correlates 1:95, 2:88, 2:420 neural indexes 2:98 neurocognitive panpsychism 2:98 Reentrant Dynamic Core Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:89 neurocognitive panpsychism 2:98 response override hypothesis retrieval stopping processes 1:212 selective memory retrieval 1:212 neurochemical mechanisms 2:101–110 awareness alertness controls 2:103 clinical consciousness anesthetics 2:101 cholinergic system 2:102 coma 2:101 desferrioxamine/prochlorperazine doses 2:102
497
498
Subject Index
neurochemical mechanisms (continued) coma anesthetics 2:101 cholinergic system 2:102 desferrioxamine/prochlorperazine doses 2:102 induction mechanisms 2:101 general discussion 2:101 phenomenal consciousness acetylcholine 2:107 adrenergic system 2:108 bipolar disorder 2:108 mood manipulation 2:108 schizophrenia 2:109 serotonin 2:104 virtual reality processes 2:106 religious/spiritual experiences 2:276t, 2:281 neurocognitive panpsychism 2:98 neuroimaging techniques anesthetic effects 1:306, 1:307f creativity research divergent thought 1:438 moment of insight 1:439, 1:441f right-hemisphere priming 1:437 functional magnetic resonance imaging (fMRI) aesthetic responses 1:5 affective response analysis 1:234, 2:382 automatic controlled processing research 1:90 binocular rivalry 2:79, 2:129 bistable perception anatomical-location versus state-change theories 1:95 binocular rivalry 1:101 general discussion 1:93, 1:99 motion perception 1:101, 1:102f neural synchronization 1:104 parietal and prefrontal cortices 1:98f, 1:103 reversal-related activations 1:101, 1:102f visual processing 1:101 blindsight 1:117 creativity research 1:439, 1:441f habit formation processes 1:320 hypnosis research general discussion 1:361 hypnotic analogue research 1:363 instrumental research 1:363 intrinsic research 1:361 inner speech 1:393 intergroup processing 2:380, 2:384–385 mind wandering research 2:36, 2:67 neural encoding mechanisms 2:81 pain perception 1:246, 1:246f placebo response analysis 2:206, 2:212 religious/spiritual experiences 2:275 REM sleep 2:357–358, 2:362, 2:363f selective attention 1:62 selective memory retrieval 1:209, 1:212f skill acquisition and learning 2:14 theory of mind research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 language 2:401, 2:402f personality traits 2:405 social cognition 2:404 thought suppression 2:67 voluntary action 1:129, 2:115–116 habit formation processes 1:320 hypnosis research general discussion 1:361 hypnotic analogue research alien control delusions 1:364–365 clinical delusions 1:363 conversion disorder paralysis 1:363–364 disrupted perception 1:363 functional (hysterical) blindness 1:363 general discussion 1:363 hallucinations 1:364 malingering/faked conditions 1:365
instrumental research 1:363 intrinsic research 1:361 lateral occipital complex (LOC) 1:306, 1:307f mind wandering research absent tasks 2:66 control mechanisms 2:67 default mode hypothesis 2:67 general discussion 2:66 mirror neurons 2:408 positron emission tomography (PET) disorders of consciousness (DOC) 1:152f habit formation processes 1:320 hypnosis research general discussion 1:361 hypnotic analogue research 1:363 instrumental research 1:363 intrinsic research 1:361 inner speech 1:393 placebo response analysis 2:206, 2:212 religious/spiritual experiences 2:275 REM sleep 2:362, 2:363f skill acquisition and learning 2:14 voluntary action 2:116 priming methods 1:437 religious/spiritual experiences 2:275 single photon emission computed tomography (SPECT) 2:275 sleep 1:307f neurological diseases 2:262t neuronal group selection theory (TNGS) 1:291, 2:89 neuronal synchronization 2:153t, 2:154 neuropil 2:109 neuroses 2:248, 2:250t, 2:252t neurosyphilis 2:262t neurotransmitters arousal system 1:304, 1:305f creativity research 1:444 psychoactive drugs 2:220, 2:223, 2:226 religious/spiritual experiences 2:275, 2:277f Newton, Isaac 2:465 Nicaraguan Sign Language (NSL) 1:455 nicotine 1:308–309, 1:309f, 2:221–222 Nielsen, Torsten 2:112–113 nightmares 1:14, 2:345 nihilism 2:253t, 2:254, 2:258t nine-dot problem 1:433, 1:434f, 1:444f nitrous oxide 1:296, 2:102, 2:222 N-methyl-D-aspartate (NMDA) channels 1:297, 2:281 nociception 2:207, 2:212 nocturnal wandering 1:14 Noe¨, Alva 1:140, 1:141, 2:96, 2:454 noetic consciousness definition 1:77–78 semantic memory 1:188 nondeclarative memory 2:263t nonfluent dysphasias 2:256t nonknowing consciousness 1:186 nonrepresentational theories 2:31 nonverbal experiences 2:38 norepinephrine 1:305f, 1:444, 2:104, 2:276t, 2:281 normal pressure hydrocephalus 2:262t normal state of consciousness (NSC) 1:9 NREM (non-REM) sleep characteristics 2:359, 2:359f general discussion 2:370 historical research 2:357 subjective experiences 1:13 nucleus basalis of Meynert (NBM) 1:304, 2:102, 2:107 number perception 1:457–458, 1:458f numerical identity 2:304 nutritional deficiencies 2:262t
O obedience 2:266 object constancy 2:149 object recognition See feature integration theory obsessive compulsive disorder 2:248, 2:252t, 2:253t, 2:254, 2:258t, 2:269t
Subject Index
occipital lateral cortex 2:363f, 2:364 occipitotemporal cortex REM sleep 2:362, 2:363f visual systems 2:123f octopus 1:29 Octopus vulgaris 1:31–32 Oedipus complex 2:241 olfactory bulb 1:185f olfactory hallucinations 2:237, 2:260–261 one-way relation of dependence 2:48 ontological questions 2:303, 2:309 open-loop process 2:13 operation span test 2:329 opiates characteristics 2:222 classifications and characteristics 2:224t pain treatment 1:247 placebo analgesic effects 2:206, 2:210, 2:212 opportunistic assimilation hypothesis 1:436 optic ataxia Balint Holmes syndrome 1:176, 2:73 dorsal ventral streams 2:125f, 2:126 optic tectum 1:28 optimal experiences 1:16 orbitofrontal cortex (OFC) aesthetic responses 1:5 memory functions 2:7 pain perception 1:246 placebo response analysis 2:206–207, 2:211, 2:213 religious/spiritual experiences 2:284 O’Regan, Kevin 1:140, 1:141, 2:96, 2:454 orexins 1:305f, 1:308, 2:104 organic disorders 2:250t, 2:252t organism theory 2:303, 2:309 Ostwald, Wilhelm 1:345 Othello’s syndrome 2:258t Ouija board 1:354 out-of-body experiences (OBEs) characteristics 1:17 psychedelics 2:225 overlearned tasks See habits over-valued ideas 2:258t own-race advantage 2:439–440
P Pain Assessment Checklist for Seniors with Limited Ability to Communicate (PACSLAC) 1:244t Pain Assessment in Advanced Dementia Scale (PAINAD) 1:244, 1:244t pain research 1:243–250 introspective awareness 2:190 pain assessment analgesic use 1:247 behavioral coma scales 1:150, 1:151, 1:245t behavioral scales 1:244t ethical considerations 1:247 general discussion 1:243, 1:248 neural correlates 1:246 noncommunicative patients 1:244 pain perception 1:246, 1:246f placebo response analysis 2:205, 2:212 type-identity theory 2:47–48 palilalia 2:256t panic disorders 2:253t, 2:254 panpsychism 1:340, 1:343–344, 2:98, 2:466 Pappenheim, Bertha (Anna O.) 2:236 paraesthesia 2:254 parafoveal vision 2:417 parahippocampal gyrus autobiographical memory 2:2 dreaming 2:363f memory functions 1:184, 1:185f, 1:186f, 2:15 working memory control mechanisms 1:216 paralinguistic speech 1:447–448 parallel searches 2:148
paralysis research Charcot’s theories 2:233 conversion disorder paralysis 1:363–364 hysterical symptoms 2:233 Janet’s theories 2:234 malingering/faked conditions 1:365 paranoia 2:250t, 2:252t, 2:255, 2:269t paraphrasias 2:256t parapraxes 2:234 parasympathetic system 2:277f, 2:281, 2:282 parietal cortex bistable perception 1:98f, 1:103 cognitive disorders 2:264 dreaming 2:363f, 2:364 memory capacity 2:337 multisensory integration 2:152 neglect 2:71 physical hallucinations 2:261 religious/spiritual experiences 2:275, 2:284 sensory memory 2:330 parietal operculum 1:246 Parkinson’s disease dementia 2:262t, 2:264 involuntary movements 2:267 placebo response analysis 2:212, 2:214 partial report procedure 2:333, 2:333f passage of time 1:193, 1:300 passivity 2:258t, 2:260t pathogenic memory structure 2:237 pattern masks 2:77 Pavlov, IP 1:333 PCP See angel dust (PCP) pedunculopontine nucleus (PPN) 1:304, 1:305f, 1:308, 2:102, 2:360 pedunculopontine tegmentum (PPT) 2:360 Penrose, Roger 2:464 Pentothal 1:296 perception aesthetic perception 1:2 bistable perception 1:93–106 basic concepts 1:93, 1:94f first-person skepticism 2:397 general discussion 1:104 neural correlates binocular rivalry 1:100, 1:101 cortical/subcortical processing 1:101 extrastriate cortex 1:101 general discussion 1:99 measurement techniques 1:99 motion perception 1:101, 1:102f neural synchronization 1:104 parietal and prefrontal cortices 1:98f, 1:103 reversal-related activations 1:101, 1:102f theoretical perspectives 1:95 theoretical models behavioral evidence 1:97 binocular rivalry 1:99 high-level theories 1:98, 1:98f low-level theories 1:97, 1:98f theoretical perspectives anatomical location theories 1:95 consciousness correlates 1:94 neural correlates 1:95 state-change theories 1:95 visual stimuli ambiguous motion 1:94f, 1:96 binocular rivalry 1:97 gamma distribution 1:94f, 1:96 reversible figures 1:96 categorical perception 1:454 color perception color-digit synesthesia 2:150 color phi phenomenon 1:198 externalist theories 2:27–28 Inverted Earth concept 2:26–27 linguistic processes 1:457 microconsciousness theory 2:94 neural encoding mechanisms 2:80
