Curious Emotions: Roots Of Consciousness And Personality In Motivated Action (Advances in Consciousness Research)

Curious Emotions

Advances in Consciousness Research Advances in Consciousness Research provides a forum for scholars from different scientific disciplines and fields of knowledge who study consciousness in its multifaceted aspects. Thus the Series will include (but not be limited to) the various areas of cognitive science, including cognitive psychology, linguistics, brain science and philosophy. The orientation of the Series is toward developing new interdisciplinary and integrative approaches for the investigation, description and theory of consciousness, as well as the practical consequences of this research for the individual and society. Series A: Theory and Method. Contributions to the development of theory and method in the study of consciousness.

Editor Maxim I. Stamenov Bulgarian Academy of Sciences

Editorial Board David Chalmers

Steven Macknik

Australian National University

Barrow Neurological Institute

Gordon G. Globus

George Mandler

University of California at Irvine

University of California at San Diego

Ray Jackendoff

Susana Martinez-Conde

Brandeis University

Barrow Neurological Institute

Christof Koch

John R. Searle

California Institute of Technology

University of California at Berkeley

Stephen Kosslyn

Petra Stoerig

Harvard University

Universität Düsseldorf

Earl Mac Cormac Duke University

Volume 61 Curious Emotions: Roots of consciousness and personality in motivated action by Ralph D. Ellis

Curious Emotions Roots of consciousness and personality in motivated action

Ralph D. Ellis Clark Atlanta University

John Benjamins Publishing Company Amsterdam
/
Philadelphia

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data Ralph D. Ellis Curious Emotions : Roots of consciousness and personality in motivated action / Ralph D. Ellis. p. cm. (Advances in Consciousness Research, issn 1381–589X ; v. 61) Includes bibliographical references and index. 1. Consciousness. 2. Emotions. 3. Motivation (Psychology) 4. Selforganizing systems. I. Title. II. Series. BF311.E483 2005 152.4--dc22 isbn 90 272 5197 5 (Eur.) / 1 58811 628 X (US) (Hb; alk. paper)

2004062771

© 2005 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

Table of contents

Introduction 1. The enactive approach to affective intentionality 1 2. Some preliminary predictions of enactivism 9 3. The “curious” emotions 14 4. Conceptualizing action versus reaction 18 5. Plan of the book 21 Chapter 1 Preconscious emotional intentionality 1. Motivation, conscious emotion, and unconscious emotion 27 2. The murkiness of emotional intentionality 31 3. Aims, objects, triggers, and symbolization-vehicles 33 4. The roles of sensation, interoception, and sensorimotor action imagery 42 Chapter 2 Motivated attention in action: How emotion creates conscious intentionality 1. Linear versus dynamical causal sequences in the brain 52 2. Conflicting theories with conflicting empirical predictions 57 3. The P300 ERP as an operational definition of perceptual consciousness 63 4. How the Mack and Rock data relate to the two types of hypotheses 63 5. The paradox of early and late selection 69 6. Attention and conscious processing 72 7. Further implications for the problems of attention and consciousness 73

1

25

47



Curious Emotions

Chapter 3 Non-consummatory motivations: Extropy and “life wish” in the self-organization of emotion 1. Intertheoretic reduction and consummatory-drive reductionism 82 2. The notion of “extropy”: A non-reductive force? 89 3. The humanistic notion of “life wish” 95 4. A possible synthesis 99 Chapter 4 Homeostasis, extropy, and boundary needs as grounding specific emotions 1. Physiological evidence for non-consummatory motivation 104 2. Novelty, constraints to freedom, and the action-consciousness connection 115 3. The importance of extropy needs in higher mammals 124 4. Existential requirements for an adequate dynamical theory of emotion 125 5. Toward an integrated physiological and phenomenological account 128 Chapter 5 Varieties of extended self and personality 1. How emotion grounds the various senses of self 2. Why not an illusory-choice model? 142 3. The embodied self and the personality 150 4. How can there be knowledge of the self? 158

79

103

131 138

Chapter 6 Learning about emotions through the arts 1. An enactive dance form for the eye 171 2. Why does art move, and not just entertain? 176 3. Love and other non-consummatory motivations 181

167

Table of contents 

Chapter 7 Dynamical systems and emotional agency: A closer look 1. The causal power of dynamical systems 193 2. How can top-down systems avoid violating causal closure? 200 3. The emotional brain as an enactive system 207 4. Objections and responses 209

189

Conclusion

215

References

223

Index

233

Introduction

.