499
500
Subject Index
perception (continued) psychoactive drugs 2:226, 2:227f visual experience 2:438, 2:438f decision-making tasks 2:76–77 explicit perception 1:50, 1:54, 2:416 face perception 1:2 implicit perception change blindness 1:50 inattentional blindness 1:54 unconscious cognition 2:416 unconscious conscious perception 2:144 implicit social cognition 1:383–388 automatic thoughts and actions 1:383 general discussion 1:387 social constructs attitudes 1:384 automatic social behavior 1:385 construct formation and change processes 1:385 general discussion 1:384 goal pursuit 1:385 measurement techniques 1:386 self-concept 1:385 stereotypes 1:384 introspection 2:191 linguistic processes 1:450, 2:413–414 Local Recurrence Theory 2:93 mental imagery 2:445–457 background information 2:446 definitions 2:446 general discussion 2:456 historical research 2:447 subjective individual differences 2:449 theoretical perspectives description (mentalese) theory 2:453 enactive theory 2:454 general discussion 2:450 picture theory 2:451, 2:451f number perception 1:457–458, 1:458f perception awareness research 1:206 perceptual learning 2:82 psychopathology first rank symptoms 2:260t general discussion 2:258 hallucinations 2:259 top-down/bottom-up cognitive processing 2:259 self-perception cognitive disorders 2:264 habits 1:324 self-specifying information 2:292, 2:295–296 subliminal perception 1:162, 1:324, 2:93, 2:144, 2:159–160, 2:416 unconscious conscious perception 2:135–146 action control 1:171–181 general discussion 2:135, 2:145 goal pursuit 2:423–433 automatic thoughts and actions 2:426 decision-making tasks 2:428 general discussion 2:423, 2:427, 2:432 goal contagion 2:427 goal establishment 2:424 goal intentions 2:425 goal-related words 2:430 goal striving 2:427 habitual behaviors 2:426 implementation intentions 2:425 monitoring processes 2:429 planning and implementation stage 2:425 revision stage 2:430 subliminal affective priming 2:427–428 historical research 2:412 linguistic processes 1:453 Local Recurrence Theory 2:93 model approaches null hypothesis model 2:138, 2:139f objective threshold/nonmonotonic approach 2:139f, 2:142 objective threshold/rapid decay approach 2:139f, 2:141 subjective threshold approach 2:139, 2:139f model validity 2:138 neural/psychological correlates 2:420
objective threshold approach 2:137, 2:139f, 2:141 subjective threshold approach 2:136, 2:139, 2:139f unconscious information processing 2:159–173 access consciousness 2:161 attention-related factors 2:171 binocluar rivalry 1:94f, 1:97, 1:99, 2:92–93, 2:165, 2:166f cognitive contents 2:160–161 criterion content 2:162 decision-making tasks 2:428 definitions 2:160 exhaustive exclusion 2:164 feedforward feedback connections 2:171 general discussion 2:171 habit-formation processes 1:323 implicit versus explicit learning 1:369, 1:378 indirect measurement strategies 2:164 invisible stimuli 2:75, 2:165, 2:166f methodological issues 2:162 narrative consciousness 2:161 nineteenth-century studies 1:331 normal versus neurologically impaired observers 2:165 phenomenal consciousness 2:161 postretinal processing 2:168 reflective consciousness 2:161 research challenges 1:375 research concerns 2:160 results and analysis 2:168 signal detection theory (SDT) 2:163 subliminal affective priming 1:233, 2:427–428 threshold measurements 2:163 unresolved issues attention consciousness relationship 2:96, 2:144 brain activation 2:144 direct indirect task relationship 2:143 implicit perception 2:144 masking techniques 2:144 subliminal perception 2:144 visual processing 2:127 unconscious information processing 2:159–173 attention-related factors 2:171 conscious level conscious content distinction 1:284 feedforward feedback connections 2:171 general discussion 2:171 implicit versus explicit learning 1:369, 1:378 invisible stimuli methods attentional limits 2:167 binocluar rivalry 1:94f, 1:97, 1:99, 2:92–93, 2:165, 2:166f flash suppression 2:166 general discussion 2:165 motion-induced blindness 2:166 semistabilized images 2:165, 2:166f transcranial magnetic stimulation (TMS) 2:82, 2:167 visual crowding 2:166f, 2:167 visual (two-transient) masking 2:77, 2:166f, 2:167 methodological issues criterion content 2:162 exhaustive exclusion 2:164 general discussion 2:162 indirect measurement strategies 2:164 invisible stimuli 2:75, 2:165, 2:166f normal versus neurologically impaired observers 2:165 signal detection theory (SDT) 2:163 threshold measurements 2:163 postretinal processing attentional limits 2:171 backward masking 2:170 binocular rivalry 2:169 blindsight 2:168 continuous flash suppression 2:170 feature inheritance 2:170, 2:170f form blindness 2:169 general discussion 2:168 motion-induced blindness 2:170 neurological patients 2:168 normal observers 2:169 transcranial magnetic stimulation (TMS) 2:170 research challenges 1:375 research concerns
Subject Index
access consciousness 2:161 cognitive contents 2:160–161 definitions 2:160 narrative consciousness 2:161 phenomenal consciousness 2:161 reflective consciousness 2:161 results and analysis general discussion 2:168 postretinal processing 2:168 visual experience 2:435–443 acetylcholine 2:107 brain processing paradox 2:441, 2:442f change blindness 2:441 general discussion 2:435, 2:442 neural mechanisms 2:75–86 attention awareness relationship 2:84 binocular rivalry 2:78 bistable perception 1:95 conscious content 2:79 cortical regions 2:82 discrimination tasks 2:76–77 encoding processes 2:79 general discussion 2:75, 2:85 perceptual threshold 2:75 reversible figures 2:78 stimuli processing 2:75 visual masking 2:77, 2:92–93, 2:166f, 2:167 rich experience impoverished input paradox color perception 2:438, 2:438f edge alignment 2:437, 2:437f edge enhancement 2:436, 2:436f edge-filling abilities 2:437, 2:438f face recognition 2:439–440 feature exaggeration 2:439 gravity and light constraints 2:438, 2:439f Hermann grid 2:436–437, 2:436f heuristic processes 2:438, 2:439f prior knowledge 2:439 psychophysical research 2:435 rich experience limited report paradox 2:440 serotonin 2:104 virtual reality processes 2:106 visual perception 2:292 visual systems 2:126 perceptual implicit memory 1:186 perceptual memory, 000 perceptual moment 2:331, 2:332f perceptual-motor learning 2:13 periacqueductal gray (PAG) arousal system 1:305f neurochemical mechanisms 2:108–109 pain perception 1:246 placebo response analysis 2:206–207, 2:213 sleep regulation 2:359f perifornical nucleus 1:305f perimetry general discussion 1:107 research background 1:109 peripheral nervous system (PNS) 2:381 perirhinal cortex autobiographical memory 2:2 familiarity 1:189 memory functions 1:184, 1:186f Perrin, Jean 1:345 persecutory delusions 2:258t perseveration 2:256t persistence question basic concepts 2:304 identity concerns 2:301 proposed solutions general discussion 2:305 personal-ontology viewpoint 2:309 psychological continuity 2:306 too-many-minds problem 2:310 psychological continuity basic concepts 2:306 beginning and ending of life 2:308
fission factor 2:307 multiple occupancy 2:307 personal-ontology viewpoint 2:309 too-many-minds problem 2:310 persistent vegetative states See vegetative states personal identity 2:301–312 concerns persistence question 2:301, 2:306 personal-ontology questions 2:303, 2:309 population question 2:302 selfish concern 2:304 general discussion 2:301 persistence question basic concepts 2:304 identity concerns 2:301 proposed solutions general discussion 2:305 personal-ontology viewpoint 2:309 psychological continuity 2:306 too-many-minds problem 2:310 psychological continuity basic concepts 2:306 beginning and ending of life 2:308 fission factor 2:307 multiple occupancy 2:307 personal-ontology viewpoint 2:309 too-many-minds problem 2:310 temporal parts theory 2:311 personality disorders dissociative amnesia 2:414 phenomenological characteristics levels of explanation 2:252t terminology 2:250t psychopathology 2:254, 2:267, 2:269t retrograde amnesia 2:3–4 personality traits 2:405 personhood criteria 1:257 perspective-reversing figures 1:94f, 1:96 perverse sexual assaults 2:238 phantom limb syndrome 2:290 phencyclidine (PCP) classifications and characteristics 2:222, 2:224t phenomenological characteristics 2:225, 2:226 phenobarbitol 2:227f phenomenal consciousness basic concepts 1:158, 2:22 characteristics 1:85, 1:86t, 2:161, 2:390 conscious content 1:285 folk theories 1:256 intentionality 1:427 neurobiological theories 2:95 personhood criteria 1:257 representational theories basic concepts 2:23 criticisms 2:26 enactive theory 2:31 externalism 2:27 internalism 2:29 subjectivity 2:30 self-awareness theories 2:161 serotonin 2:104 subjective properties 1:144, 2:390 phenomenally conscious machines 1:42 phenomenological subtraction and addition altered states of consciousness 2:219–220 information integration theory (IIT) 1:403 pheochromocytoma 2:251–252, 2:262t phobic anxiety disorder 2:253t phonological studies 1:452 phototaxis 2:122 physical hallucinations 2:261 physical symbol system hypothesis 1:37 physics concepts 2:47 physiostigmine 2:222 Piaget, Jean 1:389–402 Pick’s disease 2:262t, 2:263–264 pictorial illusions 2:129, 2:130f
501
502
Subject Index
picture theory 2:451, 2:451f pill-rolling tremor 2:267 Pinocchio illusion 2:290–291 Piraha˜ tribe 1:458, 1:458f Pisa syndrome 2:267 placebo treatment clinical trials 2:202 general discussion 2:214 placebo response effects 2:201–216 central mechanisms conditioning-based responses 2:209 dopamine release 2:212 expectation-based responses 2:210 general discussion 2:209 research controversies 2:211 conditioning-based responses brain mechanisms 2:209 endogenous opioid responses 2:206, 2:210 research controversies 2:211 stimulus response conditions 2:209 expectation-based responses basic concepts 2:210 brain mechanisms 2:211 endogenous opioid responses 2:206 research controversies 2:211 false placebo effects 2:202 mechanisms central mechanisms 2:209 general discussion 2:208 proximal mechanisms 2:212 physiological outcomes endogenous opioid responses 2:206, 2:210, 2:212 general discussion 2:205 neuroimaging research 2:206, 2:212 pain physiology 2:205, 2:212 peripheral physiological processes research 2:208 proximal mechanisms depression 2:213 dopamine release 2:214 general discussion 2:212 Parkinson’s disease 2:214 placebo analgesic effects 2:212 as reporting biases demand characteristics 2:204 Hawthorne effects 2:204 response bias 2:204 self-reporting issues 2:203 research background 2:201 as statistical artifacts natural symptom fluctuation 2:203 participant sampling bias 2:203 regression to the mean 2:203 research background 2:202 spontaneous remission 2:203 research background 2:201 place perception 2:262–263 Place, UT 1:345–346 planchette 1:354 Plato aesthetics 1:1 consciousness studies 1:330 memory studies 1:370–371 mental imagery 2:447 mind body problem 1:340–341 personal identity studies 2:303 Polanyi, Michael 1:450–451 pontine tegmentum 2:363f pop out features binding problem 2:148, 2:148f inattentional blindness 1:56 selective attention 1:65, 1:65f population coding 2:153, 2:153t population question 2:302 porphyria 2:262t positron emission tomography (PET) disorders of consciousness (DOC) 1:152f habit formation processes 1:320 hypnosis research
general discussion 1:361 hypnotic analogue research alien control delusions 1:364–365 clinical delusions 1:363 conversion disorder paralysis 1:363–364 disrupted perception 1:363 functional (hysterical) blindness 1:363 general discussion 1:363 hallucinations 1:364 malingering/faked conditions 1:365 instrumental research 1:363 intrinsic research 1:361 inner speech 1:393 placebo response analysis 2:206, 2:212 religious/spiritual experiences 2:275 REM sleep 2:362, 2:363f skill acquisition and learning 2:14 voluntary action 2:116 postattentive awareness 1:70 postcardiac arrest 2:262t posterior cingulate gyrus 2:363f posterior parietal cortex (PPC) automatic controlled processing research 1:90, 1:90f creativity 1:438 Duplex Vision Theory 2:91 habit formation processes 1:320 motor-skill learning 2:15 visual systems 2:123, 2:123f voluntary action 1:124, 1:124f, 1:125f posterior superior parietal lobe (PSPL) 2:274, 2:277f, 2:279 posteromedial cortex 1:152, 1:152f posthypnotic suggestions 1:351–352 postsynaptic proteins 2:109–110 posttraumatic nightmares 2:345–346, 2:348 posttraumatic stress disorder (PTSD) 2:253t, 2:353 posturing 2:266 poverty of content 2:256t preattentive feature maps 1:63 prefrontal cortex bistable perception 1:98f, 1:103 consciousness development processes 1:221–229 age-related constraints 1:227 Capgras syndrome 1:223, 2:188 general discussion 1:221, 1:227 growth process 1:223 levels of consciousness 1:94, 1:221, 1:222, 1:225f minimal consciousness 1:222, 1:225f phenomenological onset 1:222 preschooler prefrontal patient study 1:225 reflective (recursive) consciousness 1:223, 1:225f self-awareness theories 1:223, 1:224 subjective perspective labels 1:226 dreaming 2:363f, 2:364 fringe consciousness 2:461 growth process 1:223 habit formation processes 1:320 intergroup processing 2:382, 2:384–385 memory functions 1:184, 1:186f mind wandering research 2:67 motor-skill learning 2:15 pain perception 1:246 perceptual awareness 2:84 placebo response analysis 2:206–207, 2:211, 2:213 religious/spiritual experiences 2:275, 2:277f, 2:278, 2:284 selective memory retrieval 1:209, 1:212f, 1:213f theory of mind research 2:401–402 transient hypofrontality theory 2:219 visual systems 2:123 Premature Infant Pain Profile (PIPP) 1:244t premotor cortex 1:124, 1:124f, 1:125f, 2:127 preschooler prefrontal patient study 1:225 pre-Socratic philosophers 1:340 pressure of speech 2:256t presupplementary motor area (pSMA) automatic controlled processing research 1:90, 1:90f voluntary action 1:124, 1:124f, 1:125f, 2:115–116 Prichep, Leslie 1:311 primary consciousness 1:24, 1:284
Subject Index
primary delusions 2:260t primary motor cortex See motor cortex primary repression 2:239 primary visual cortex bistable perception 1:101 blindsight 1:107, 2:127 encoding processes 2:79 implicit perception 2:417 Livingstone Hubel model 2:151 Local Recurrence Theory 2:93 motor-skill learning 2:15 perceptual awareness 2:82, 2:420 REM sleep 2:362 residual islands/tags 1:117 schematic diagram 2:123f visual field defects early human evidence 1:108, 1:109f early nonhuman primate evidence 1:109 recent human evidence 1:112 recent nonhuman primate evidence 1:111 vivid mental imagery 2:449–450 voluntary action 1:124f primates visual recognition studies 1:24 visual system research binding problem 1:64–65 bistable perception 1:101 blindsight 1:26, 2:128 early nonhuman primate evidence 1:109 recent nonhuman primate evidence 1:111 selective attention 1:63 visually-guided behaviors 2:123 priming methods automatic social behavior 1:385 compound remote associate (CRA) problems 1:438 creativity research 1:437 evaluative priming 1:386 goal pursuit 2:427–428, 2:431 habit-formation processes 1:325 implicit memory research 1:371, 2:414 intergroup processing 2:384 memory characteristics 2:263t right-hemisphere priming 1:437 semantic priming 1:437, 2:418 theoretical controversies 1:160–161 unconscious information processing attentional limits 2:171 backward masking 2:170 binocular rivalry suppression 2:169 feedforward feedback connections 2:171 habit-formation processes 1:323 subliminal affective priming 1:233, 2:427–428 Prince, Morton 2:231 principle of the necessity of identity 2:48 Principles of Psychology (James) 1:83, 1:123, 1:330, 2:375, 2:412, 2:460 prior knowledge 2:439 private speech definition 1:389 developmental characteristics 1:392 self-regulation 1:390–391, 1:392 privileged access concept 1:258, 2:189 problem-solving experiences 1:433 procedural implicit memory 1:186, 2:1, 2:15, 2:263t process dissociation procedure 1:377 propofol 1:296 proportional awareness 2:293 propositional attitudes 1:256 propositions 2:20 propranolol 1:444 proprioception 2:292 prosopagnosia 2:258 prospective memory 2:263t proximal decisions 1:268 Prozac 2:223 pseudoakathisia 2:266 pseudohallucinations 2:250t, 2:260