The enactive approach to affective intentionality

Enactive approaches to intentionality and consciousness propose that such mental processes as feeling, representation, and consciousness, including perceptual consciousness, can result only from a self-organizational system that in an important sense acts upon rather than only reacting to its environment, and one that appropriates, organizes, and replaces its own micro-constituents on an as-needed basis rather than being only a causal epiphenomenon that is built up from the interactions of the micro-constituents. “Enaction” in this context is meant to differ from mere “activity” by including some sort of primitive causal agency, and to define it requires delving into the distinction between living and nonliving systems in terms of their dynamical self-organization, which gives some systems a power to use their environment and their own constituents to maintain stable structure and function across exchanges of energy and materials. The term “enactive” was first used in this way by Varela et al. (1991), while simultaneously Newton was developing an “action theory of understanding” (Newton 1982, 1989a, 1996) and a “sensorimotor theory of cognition” (Newton 1993); during the same period, Ellis (1986, 1990, 1995) was proposing that consciousness depends on a self-organizing motivational process; Kauffman (1993), Kelso (1995) and others discussed below were developing the theory of self-organization along scientific lines; and Thelen and Smith (1996) applied this theory in developmental psychology. Meanwhile, Panksepp’s animal emotion studies (for example, Panksepp et al. 1984; Panksepp 1998, 2003) were revealing more and more that a number of relatively independent, unconditioned emotional brain systems are fueled by action-initiating circuits deep in the subcortex; and Ito (1993) and Schmahmann and his associates (for example, see Schmahmann 1997) were finding that the cerebellum, traditionally associated with coordinating action routines, also plays a vital role in emotional and higher cognitive processes (see also Courtemanche et al. 2002; Gottwald et al. 2003).



Curious Emotions

It is a short step from these observations that an enactive approach should offer a different understanding of the affective dimension from those that construe emotions and feelings as reactions to perceptual inputs combined with the interoception of internal body states. Damasio (2003) characterizes interoception as just another form of perceptual input originating from the viscera of the body. By contrast to this interoception-plus-associated-perceptual-imagery approach, the enactivist account holds that emotional feelings resemble other conscious states in that they are not constituted by a receiving and processing of inputs, but rather by initiating action patterns relative to actual or possible environmental affordances. Consciousness, for example, depends more on efferent than afferent enervation. By “efferent” in this context, I mean nervous impulses that travel away from the central nervous system’s most primary action-initiating centers – beginning with the feedback loop including the periaqueductal gray (PAG), thalamus and hypothalamus (Panksepp 2000a), from whence the developing action commands then spread through the cerebellum and its loops with the thalamus and hypothalamus (Haines et al. 1997), and eventually lead to a real or imaginary bodily action by activating the motor, premotor, and supplementary motor areas, which then (when disinhibited by the orbitofrontal and anterior cingulate areas – see Faw 2003) relay the efferent commands to the body’s extremities. Afferent signals, by contrast, refer in this context to those signals that travel from the peripheral parts of the nervous system inward to the more central areas. It is now widely acknowledged that consciousness and genuine intentionality can never result from the afferent processing of perceptual inputs alone (Aurell 1984, 1989; Faw 2000; Mack & Rock 1998; Panksepp 1998; Posner & Rothbart 1992, 1998, 2000; Nikolaev et al. 2001; and others to be cited below). Indeed, I shall argue in agreement with Newton (1996) that we understand the world by anticipating how we could act upon it relative to organismic purposes. When the action affordances of environmental situations are anticipated, those anticipated affordances create the action imagery that grounds conscious understanding of the environment. This might also be called a “pragmatic” view, in the technical philosophical sense, but its affective implications are also paradoxically less “pragmatic” in the everyday sense, because it acknowledges that the system is motivated to maintain its patterns of enaction for their own sake, and not merely for instrumental reasons. This view is also consistent with Merleau-Ponty’s (1941, 1942) phenomenological account of the roles of the “body schema” and “body image” in consciousness. The body image is constituted by interoceptive and proprioceptive afferent information of the kind stressed by Damasio (2003). The