psilocybin classifications and characteristics 2:222, 2:224t color perception 2:227f phenomenological characteristics 2:225 religious/spiritual experiences 2:284 psychedelics characteristics 2:222 classifications and characteristics 2:224t 5-HT receptors 2:104 misrepresentations of experiences 1:12 passage of time 1:194 phenomenological characteristics 2:223 psychiatric diagnoses 2:247 psychic pillow 2:266 psychoactive drugs 2:217–229 classifications and characteristics 2:221 color perception 2:226, 2:227f form constants 2:226 general discussion 2:228 historical background 2:217 passage of time 1:194 phenomenological characteristics 2:223 religious/spiritual experiences 2:284 transient hypofrontality theory 2:219, 2:220f psychodynamic theories of the unconscious 2:231–244 Adler’s theories 2:241 associationism 2:232 competing theories 2:241 consciousness definitions 2:232 general discussion 2:243 Jung’s theories 2:241 research background 2:231 sensory-motor reflex physiology 2:232, 2:233 sexual repression castration anxiety 2:241 Freud’s theories 2:237 instinctual drives 2:239 mental structural theory 2:240 Oedipus complex 2:241 seduction hypothesis 2:238 sexual instinctual drive 2:239 two-stage repression 2:239 unconscious hysterical symptoms Breuer’s theories 2:234 Charcot’s theories 2:233 clinical unconsciousness 2:235 Freud’s theories 2:234 Janet’s theories 2:234 rational therapies Breuer’s therapies 2:236 Delboeuf ’s therapies 2:236 Freud’s therapies 2:237 Janet’s therapies 2:235 magnetic diseases 2:235 research background 2:233 psychological continuity basic concepts 2:306 beginning and ending of life 2:308 fission factor 2:307 multiple occupancy 2:307 personal-ontology viewpoint 2:309 too-many-minds problem 2:310 psychological distance 1:226–227 psychological moment/psychological present 2:331 psychoneuroses 2:237 psychopathology 2:245–271 basic concepts 2:246 cognitive disorders basic concepts 2:261 delirium 2:261, 2:262t dementia 2:261, 2:262t identity perception 2:263 insight 2:265 memory 2:263, 2:263t place perception 2:262–263 self-perception 2:264 time perception 2:262 visuospatial information 2:264
503
504
Subject Index
psychopathology (continued) diagnostic concerns 2:268 disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257 misidentification syndromes 2:257 self-harm risks 2:255 emotional disorders 2:252 general discussion 2:245 mood and affect disorders 2:252, 2:253t movement disorders characteristics 2:265 involuntary movements 2:267 voluntary movements 2:266 perception disorders first rank symptoms 2:260t general discussion 2:258 hallucinations 2:259 top-down/bottom-up cognitive processing 2:259 personality disorders 2:267, 2:269t phenomenological characteristics delirium versus dementia 2:249–250 form versus content 2:251 functional versus organic disorders 2:248–249 general discussion 2:247 levels of explanation 2:251, 2:252t psychosis versus neurosis 2:248 reactive versus instrumental behavior 2:251 terminology 2:250t visual versus auditory hallucinations 2:249 psychiatric diagnoses 2:247 psychopaths 2:269t psychophysics 2:435 psychosemantic theories 2:21 psychoses 2:248, 2:250t, 2:252t pulvinar 2:123f puns 2:256t pupil grating responses (PGRs) 1:114, 1:115f pupillometry implicit processing 1:114, 1:115f visual stimuli testing 1:108 putamen 1:124f Putnam, Hilary 1:347 Pylyshyn, Zenon 2:453
Q Qigong 2:274, 2:283 qualia action control 1:179 consciousness intentionality relationship 1:417, 2:22 displaced perception theories 2:195 higher-order thought theory 1:426 linguistic processes 1:449 qualia space (Q-space) analyses experience properties 1:411 experience shape 1:410 geometrical analyses 1:412 unconscious information processing 1:284, 2:160–161, 2:168 See also phenomenal consciousness qualitative identity 2:304 quantum phenomena 2:464 quasipictorial theory 2:451
R Rabbit syndrome 2:267 racial bias 2:381 radical emergentism 1:343–344 radical plasticity thesis 1:142 raphe nucleus 1:307, 2:104, 2:108–109, 2:277f rational action 1:287 readiness potential study 1:127, 1:128f, 1:269, 1:280, 1:281f, 2:115 reality monitoring 2:6
Reber, Arthur 1:372, 1:379 receptive fields edge enhancement 2:436f, 2:437 visual cortex 1:63–64 recollective experiences amnesia 1:189 autobiographical memory 1:77 functional role 1:191 recovered memories 2:37 recurrent dreams 2:345 recurrent intrusive thoughts 2:258t recurrent processing model 2:156 reduced speech 2:256t reelin expression 2:109–110 Reentrant Dynamic Core Theory access phenomenal consciousness relationship 2:95 attention consciousness relationship 2:96 basic concepts 2:89 neurocognitive panpsychism 2:98 reflective consciousness 1:223, 1:225f, 2:161, 2:177, 2:344 reflex hallucinations 2:259 Reid, Thomas 2:116 relational binding 1:189 religious delusions 2:258t religious/spiritual experiences 2:273–287 altered states of consciousness 2:273 brain activation amygdala 2:277f, 2:280 anterior cingulate cortex (ACC) 2:277f, 2:278 autonomic nervous system (ANS) 2:281, 2:282 beta(b)-endorphins 2:276t, 2:281 electroencephalography (EEG) 2:275–276, 2:283 general discussion 2:277 hippocampus 2:277f, 2:280 hypothalamus 2:277f, 2:280, 2:281 language areas 2:280 limbic system 2:280 posterior superior parietal lobe (PSPL) 2:277f, 2:279 prefrontal cortex 2:277f, 2:278 schematic diagram 2:277f serotonin 2:282 thalamus 2:277f, 2:279 neuroimaging research 2:275 psychological research 2:285 psychotropic drugs 2:284 research challenges 2:286 Yoga meditation 2:284 remembering imaging systems 1:79 remembering, process of 1:77, 1:188–189 remorse 2:255, 2:269t remote associates task (RAT) 1:436–437 REM sleep altered states of consciousness characteristics 1:12 coma 2:367, 2:368f locked-in syndrome (LIS) 2:370 recovered states 2:369 sleep disturbances 2:369 vegetative states 2:369 characteristics 2:359f, 2:360 costly signaling theory (CST) 2:349 dream content analysis 2:345 general discussion 2:370 historical research 2:357 lucid dreaming characteristics 1:14, 1:16 illustrative diagram 1:148f sentinel theory 2:349 renal failure 2:262t repetition blindness 2:417 repetition compulsion 2:240, 2:256t repetitive dreams 2:345 representational content 2:19 representational neglect 2:72 Representation Theory of Mind (RTM) 1:423 repression olfactory hallucinations 2:237 primary repression 2:239
Subject Index
seconday repression 2:239 sexual repression castration anxiety 2:241 Freud’s theories 2:237 instinctual drives 2:239 mental structural theory 2:240 Oedipus complex 2:241 seduction hypothesis 2:238 sexual instinctual drive 2:239 two-stage repression 2:239 response override hypothesis 1:205–219 basic concepts 1:205, 1:206f general discussion 1:217 perception awareness research 1:206 retrieval stopping processes Freudian suppression model 1:214 individual differences 1:214 intrusive memory awareness 1:215 neurobiological theories 1:212 Think/No-Think paradigm (TNT) 1:210, 1:211f, 1:212f, 1:213f thought suppression 1:216 working memory control mechanisms 1:216 selective retrieval inhibition mechanisms 1:209 inhibitory control processes 1:207 neurobiological theories 1:209 retrieval-induced forgetting (RIF) 1:206f, 1:208 retrieval stopping processes 1:210 resting cerebral metabolism 1:151, 1:152f restructuring experiences 1:433 reticular thalamic nucleus (NRT) 1:303, 1:308, 2:277f, 2:279, 2:359 retina 2:123f retinotopic maps 2:155–156 retrieval-induced forgetting (RIF) 1:206f, 1:208 retrograde amnesia 1:12, 1:79, 1:189, 2:3–4, 2:15, 2:262, 2:414 reverse hierarchy theory (RHT) 1:66 reversible figures 1:96, 2:78 Ribot, The´odule 2:117 riddles 1:440 Riddoch syndrome 2:84 right-left disorientation 2:264 Ritalin 2:223 ritual behaviors 2:258t Roback, AA 1:334 robotic systems 1:38, 1:41, 1:257 Rosenthal, David conscious content 1:284 higher-order thought theory 1:165, 1:285, 1:348, 1:426 phenomenal consciousness 2:390 rostral anterior cingulate cortex (rACC) 2:206, 2:211, 2:213 Rubin’s face-vase illusion 1:93, 1:94f, 1:96 rumination 2:431 runner’s high 1:16, 2:219–220 Russell, Bertrand 1:333, 1:422, 2:47, 2:390 Ryle, Gilbert 1:333, 1:344, 2:192
S saccade 2:107 saccade-contingent research techniques 1:50 samadhi state of meditation 1:15–16 same-race memory bias 2:381 Sapir Whorf hypothesis 1:390, 1:456 sarcoidosis 2:262t schizoid personality disorder 2:269t schizophrenia creativity 1:443 dysfunctional self-talk 1:397 memory capacity 2:337 neurochemical mechanisms 2:109 psychopathology alien control delusions 2:260t bodily hallucinations 2:260t disordered thought and speech 2:250t, 2:256t ego boundaries 2:257 first rank symptoms 2:260t
form versus content 2:251 functional versus organic disorders 2:248–249 grandiose delusions 2:258t hallucinations 2:259 levels of explanation 2:252t mood and affect disorders 2:252, 2:253t movement disorders 2:266 passivity 2:258t, 2:260t phenomenological characteristics 2:247 primary delusions 2:260t psychiatric diagnoses 2:247 self-harm risks 2:255 thought possession 2:258t visual versus auditory hallucinations 2:249, 2:259, 2:260t schizotypal personality disorder 2:269t Schooler, Jonathan 2:61 Schopenhauer, Arthur 2:465 Schroeder staircase 1:96 scientific epiphenomenalism 1:274 scientific realism 1:345 scientific revolution era 1:341 scopolamine 2:222, 2:224t, 2:225 scotoma 1:108, 2:82, 2:107 Searle, John 1:161, 1:423, 1:453, 2:116 seconday repression 2:239 sedatives characteristics 2:221 classifications and characteristics 2:224t color perception 2:227f pain treatment 1:247 seduction hypothesis 2:238 seizures 2:262t selective attention 1:61–75 attention awareness relationship attentional blink 1:67, 1:234 awareness of attention 1:71 awareness of awareness 1:70, 1:71f blindsight 1:69, 1:69f change blindness 1:47–59, 1:67, 1:86, 1:233–234, 2:85, 2:96, 2:167 general discussion 1:66 gist situations 1:68 inattentional blindness 1:47–59, 1:67, 1:233–234, 2:85, 2:167, 2:335 neural mechanisms 2:84 neurochemical mechanisms 2:103 otherwise-engaged attention 1:68 postattentive awareness 1:70 basic concepts 1:61 binding problem 1:62, 1:235 creativity 1:442–443 general discussion 1:71 neurochemical mechanisms 2:103 object recognition feature characteristics 1:65, 1:65f feature integration theory 1:63 feedforward models 1:65 reverse hierarchy theory (RHT) 1:66 selective memory retrieval inhibition mechanisms 1:209 inhibitory control processes 1:207 neurobiological theories 1:209 retrieval-induced forgetting (RIF) 1:206f, 1:208 retrieval stopping processes Freudian suppression model 1:214 individual differences 1:214 intrusive memory awareness 1:215 neurobiological theories 1:212 Think/No-Think paradigm (TNT) 1:210, 1:211f, 1:212f, 1:213f thought suppression 1:216 working memory control mechanisms 1:216 selectivity controls 1:55 self-awareness theories 2:187–199 body awareness 2:289–300 affective awareness 2:290 conceptual awareness 2:290 direct awareness 2:293
505
506
Subject Index
self-awareness theories (continued) general discussion 2:289, 2:299 high-level body awareness 2:290 low-level body awareness basic concepts 2:290 body-internal spatial coding 2:291 moment-to-moment representation 2:291 neurological mechanisms 2:292 object-relative spatial coding 2:291 orientational coding 2:292 proprioception 2:292 proportional awareness 2:293 self-specifying information 2:292, 2:295–296 self-world dualism 2:296 semantic awareness 2:290 somatic proprioception 2:295 spatial perceptual frames of reference 2:296 visual perception 2:292 consciousness development processes 1:223, 1:224 dream content analysis 2:343 inner speech 1:398 phenomenal consciousness 2:161 situatedness 2:187 voluntary action 2:118 self-concept 2:313–326 general discussion 2:314, 2:324 implicit associations 1:385 self-concept coherence emergence of self-structure 2:315–316, 2:316f general discussion 2:314 global self-evaluation 2:317 maintenance processes general discussion 2:319 self-concept change 2:321 self-evaluation maintenance 2:319 self-verification 2:321 self-concept certainty 2:318 self-concept differentiation 2:315, 2:316f self-organization 2:314 self-regulation general discussion 2:322 nonoptimal consistency 2:323 nonoptimal flexibility 2:323 threatened egotism 2:323–324 self-consciousness body awareness 2:289–300 affective awareness 2:290 conceptual awareness 2:290 direct awareness 2:293 general discussion 2:289, 2:299 high-level body awareness 2:290 low-level body awareness basic concepts 2:290 body-internal spatial coding 2:291 moment-to-moment representation 2:291 neurological mechanisms 2:292 object-relative spatial coding 2:291 orientational coding 2:292 proprioception 2:292 proportional awareness 2:293 self-specifying information 2:292, 2:295–296 self-world dualism 2:296 semantic awareness 2:290 somatic proprioception 2:295 spatial perceptual frames of reference 2:296 visual perception 2:292 introspection 2:177 self-deception 2:190–191 self-esteem 2:317, 2:431 self-harm risks 2:255 self-knowledge concept 2:194 self-perception cognitive disorders 2:264 habits 1:324 self-regulation hyperactivity 1:398 inner speech 1:394 private speech 1:390–391, 1:392
self-concept coherence general discussion 2:322 nonoptimal consistency 2:323 nonoptimal flexibility 2:323 threatened egotism 2:323–324 self-relevant information 2:314, 2:316f self-specifying information 2:292, 2:295–296 self-statement questionnaire method 1:391 self-talk definition 1:389 developmental characteristics 1:392 dysfunctional self-talk autism 1:398 general discussion 1:396 hyperactivity 1:398 schizophrenia 1:397 self-regulation 1:394 self-world dualism 2:296 semantic dementia 2:264 Semantic disease 2:262t semantic memory 1:188, 1:448–449, 2:2, 2:9, 2:263t semantic priming 1:437, 2:418 semiexplicit perception 1:50 semistabilized images 2:165, 2:166f Sense and Sensibility (Austen) 2:189 sensorimotor imagery theory 2:454 sensorimotor intentions 1:125–126, 1:125f sensory consciousness 1:24 sensory memory 2:327–339 capacity limits to focus of attention 2:335, 2:336f, 2:413 definitions 2:327 general discussion 2:337 theoretical models 2:329, 2:330f time-dependent activated memory auditory stream segregation 2:333f basic concepts 2:331 discrete moment versus continuous moment hypothesis 2:332f partial report procedure 2:333f schematic diagram 2:330f tone memory experiment 2:335f Sensory Modalities Assessment and Rehabilitation Technique 1:151 sensory-motor reflex physiology 2:232, 2:233 sensory-motor theory of consciousness 1:141 sensory neglect 2:72 sentinel theory 2:349 septum 1:185f sequence learning 1:372f, 1:373, 2:14 serial searches 2:148 serotonin anesthetic agents 2:101–102 arousal system 1:305f phenomenal consciousness 2:104 religious/spiritual experiences 2:275, 2:276t, 2:282 sleep regulation 2:359f sevoflurane 1:296, 1:301 sexual repression castration anxiety 2:241 Freud’s theories 2:237 instinctual drives 2:239 mental structural theory 2:240 Oedipus complex 2:241 seduction hypothesis 2:238 sexual instinctual drive 2:239 two-stage repression 2:239 shared delusions 2:258t shock 2:256t short-term body-image 2:291 short-term memory change blindness 1:51, 2:441 insight problem-solving experiences 1:435–436 theories of consciousness 1:136 visual short-term memory (vSTM) 1:48, 1:51, 2:440 shyness, conscious 2:419 Sidis, Boris 2:231 signal detection theory (SDT) blindsight 1:118 cognitive processes 1:143
Subject Index
mind wandering 2:62 placebo treatment 2:204 unconscious conscious perception 2:137, 2:163 sign language 1:395–396, 1:454 silent behavior 2:256t simultanagnosia 2:73, 2:150–151, 2:150f single photon emission computed tomography (SPECT) 2:275 situatedness 2:187 size-contrast illusions 2:129, 2:130f skill acquisition and learning 2:13–18 action pattern organization 2:14 general discussion 2:13, 2:17 integration consensus 1:289, 2:13 neuroimaging research 2:14 perceptual-motor learning 2:13 sequence learning 2:14 Skinner, BF 1:25, 1:333, 1:335, 2:464 sleep 2:357–373 cerebral metabolism 1:152f consciousness altered states of consciousness brain-injured patients 2:367, 2:368f coma 2:367, 2:368f locked-in syndrome (LIS) 2:370 recovered states 2:369 sleep disturbances 2:369 sleep-related experiences 1:12, 2:248 vegetative states 2:369 arousal awareness states 2:365 cortical regulation 1:304, 1:305f definitions 2:358 dreaming 2:341–355 background information 2:341 bad dreams 1:14, 2:345 characteristics 1:13, 2:361 consciousness functions 1:292 cross-cultural studies 2:346 dream content analysis bizarreness 2:344 dream characters 2:343 dream self 2:343 emotional experiences 2:344 general discussion 2:342 misfortunes 2:343–344 reflective consciousness 2:344 sensory modalities 2:342 social interactions 2:343, 2:351 threatening events 2:344 dream themes 2:342 evolutionary functions costly signaling theory (CST) 2:349 general discussion 2:348 psychological problem-solving and creativity 2:350 sentinel theory 2:349 simulation functions 2:350 false awakening 1:10–11 general discussion 2:354, 2:370 historical research 2:357 hypnosis 1:352 illustrative diagram 1:148f lucid dreaming characteristics 1:14, 1:16, 2:364 illustrative diagram 1:148f neural mechanisms 2:219–220 neurobiological correlates 2:362 recurrent dreams 2:345 simulation functions general discussion 2:350 play behaviors 2:351 social simulation hypothesis 2:351 threat simulation theory (TST) 2:352 theoretical perspectives continuity hypothesis (CH) 2:347 evolutionary functions 2:348 psychoanalytic theories 2:346 psychological healing theories 2:348 random activation theories (RATs) 2:347
false awakening 1:10–11 general discussion 2:370 historical research 2:357 hypothalamic regulation 1:307, 1:309f illustrative diagram 1:148f K complexes 2:359, 2:367, 2:368f neuroimaging research 1:307f NREM (non-REM) sleep characteristics 2:359, 2:359f coma 2:367, 2:368f general discussion 2:370 historical research 2:357 information integration theory (IIT) 1:408–409 locked-in syndrome (LIS) 2:370 recovered states 2:369 sleep disturbances 2:369 subjective experiences 1:13 vegetative states 2:369 REM sleep altered states of consciousness characteristics 1:12 coma 2:367, 2:368f locked-in syndrome (LIS) 2:370 recovered states 2:369 sleep disturbances 2:369 vegetative states 2:369 characteristics 2:359f, 2:360 costly signaling theory (CST) 2:349 dream content analysis 2:345 general discussion 2:370 historical research 2:357 lucid dreaming characteristics 1:14, 1:16 illustrative diagram 1:148f sentinel theory 2:349 sleep spindles 2:359, 2:367, 2:368f sleep stages 2:359, 2:359f transcranial magnetic stimulation (TMS) 2:366–367 sleep mentation 1:13 sleep paralysis 1:13 sleepwalking 1:14, 1:86–87 slurred speech 2:256t Smart, JCC 1:345–346 Snodgrass, M 2:142 Snow, John 1:300 social cognitive neuroscience 2:379–388 behavior modification studies 2:386 categorization 2:380 categorization evaluation relationship 2:382 complex situations 2:385 evaluation processes 2:381 general discussion 2:386 intergroup processing 2:379 motivational effects 2:383 neuroimaging research 2:380 same-race memory bias 2:381 self-categorization 2:383 theoretical perspectives 2:380 social phobias 2:251–252, 2:252t social situatedness 2:188 sodium thiopental 1:296 somatic markers 1:172–173 somatic proprioception 2:295 somatoform disorder 2:253t, 2:258t somatognosia 2:264–265 somatoparaphrenia 2:72, 2:261 somatosensory cortex body awareness 2:292 pain perception 1:246 somatosensory cortical potentials 1:151, 2:207 source memory 2:1–2 spatial action 1:171–181 action control action-effect codes 1:178f basic concepts 1:177, 1:178f conscious control action planning connection 1:179, 2:15 conscious will 1:173, 1:286, 1:286f motor programs