Introduction

brain forms this body image out of information received from the body. The body schema, by contrast, is based on sensorimotor action imagery of the kind recently studied by Jeannerod (1997). This imagery results from sending information rather than merely from receiving it. Merleau-Ponty was well aware that the “phantom limb” experience of amputees results primarily from the efferent action commands still being sent to the nonexistent limb. If intentionality is rooted in the action imagery of the body schema – enacted by efferent rather than afferent processes and originating more deeply in the midbrain and brain-stem – then the intentionality of emotion, its “aboutness,” is not exhausted by noting correlations between afferent interoceptions and their accompanying perceptual inputs. That would lead to construing the emotional organism as more reactive and less active than it actually is, and would lead to concretely different specific predictions about the affective life of animals. It also leads to a different phenomenological account of the meanings of feelings, and thus different decisions about how to act on them. Not only exploration, curiosity, play, and social bonding, I shall argue, but also fear and anger are better understood if organisms are endogenously motivated to act in certain ways and then use environmental conditions as action affordances, rather than behaving in those ways because of re-action to afferent stimulation from the environment combined with afferent interoception from their own viscera. To be sure, perception and interoception are important in the feeling of emotions, and like Damasio, Merleau-Ponty emphasizes them; but they cannot be regarded as the whole story. Major advances in the understanding of emotion have taken place in the past few years. But with a few notable exceptions (for example, Gendlin 1992a, b; Panksepp 1998), these advances have focused primarily on one half of the action-reaction system of the living organism. They have interpreted emotional content and feelings as resulting essentially from an interoceptive mapping of the body’s functions (Ben Ze’ev 2000, 2002; Damasio 1994, 1999, 2003; Griffiths 1997; LeDoux 1996; Rolls 1999). But interoception is essentially afferent, like the receiving of perceptual input through vision or audition. So what is left out of the interoception-plus-concurrent-imagery picture is a full enough basis for understanding the intentionality of feelings – what they are about. We understand that feelings are partly about our bodies and how they are faring; but we also need an account of how the feelings are about the world, about the ways we can or cannot act toward it, and about why we want to act in those ways. And it is here that we have fallen back on speculation and ad hoc hypothesis to fill in the gaps of our understanding. It is frequently hypothesized that





Curious Emotions

natural selection somehow made it “just so” that we have these particular emotional reactions, and then that classical conditioning causes us to learn new environmental associations to the old emotions. I agree that natural selection plays a role in the innate response patterns; but so does the nature of selforganization. And while classical conditioning plays some role in leading us to have the eventual emotional responses that adults have, there are more sophisticated mechanisms at work than mere classical or even instrumental conditioning – mechanisms relating to the nature of self-organizing systems. The action affordances of different environments enable different feeling processes partly because of endogenous aspects of organismic structure and function. We should not expect such complexity to be accounted for by simple stimulus generalization based on learning. An infinite number of combinations of environmental conditions can enable a given holistic self-organizational “basin of attraction,” or pattern into which the system is inclined to settle; yet not just any environmental conditions will do, and there are also an infinite number of combinations that can thwart a given holistic aim. Correlatively, careful phenomenological reflection can suggest that the “aboutness” of feelings is often considerably more complex than it initially appears to be, and that nuanced feelings often defy stock categorization. Kauffman (1993) points out that natural selection explanations become much more adequate when supplemented by an understanding as to why certain patterns of self-organization are better designed to maintain themselves across replacements of their components, and more robust with regard to the types of components and environmental conditions they are able to use. If natural selection explains the techniques of the famous knuckle ball pitcher Hoyt Wilhelm, by showing that he could not have survived in the Major Leagues without an effective knuckle ball, then self-organizational theory addresses in the present tense the actual mechanics of why the “knuckle balls” of our living systems should work in the kinds of patterns they do. The first three chapters of this book will explore this concept of selforganization from the standpoint of distinguishing living and genuinely active systems from non-living and merely reactive systems, showing that only the former are capable of emotional feelings, and that fully intentional emotions, whether conscious or not, actually ground and shape all other conscious states. One of the important implications of Kauffman’s analysis of self-organization for affective processes is that self-organization motivates not only homeostatic drive reduction, but also what I call “extropy” – the maintenance of a suitably complex and higher-energy pattern of overall activity for the organism. In my view, not only are many of the endogenous emotional systems described