507
508
Subject Index
spatial action (continued) computer-programming analogy 1:173 sensorimotor processing 1:174, 2:111 sensory effects 1:172 sensorimotor processing feedforward feedback connections 1:174 multiple-pathway models 1:175, 1:175f, 1:178f single-pathway models 1:175, 1:175f somatic markers 1:172–173 voluntary action 1:171 executive ignorance 1:171 visuospatial information 2:264 spatial neglect 1:87, 2:71 spatial situatedness 2:187 specious present doctrine 1:196, 1:197f speech See language Sperling, George 2:333, 2:440 spiders 2:220f Spinoza, Baruch 1:342 spirituality See religious/spiritual experiences splat-contingent research techniques 1:50 split-brain studies 1:448 spontaneous discoveries 1:431 squid (Loligo) 1:31–32 stalking behaviors 2:255 startle eyeblink tests 2:384 state-change theories 1:95 state consciousness 1:157, 1:284 stereotypes (implicit associations) 1:384 stereotypy 2:266 stimulants 2:221–222, 2:224t stranger violence 2:255 streams of consciousness 1:140, 1:354, 2:58, 2:412 striate cortex bistable perception 1:101 blindsight 1:107, 2:127 encoding processes 2:79 implicit perception 2:417 Livingstone Hubel model 2:151 Local Recurrence Theory 2:93 motor-skill learning 2:15 perceptual awareness 2:82, 2:420 REM sleep 2:362 residual islands/tags 1:117 schematic diagram 2:123f visual field defects early human evidence 1:108, 1:109f early nonhuman primate evidence 1:109 recent human evidence 1:112 recent nonhuman primate evidence 1:111 vivid mental imagery 2:449–450 voluntary action 1:124f Stroop color naming task consciousness development processes 1:227 dual processing theory 1:85 hypnotic suggestibility 1:357, 1:362 linguistic processes 1:452 structural theory 2:240 subconscious 2:231, 2:235 subcortical dementia 2:264 subcortical visual processing 1:101 subdural hematoma 2:262t subitizing 1:457–458 subjective property 2:389–400 Cognitive Revolution 2:389 first-person knowledge 2:394 first-person skepticism 2:397 historical research 2:389 phenomenal consciousness 1:144, 2:390 scientific research 2:392 testimonies of consciousness 2:399 subjectivity 2:30 subliminal affective priming 1:233, 2:427–428 subliminal conscious 2:235 subliminal perception 1:162, 1:324, 2:93, 2:144, 2:159–160, 2:416 substance dualism 2:45
substantia nigra 1:124f, 2:108–109 suggestibility See hypnosis suicide 2:253t, 2:254, 2:255 superconducting quantum interference device (SQUID) sensors 1:129 superego 2:240 superimposed stimuli 1:55 superior colliculus multisensory integration 2:152 visual systems 2:123f superior temporal gyrus 2:152 supervenience 2:26 supervenience thesis 2:48 supplementary motor area (SMA) 1:124, 1:124f, 1:125f, 2:115–116 supramodal regions 2:84 Sustained Attention to Response Task (SART) 2:65–66 Sydenham’s chorea 2:267 symbolic imagery 2:453 symbolic interactionism 2:375 synchronous coconsciousness 1:199 synchrony 1:439, 2:153t, 2:154 synesthesia 2:150 synthetic phenomenology 1:43 syphilis 2:262t systematic processing 2:7 systemic lupus erythematosus 2:262t
T talking cure 2:236 tangentiality 2:256t tardive dyskinesias 2:267 task-unrelated/stimulus-independent thoughts 2:58 telencephalon 1:28 teleosemantic theories 2:22 television technology 2:107 temporal lobe conscious awareness 1:189 hallucinations 2:260–261 implicit memory 1:187 neglect 2:71 sensory memory 2:330 time perception disorders 2:262 temporal parts theory 2:311 temporoparietal junction (TPJ) automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 general discussion 2:401–402 language processing 2:402–403 personality traits 2:405 social cognition 2:404 texture segmentation 1:65 thalamic intralaminar nuclei (ILN) 1:304 thalamocortical system 1:25, 1:28 thalamus arousal system 1:304 central medial thalamus 1:308, 1:309f consciousness development processes 1:222 intralaminar thalamus 1:308, 1:309f mammalian avian comparison research 1:28–29 memory functions 1:185f, 1:186f neglect 2:71 neurochemical mechanisms 2:103, 2:108–109 pain perception 1:246 placebo response analysis 2:206–207, 2:213 religious/spiritual experiences 2:275, 2:277f, 2:279, 2:284 reticular thalamic nucleus (NRT) 1:303, 1:308, 2:277f, 2:279, 2:359 sleep regulation NREM (non-REM) sleep 2:359, 2:359f REM sleep 2:360, 2:363f thalamocortical switch hypothesis 1:306, 1:307f, 1:310 visual systems 1:62–63, 1:62f, 1:101, 2:123, 2:123f voluntary action 1:124, 1:124f, 1:125f Thanatos 2:240 theobromine 2:221–222 theophylline 2:221–222
Subject Index
theories of consciousness artificial intelligence (AI) global workspace theory 1:38 machine consciousness 1:39 model approaches 1:38 cognitive processes 1:135–146 hard problem of consciousness 1:144 higher-order thought theory 1:426 historical research 1:135 illusory theories apparent mental causation theory 1:141 general discussion 1:140 multiple drafts model 1:140 influential theories general discussion 1:137 global workspace theory 1:137 information integration theory (IIT) 1:139 intermediate level theory 1:138 learning process theories general discussion 1:141 higher-order Bayesian decision theory 1:143 radical plasticity thesis 1:142 sensory-motor theory 1:141 precursor theories automatic controlled processing 1:137 central executive system 1:136 modularity 1:137, 1:451 supervisory attentional system 1:136 theory of mind 2:401–410 consciousness development processes 1:223, 1:226 dreaming 2:364 false beliefs event-related potentials (ERPs) 2:407 general discussion 2:401, 2:408 mirror neurons 2:408 neuroimaging research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 language 2:401, 2:402f personality traits 2:405 social cognition 2:404 folk theories 1:252, 1:253f, 2:188–189 general discussion 2:408 mental states 1:256, 2:401 neuroimaging research automatic controlled processing comparisons (dual processing theory) 2:405, 2:406t, 2:407t executive control 2:403 language 2:401, 2:402f mirror neurons 2:408 personality traits 2:405 social cognition 2:404 theory of neuronal group selection (TNGS) 1:291, 2:89 theory of symbolic interactionism 2:375 thiamine 2:2, 2:262t Think/No-Think paradigm (TNT) 1:210, 1:211f, 1:212f, 1:213f think out loud method 1:391 third-person perspective cognition research 2:389 dream content analysis 2:343 psychopathology 2:245–271 basic concepts 2:246 cognitive disorders basic concepts 2:261 delirium 2:261, 2:262t dementia 2:261, 2:262t identity perception 2:263 insight 2:265 memory 2:263, 2:263t place perception 2:262–263 self-perception 2:264 time perception 2:262 visuospatial information 2:264 diagnostic concerns 2:268 disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257
misidentification syndromes 2:257 self-harm risks 2:255 emotional disorders 2:252 general discussion 2:245 mood and affect disorders 2:252, 2:253t movement disorders characteristics 2:265 involuntary movements 2:267 voluntary movements 2:266 perception disorders first rank symptoms 2:260t general discussion 2:258 hallucinations 2:259 top-down/bottom-up cognitive processing 2:259 personality disorders 2:267, 2:269t phenomenological characteristics delirium versus dementia 2:249–250 form versus content 2:251 functional versus organic disorders 2:248–249 general discussion 2:247 levels of explanation 2:251, 2:252t psychosis versus neurosis 2:248 reactive versus instrumental behavior 2:251 terminology 2:250t visual versus auditory hallucinations 2:249 psychiatric diagnoses 2:247 Thorndike, Edward 1:433 thought disordered thought and speech characteristics 2:255, 2:256t, 2:258t ego boundaries 2:257 misidentification syndromes 2:257 self-harm risks 2:255 thought blocks 2:256t thought broadcasts 2:250t thought language connection 1:389, 1:456 thought listing method 1:391 thought possession 2:258t thought suppression 1:216, 2:60, 2:67 threatened egotism 2:323–324 threat simulation theory (TST) 2:352 thyroid disease 2:262t thyrotoxicosis 2:251–252 Tibetan Buddhism 2:274, 2:283, 2:284 tics 2:266–267 time 1:193–204 anesthetic effects 1:300 A-series B-series theory 1:195 cognitive disorders 2:262 passage of time 1:193, 1:300 psychological characteristics 1:195 specious present doctrine 1:196, 1:197f time consciousness models act content model 1:199 color phi phenomenon 1:198 diachronic coconsciousness 1:199 flash-lag illusion 1:198 general discussion 1:198 Husserl’s phenomenological model 1:200, 1:201f hybrid model 1:202 overlap model 1:199 protention retention theory 1:200 synchronous coconsciousness 1:199 time-dependent activated memory auditory stream segregation 2:333f basic concepts 2:331 discrete moment versus continuous moment hypothesis 2:332f partial report procedure 2:333f schematic diagram 2:330f tone memory experiment 2:335f tip-of-the-tongue state 1:448–449, 2:34, 2:461 Titchener, Edward 1:332, 2:390, 2:448 token-identity theory 2:45 Tononi, Giulio consciousness concepts 1:157 dynamic core hypothesis (DCH) 1:290 global workspace theory 1:158–159