Introduction

by Panksepp (1998) driven by extropy concerns – play, seeking/curiosity, nurturance/bonding, and even some aspects of sexual lust – but also Panksepp’s discussions of the roles of the periaqueductal gray (PAG) and generally of the upper brainstem in activating the system lead to an enactivist account of the intentionality of affect: the intentionality of feelings is better understood in terms of the actions they aim at, in relation to multiply-realizable environmental affordances, than in terms of the perceptual objects that are thought to cause or trigger the feelings. Again, interoception and perception combined give us only afferent information, when what is needed is efferent information – information about the complex patterns of actions that our organisms are trying to execute, using the environment in a fluid and multiply-realizable way. In fact, I shall argue that to view emotions and feelings as being caused by concurrent or even remembered or imagined external events is to fall into a serious misunderstanding – essentially, the oversimplifying distortion of the meaning of subjective events that Husserl (1913) calls the “natural attitude.” I.e., we try to apply to ourselves the same categories that are so successful in explaining causal relations between external objects in order to understand what “causes” us to feel the way we do, and then try to construe the resulting oversimplified and distorted causal account as the intentional meaning of the feeling. The classic example is the philosopher who, when asked how he felt about his wife, began by noticing the constant conjunction between certain feelings in himself and certain events in the environment, and at last concluded that there was a statistical probability that certain facts about his wife had caused certain feelings in himself. Clinical psychologists would call such a response “alexithymic” – unable to read the meaning of our own feelings. But to greater or lesser extents, we are all alexithymic in the same way as the philosopher in the example, because we fall victim to essentially the same mistake of reasoning. That is why the natural attitude is so “natural.” Antonio Damasio, in all his works, has offered a useful framework in which to understand emotional feelings essentially as reactions to interoception of internal body states associated with perceptual images. This is a valuable starting point for exploring the workings of the affective dimension, but it is only half the picture, and tends to lead either to oversimplified accounts of the intentional meanings of feelings, or to a dissatisfying silence on that score. My sympathy with a loved one, admiration for a sunset, or anger upon being cut off in traffic are not “about” my internal body states alone. At first glance, they appear as being about the loved one, the sunset, or the motorist who cut me off in traffic, but we cannot settle for such superficial, unanalyzed accounts of their meaning. Even these assumed intentional objects are often mislead-





Curious Emotions

ing, because the emotional life intends much more than what is immediately present in concurrent perception or imagination. The sunset may work as a symbol for the poignant finitude of the existential condition; the evil motorist may be only a symbol of all that oppresses and bullies me. The loved one may offer an opportunity to enact a highly complex energization of entire dimensions of my being. We have only begun to scratch the surface of understanding what our emotional feelings are telling us about how we are faring in our overall attempts to integrate the various motivational purposes that fit into the goal orientations of a complex, self-organizing system. The intentionality of feelings – what they are “about” – includes aims as well as environmental affordances and triggering objects. And many of these aims cannot be grasped within a simple stimulus-response framework that would treat conscious states merely as reactions to inputs and to the interoception of internal body states. On the contrary, many emotions and feelings have to do with aims that are active rather than reactive. In fact, it may well be that, as Newton (1989a) and Humphrey (2000) suggest, we are first motivated to perform actions, and then understand the independent reality of the external world in terms of the way it resists us. To define feeling too much in terms of interoception leads to confusion about the intentional meanings of both conscious feelings and unconscious emotions. As Salmela (2002) points out in agreement with Merleau-Ponty (1941), even unconscious affective processes often seem to have had intentional meaning prior to conscious reflection on them. To define feeling simply as the consciousness of emotions, and then to define emotions in turn as bodily processes available to interoception, blurs rather than clarifying this problem. Damasio, for example, defines feeling as interoceptive access to the bodily processes that constitute emotions; feeling, then, is the awareness/consciousness of an emotion which we simultaneously associate with some triggering image from the perceptual world or from perceptual imagination. But then the question arises: when we experience feelings, why are we so often un-aware of the body states that Damasio defines as emotions, and about which interoception is presumably informing the brain? If feeling is awareness of interoceptive information, then it should be impossible by definition to have feelings and remain unaware of the interoception. Damasio’s answer here, of course, is that we are often unaware or unconscious of the interoceptive information that our brains are receiving (Damasio 2003: 87ff.). But this explanation seems to straightforwardly contradict Damasio’s definition of feelings as the awareness of emotions. Damasio seems to want it both ways. He says that