509
510
Subject Index
Tononi, Giulio (continued) information integration theory (IIT) 1:139, 1:311, 1:403–416 integration consensus 1:288 Reentrant Dynamic Core Theory 2:89 too-many-minds problem 2:310 top-down conscious processing automatic controlled processing comparisons (dual processing theory) 1:89–90, 2:96 top-down/bottom-up cognitive processing 2:259 topographic theory 2:234–235, 2:241 torticollis 2:267 Tourette’s syndrome 2:266–267 trances 1:353 tranquilizers 2:221 transcendental meditation 2:220–221, 2:274, 2:283 transcranial magnetic stimulation (TMS) action control 1:261 perceptual awareness 2:82 unconscious information processing 2:167, 2:170 wakefulness sleep comparison studies 2:366–367 transference 2:240 transient hypofrontality theory 2:219, 2:220f transient ischemic attack (TIA) 2:262t transitive consciousness 1:157, 1:160, 1:284 Transitivity Principle (TP) 1:160 transsaccadic memory 1:48 transverse intentionality 1:202 traumatic brain injuries 2:262t traumatic childhood experiences 2:238 Treisman, Anne 2:155 tremors 2:267 tuberculosis 2:262t tuberomammilary nucleus 1:305f, 1:307 tumors 2:262t tune-outs 2:36 Turing test 1:41, 1:43 Twin Earth paradigm 1:347, 2:29 two-string problem 1:434f, 1:436–437, 1:444f Tye, Michael 2:195 type-identity theory 2:47–48
U unconscious cognition 2:411–421 affective states 2:418 attentional control 2:413, 2:419 capacity limits 2:413 conscious shyness 2:419 historical research 2:412 implicit learning 2:415 implicit memory 2:414 implicit perception 2:416 implicit thought 2:418 neural/psychological correlates 2:420 unconscious conscious perception 2:135–146 action control 1:171–181 action-effect codes 1:178f afferent versus efferent action 2:113 basic concepts 1:177, 1:178f binding problem 2:152 conscious awareness 2:14, 2:15, 2:111 conscious control action planning connection 1:179, 2:15 conscious will 1:173, 1:286, 1:286f, 2:111, 2:116 folk theory of mind concept 1:253–254, 1:253f intentionality 1:253–254, 1:253f, 2:115 motor programs computer-programming analogy 1:173 sensorimotor processing 1:174, 2:111 sensory effects 1:172 pathological disorders 2:117 self-awareness theories 2:118 sensorimotor processing feedforward feedback connections 1:174 multiple-pathway models 1:175, 1:175f, 1:178f single-pathway models 1:175, 1:175f somatic markers 1:172–173 voluntary action 1:171
executive ignorance 1:171 general discussion 2:135, 2:145 goal pursuit 2:423–433 automatic thoughts and actions 2:426 decision-making tasks 2:428 general discussion 2:423, 2:427, 2:432 goal contagion 2:427 goal establishment 2:424 goal intentions 2:425 goal-related words 2:430 goal striving 2:427 habitual behaviors 2:426 implementation intentions 2:425 monitoring processes 2:429 planning and implementation stage 2:425 revision stage 2:430 subliminal affective priming 2:427–428 historical research 2:412 linguistic processes 1:453 Local Recurrence Theory 2:93 model approaches null hypothesis model 2:138, 2:139f objective threshold/nonmonotonic approach 2:139f, 2:142 objective threshold/rapid decay approach 2:139f, 2:141 subjective threshold approach 2:139, 2:139f model validity 2:138 neural/psychological correlates 2:420 objective threshold approach 2:137, 2:139f, 2:141 subjective threshold approach 2:136, 2:139, 2:139f unconscious information processing 2:159–173 attention-related factors 2:171 conscious level conscious content distinction 1:284 decision-making tasks 2:428 feedforward feedback connections 2:171 general discussion 2:171 habit-formation processes 1:323 implicit versus explicit learning 1:369, 1:378 invisible stimuli methods attentional limits 2:167 binocluar rivalry 1:94f, 1:97, 1:99, 2:92–93, 2:165, 2:166f flash suppression 2:166 general discussion 2:165 motion-induced blindness 2:166 semistabilized images 2:165, 2:166f transcranial magnetic stimulation (TMS) 2:82, 2:167 visual crowding 2:166f, 2:167 visual (two-transient) masking 2:77, 2:166f, 2:167 methodological issues criterion content 2:162 exhaustive exclusion 2:164 general discussion 2:162 indirect measurement strategies 2:164 invisible stimuli 2:75, 2:165, 2:166f normal versus neurologically impaired observers 2:165 signal detection theory (SDT) 2:163 threshold measurements 2:163 nineteenth-century studies 1:331 postretinal processing attentional limits 2:171 backward masking 2:170 binocular rivalry 2:169 blindsight 2:168 continuous flash suppression 2:170 feature inheritance 2:170, 2:170f form blindness 2:169 general discussion 2:168 motion-induced blindness 2:170 neurological patients 2:168 normal observers 2:169 transcranial magnetic stimulation (TMS) 2:170 research challenges 1:375 research concerns access consciousness 2:161 cognitive contents 2:160–161 definitions 2:160 narrative consciousness 2:161
Subject Index
phenomenal consciousness 2:161 reflective consciousness 2:161 results and analysis general discussion 2:168 postretinal processing 2:168 subliminal affective priming 1:233, 2:427–428 unresolved issues attention consciousness relationship 2:96, 2:144 brain activation 2:144 direct indirect task relationship 2:143 implicit perception 2:144 masking techniques 2:144 subliminal perception 2:144 visual processing 2:127 unconscious emotion 1:236 unconscious idea 2:232 unconscious intelligence 2:462 unconsciousness, definition of 2:232 unilateral neglect 2:71, 2:92–93 unwanted thoughts inner speech 1:397 metaconsciousness/meta-awareness 2:36 neuroimaging research 2:67 retrieval stopping processes 1:210, 1:215, 1:216 See also dysfunctional self-talk uremia 2:262t utilization behavior 1:273–274
V vascular diseases 2:262t vegetative states 1:147–156 cerebral metabolism 1:152f clinical definitions 1:149 consciousness states 2:365–366 diagnostic criteria 1:150, 1:244 external stimulation 1:153f historical research 1:148 illustrative diagram 1:148f pain assessment ethical considerations 1:247 general discussion 1:243, 1:248 pain perception 1:246, 1:246f sleep states 2:369 treatment strategies 1:153 ventral lateral preoptic nucleus 1:305f, 1:307 ventrolateral prefrontal cortex (VLPC) 1:209 verbal auditory hallucinations 1:397 verbal overshadowing 1:399–400, 2:38 verbigeration 2:256t verificationism 1:344 vertebrates consciousness studies 1:27 visual systems 2:122–123 See also primates viral infections 2:262t Virola tree 2:224t virtual reality processes 2:106 visual agnosia 2:124–125, 2:124f, 2:125f Visual Analogical Scale (VAS) 1:243 visual cortex acetylcholine 2:106 basic concepts 1:62–63, 1:62f bistable perception 1:97, 1:98f, 1:101 blindsight 1:107, 2:127 Duplex Vision Theory 2:91 encoding processes 2:79 Livingstone Hubel model 2:151 microconsciousness theory 2:94 perceptual awareness 2:82 schematic diagram 2:123f serotonin 2:104 subcortical visual processing 1:101 visual pathways 1:119, 1:119f voluntary action 1:124f visual crowding 2:166f, 2:167