Introduction

1. A feeling is consciousness/awareness of an emotion. 2. The consciousness/awareness of an emotion is the perception of a body state/body map (through proprioception/interoception). 3. The reason we usually are not aware of the body states/body maps of which our feelings (conscious emotions) are the awareness is that we are usually aware of them on an unconscious basis. Are 1, 2, and 3 logically consistent? Perhaps, if we are careful about what is meant by “unconscious awareness.” But even when we are consciously aware of our emotions, we do not seem to be consciously aware of most of the body states that constitute the emotion. In those cases where we have a conscious feeling, the question remains: If the emotion of which the feeling is the conscious awareness is a body state, then how can we be consciously un-aware of the body states of which the feeling is supposedly the awareness? To be sure, there is nothing wrong with saying that a feeling can include unconscious awareness of interoceptive information. The problem arises when Damasio wants to equate feeling with “knowing that” we have an emotion, and thus assume that the interoceptive information, plus the perceptual trigger with which it is associated, can add up to and exhaust the entire conscious content of the feeling. If this were so, then to empty the feeling of its consciousness of the bodily interoception should be to empty it of all consciousness whatever, except for the perceptual consciousness with which we associate it. But here again, we often do have feelings, and some understanding as to what they are “about,” without much consciousness of the interoceptive information that underlies them. Conversely, if the conjunction of interoception and perception were the whole story, then this information should tell us what our bodily system wants in relation to the given triggering event. But in many instances, we lack understanding as to why we feel the way we do toward the trigger, because we cannot focus clearly enough on what our bodies want to happen in all the various respects in which that particular trigger might be relevant. When we do gain such understanding, we often do so by shifting our attention away from mere interoception, and by paying attention to the efferent action affordances of the environment, which are multiply realizable. Part of the danger of modeling feeling as an “awareness” of interoceptive “information” from the body is that terms like “awareness” and “information” lend themselves to ambiguity, because they can be used in both a mental and a non-mental sense. Of course, the brain is “aware” of interoceptive data just as a billiard ball is “aware” of the angle and velocity of the ball striking it. But this





Curious Emotions

causal relation is not a mental or intentional relation. So, if we say that feelings are unconsciously aware of the body states that constitute emotions, then it turns out that the difference between feeling and emotion in Damasio’s theory really just depends on an extra layer of (often non-mental) “encoding” added to the original body states, but now at the cortical level. However, studies of event related potentials (ERPs) in the brain show that extensive cortical processing and encoding of information can occur absent any consciousness whatever (see Aurell 1989; Bachmann 2000; Runeson 1989; Shevrin 2001; and others to be cited later). Such an encoding does not lead to intentional (mental) content any more than, if someone were to replicate the face of George W. Bush on the side of Mount Rushmore, the mountain would then be “aware” of Bush. So the sense in which the brain is “aware” of emotions may or may not have anything to do with what the feelings are “about” in the conscious or intentional sense. To conflate this “awareness” with what the feelings are actually conscious of (or aware of in the intentional sense) creates confusion. It assumes, but does not clarify, the needed three-way distinction between (i) consciousness, (ii) unconscious yet mental or intentional “awareness,” and (iii) causal sequences that do not involve awareness in a meaningful sense because they do not involve “mentality” and “intentionality”. Once we sort out the meanings of these mentalistic terms, Damasio’s account seems to imply that emotions prior to being elevated to “feeling” status would not involve intentionality – a conclusion that I do not think Damasio himself would want to endorse. Correlatively, if we assume with Damasio that feelings are simply the interoceptive “awareness of ” emotions, which in turn are body states available to interoception, what is missing is a way to feel the aims of the feelings, since we cannot know these based merely on a constant conjunction of interoceptions with triggering objects. We then end up trying to get the “body image,” which is interoceptive and thus passive, to carry the weight of the “body schema,” which is enactive. I shall argue in this book, in agreement with Salmela and Merleau-Ponty, that unconscious emotions do already have intentional meaning, but that concurrent perceptual and imaginative objects, as well as interoceptive data, do not exhaust this intentional meaning. The intentional meaning has to do with the aims of self-organization, which are complex, relational, and oriented toward maintaining a pattern of self-initiated action, relative to available or potentially available environmental affordances. On this view, emotions are intentional – they are “about” some situation to which they refer – with or without any need for perceptual or interoceptive representations to accompany them. We do not cause emotions to have intentionality by associating them with a perceptual

Introduction

or interoceptive content of which we are already conscious. On the contrary, our consciousness of the world is rooted in the way we expect it to have affordances for possible actions, and those in turn are already motivated by the intentionality of emotion. To treat either emotions or feelings primarily as reactions to either internal or external inputs is therefore to put the cart before the horse. As Damasio himself has argued in his earlier work (Damasio 1994, 1999), it is precisely because emotions arise from the action proclivities of the organism (in my terms, the enactive self-organizing system; in Damasio’s terms, the action components of the “as-if body loop”) that there is a difference between conscious and nonconscious information processing to begin with. Information processing can be conscious only if it is motivated by organismic purposes, where “organismic” must be defined by distinction against “inorganic” or “merely reactive.” If feelings are in some sense conscious takes on emotion (I want to avoid oversimplifying here; feelings are not simply conscious emotions or consciousness of emotions, as we shall see), and if the emotions in turn have their intentional meanings in terms of organismic action proclivities, then the meanings of feelings too must include those action proclivities. Feelings will then be felt not only in relation to the interoception of body states, which is afferent, but also the efferent sending of action commands, which when inhibited lead to sensorimotor action imagery of the kind that Jeannerod (1997) and Newton (1982, 1996, 2000) discuss. Emotions tell us that we are being facilitated or thwarted in the complex patterns of activity that our total self-organizational organisms want to execute in order to achieve and maintain holistic balance at the appropriate level of energy and complexity.