visual field defects basic concepts 1:107 implicit processing 1:114 research background early human evidence 1:108, 1:109f early nonhuman primate evidence 1:109 follow-up research 1:113 recent human evidence 1:112, 1:112f, 1:113f recent nonhuman primate evidence 1:111 visual hallucinations 2:249, 2:260 visual masking 2:77, 2:92–93, 2:166f, 2:167 visual neglect 2:72 visual short-term memory (vSTM) 1:48, 1:51, 2:440 visual systems 2:121–134 binding problem Balint Holmes syndrome 2:150, 2:150f Livingstone Hubel model 2:151 neural encoding mechanisms 2:83–84 psychophysical research 2:148, 2:148f selective attention 1:62, 1:235 spatiotemporal distinctions 2:149 synesthesia 2:150 theoretical solutions convergence coding 2:153, 2:153t feature integration theory 2:155, 2:155f general discussion 2:152 population coding 2:153, 2:153t recurrent processing model 2:156 synchrony model 2:153t, 2:154 bistable perception 1:93–106 basic concepts 1:93, 1:94f first-person skepticism 2:397 general discussion 1:104 neural correlates binocular rivalry 1:100, 1:101 cortical/subcortical processing 1:101 extrastriate cortex 1:101 general discussion 1:99 measurement techniques 1:99 motion perception 1:101, 1:102f neural synchronization 1:104 parietal and prefrontal cortices 1:98f, 1:103 reversal-related activations 1:101, 1:102f theoretical perspectives 1:95 theoretical models behavioral evidence 1:97 binocular rivalry 1:99 high-level theories 1:98, 1:98f low-level theories 1:97, 1:98f theoretical perspectives anatomical location theories 1:95 consciousness correlates 1:94 neural correlates 1:95 state-change theories 1:95 visual stimuli ambiguous motion 1:94f, 1:96 binocular rivalry 1:97 gamma distribution 1:94f, 1:96 reversible figures 1:96 blindsight 1:107–122 access consciousness 2:161 affective blindsight 1:233 attention awareness relationship 1:69, 1:69f basic concepts 1:107, 2:127 conscious awareness 1:119, 2:390 implicit perception 2:417 implicit processing 1:114, 1:115f occurrences 1:115 pupillometry 1:114, 1:115f rehabilitation strategies 1:120 research background early human evidence 1:108, 1:109f early nonhuman primate evidence 1:109 follow-up research 1:113 recent human evidence 1:112, 1:112f, 1:113f recent nonhuman primate evidence 1:111 research implications
511
512
Subject Index
visual systems (continued) controversial issues 1:116, 1:161 decision criteria/response bias 1:117 experimental approaches 1:116 normal, degraded vision comparisons 1:118 residual islands/tags 1:117 sensorimotor pathways 1:176 stimulus parameters 1:116 subjective threshold approach 2:141 unconscious information processing 1:284, 2:168 visual pathways 1:119, 1:119f voluntary action 1:287 body awareness 2:292 dorsal ventral streams basic concepts 2:123 blindsight 2:127 general discussion 2:132 grasping abilities 2:125, 2:125f optic ataxia 2:125f, 2:126 perception action frames of reference computational requirements 2:126 conscious perception 2:127 dissociation theories 2:129 grasping abilities 2:129, 2:130f hollow face illusions 2:131–132, 2:131f neuroimaging research 2:129 pictorial illusions 2:129, 2:130f time-control studies 2:132 visuomotor responses 2:129, 2:131f schematic diagram 2:123f visual agnosia 2:124–125, 2:124f, 2:125f visual processing 2:127 Duplex Vision Theory 2:91 evolutionary origins 2:122 functional (hysterical) blindness 1:363, 2:417 general discussion 2:122 heuristic processes 2:438, 2:439f mental imagery 2:445–457 background information 2:446 definitions 2:446 general discussion 2:456 historical research 2:447 subjective individual differences 2:449 theoretical perspectives description (mentalese) theory 2:453 enactive theory 2:454 general discussion 2:450 picture theory 2:451, 2:451f Muller Lyer illusion 1:451, 1:451f object recognition feature characteristics 1:65, 1:65f feature integration theory 1:63 feedforward models 1:65 reverse hierarchy theory (RHT) 1:66 parafoveal vision 2:417 perceptual moment 2:331 receptive fields 2:436f, 2:437 schematic diagram 1:62f self-specifying information 2:292, 2:295–296 visual experience 2:435–443 acetylcholine 2:107 brain processing paradox 2:441, 2:442f categorical perception 1:454 change blindness 2:441 general discussion 2:435, 2:442 neural mechanisms 2:75–86 ambiguous motion 1:94f, 1:96 attention awareness relationship 2:84 binocular rivalry 2:78, 2:128–129 bistable perception 1:95 conscious content 2:79 cortical regions 2:82 discrimination tasks 2:76–77 encoding processes 2:79 general discussion 2:75, 2:85 perceptual threshold 2:75 reversible figures 1:96, 2:78 stimuli processing 2:75
visual masking 2:77, 2:92–93, 2:166f, 2:167 visual processing 2:128–129 rich experience impoverished input paradox color perception 2:438, 2:438f edge alignment 2:437, 2:437f edge enhancement 2:436, 2:436f edge-filling abilities 2:437, 2:438f face recognition 2:439–440 feature exaggeration 2:439 gravity and light constraints 2:438, 2:439f Hermann grid 2:436–437, 2:436f heuristic processes 2:438, 2:439f prior knowledge 2:439 psychophysical research 2:435 rich experience limited report paradox 2:440 serotonin 2:104 subjective experiences 2:390 virtual reality processes 2:106 visual hemineglect 2:85 visual masking 2:77 visual pathways 1:119, 1:119f visual short-term memory (vSTM) 1:48, 1:51, 2:440 visuospatial information 2:264 visuomotor responses blindsight 2:127 change blindness 1:50 dorsal ventral streams 2:123, 2:129, 2:131f evolutionary development theories 2:122 inattentional blindness 1:54 pictorial illusions 2:129 time-control studies 2:132 visuospatial information 2:264 vitalism 1:344–345 vitamin B1 2:2 vitamin B12 2:262t Vividness of Visual Imagery Questionnaire (VVIQ) 2:449 voice onset time (VOT) 1:454 volatile anesthetic agents 1:296 voluntary action 1:171–181 action control action-effect codes 1:178f afferent versus efferent action 2:113 basic concepts 1:177, 1:178f binding problem 2:152 characteristics 1:171 conscious awareness 2:15, 2:111 conscious control action planning connection 1:179, 2:15 conscious will 1:173, 1:286, 1:286f, 2:111, 2:116 folk theory of mind concept 1:253–254, 1:253f historical research 2:111 intentionality 1:253–254, 1:253f, 2:115 motor programs computer-programming analogy 1:173 sensorimotor processing 1:174, 2:111 sensory effects 1:172 pathological disorders 2:117 self-awareness theories 2:118 sensorimotor processing feedforward feedback connections 1:174 multiple-pathway models 1:175, 1:175f, 1:178f single-pathway models 1:175, 1:175f sequence learning 2:14 somatic markers 1:172–173 causal theories 2:111 characteristics 1:171 conscious will action initiation 1:132 brain activation 1:123 dual causation pathways 1:286f historical research 2:111 legal implications 1:132 long-term intentions 1:132 readiness potential study 1:173, 1:280, 2:116 theoretical perspectives 1:286 executive ignorance 1:171 movement disorders 2:266 neural mechanisms 2:111–120
Subject Index
afferent versus efferent action 2:113 conscious awareness 2:15, 2:111 historical research 2:111 intentionality 2:115 pathological disorders 2:117 self-awareness theories 2:118 neuroscience research 1:123–133 action initiation 1:132 basic concepts action-initiation steps 1:125f, 1:126 brain area locations 1:124, 1:124f schematic diagram 1:125f sensorimotor intentions 1:125–126, 1:125f willed intentions 1:125, 1:125f consciousness general discussion 1:126 movement control 1:126 movement initiation 1:127 readiness potential study 1:127, 1:128f conscious will 1:123 legal implications 1:132 long-term intentions 1:132 neuroimaging techniques 1:129 Wegner’s model 1:130, 1:131f philosophical theories 1:131 Vygotsky, Lev private speech 1:389, 1:392 thought language connection 1:390
W wakefulness altered states of consciousness 1:12 anesthetic effects 1:302 consciousness functions 1:292 definition 1:147, 1:148f historical research 2:357 transcranial magnetic stimulation (TMS) 2:366–367 Wallace, Alfred Russell 1:25 Wallas, Graham 1:432 watercolor illusion 2:438, 2:438f Watson, John 1:25, 1:331, 1:333, 2:448–449 waxy flexibility 2:266 Wearing, Clive 1:79 Wegner, Daniel apparent mental causation theory 1:141 free will theory 1:123, 1:179, 1:274, 1:286, 1:286f illusory experiences 1:141, 1:286, 1:360 ironic processing theory 1:216, 2:60 unwanted thoughts temporal dissociations 2:36 thought suppression studies 1:216 voluntary action studies 1:130, 1:131f Wernicke Korsadov syndrome 2:262t Wernicke’s area 1:399 Wessex Head Injury Matrix 1:151
Western Neuro-Sensory Stimulation Profile 1:151 Wheatley, Thalia 1:274 whistle speech 1:447 white bear paradigm 1:216, 2:60, 2:67 white matter 2:71 Whyte, Lancelot 2:232 wide content 2:27 willful causation See conscious will Wilson’s disease 2:262t Winkler, Istva´n 2:334–335 Wisconsin Card Sorting Test 2:6 Wittgenstein, Ludwig 1:333, 1:344, 2:397–398, 2:448–449 word salad 2:256t working memory altered states of consciousness 2:219 attentional control 2:62f, 2:64 capacity limits 2:335, 2:336f, 2:413 characteristics 2:263t consciousness development processes 1:223 control mechanisms 1:216 general discussion 2:337 habit formation processes 1:320 information flow 2:58, 2:59f inner speech 1:396 insight problem-solving experiences 1:435–436, 2:68 theories of consciousness 1:136 World Knot theory 2:465 writer’s cramp 2:267 Wundt, Wilhelm introspectionism 1:332, 2:390 introspective awareness 2:448 mind body problem 2:465 visual experience studies 2:436 voluntary action studies 2:111
X xenon 1:296, 1:307f
Y Yoga meditation 2:283, 2:284 Yoga Nidra guided meditation 2:284 yohimbe 2:221–222
Z Zajonc, Robert 2:376 Zeki, Semir 2:94 Zen meditation 2:283 zombies 1:4, 1:86–87, 1:280 zone of proximal development 1:390 zone-outs 2:36
513