. Some preliminary predictions of enactivism The enactivist and self-organizational approach to the neurophysiological correlates of consciousness and intentionality, during the approximate two decades since it began to be widely discussed, has already yielded a number of confirmed predictions. Six of these predictions are particularly relevant in exploring the meaning of affective processes, and will figure into the main arguments of this book. When the initial expositions of the theory were first proposed, as far back as Ellis (1986), Newton (1982, 1993), and Varela et al. (1991), none of these predictions had yet been confirmed, so the fact that they now have is intriguing:





Curious Emotions

1. The working of mirror neurons (Gallese 1998; Rizzolatti & Gallese 1998; Stamenov & Gallese 2002) confirms that many of the same neural systems are activated when we observe another person performing an action as when we perform the action ourselves, and that it is the other person’s intention to achieve a specific goal that is mirrored, rather than the literal physical movement (Wohlschläger & Bekkering 2002). For example, an infant imitating an adult’s pointing her left hand to her left ear will point with the right hand to the left ear if the left hand is not available. These findings are consistent with and were predicted by Newton’s (1993, 1996) assessment of how we understand other people by initiating the corresponding goal-oriented action commands in ourselves and then inhibiting them, in order to imagine the corresponding actions in others. In effect, we empathize by imagining ourselves doing what the other is doing (Newton 1996: Chapter 6). Correlatively, this prediction is interrelated with the Jeannerod (1994) finding that action imagery involves many of the same neural correlates as executing the action, except that in the imagined action we inhibit the action command – a finding that was sketchy when the enactivist theory was developed (Ellis1986, 1990, 1995; Humphrey 1986; Newton 1993) but is highly consistent with the theory. What the mirror neuron findings add is that our understanding of others is facilitated by this kind of action imagery, just as Newton had proposed. Moreover, when we image an action, the action command does not simply stop; it continues being sent, but is inhibited at a certain point in the cortex. 2. The Mack and Rock (1998) inattentional blindness findings confirm the enactivist prediction (Ellis 1986, 1995) that a prior attentional act is required for perceptual consciousness. Equally important, in many of the standard experiments in the Mack and Rock study, there apparently is no implicit knowledge of the unattended material. I.e., in most of the experimental designs where lack of attention prevents consciousness of presented stimuli, subjects’ answers to questions about the unattended stimulus are no better than chance (Mack & Rock 1998: 13). This often neglected aspect of the Mack and Rock research is uniquely predicted by enactivism, and contradicted by reactivist views – in which consciousness and knowledge would consist primarily of a receiving and processing of inputs. A reactivist view of consciousness and mentality would entail that there should be implicit knowledge prior to motivated attention, given the usual assumption that information normally is extensively processed prior to any consciousness, and that intentional goal directedness (as in the unconscious action imagery emphasized by enactivism) should play little part

Introduction

in the processing. If consciousness really were a correlate only of the receiving and processing of the input, then there should be consciousness of the information even if we are not previously motivated to actively direct our attention to it. The implications of these inattentional blindness findings call for careful analysis, and they will be discussed extensively in this book, especially Chapter 2. 3. Evidence is increasingly mounting that, as Damasio (1999) summarizes, a necessary and sufficient condition for “core consciousness” is that the emotional brain areas be intact. Damasio points out that there is consciousness in many cases of akinetic mutism, but not in permanent vegetative states, because in the former case the damage is mostly cortical, whereas in the latter it is further down in the brain. While reactivist approaches can grant that emotional areas are necessary for consciousness (functioning as a “power switch” for the cortex), they cannot grant that activation of the emotional areas is a sufficient condition for consciousness, as will be discussed more extensively in the body of this text. The processing of received information is more cortical than subcortical, so on the theory that consciousness is a receiving and processing of information, we should expect that removal of the brain areas that process information would remove consciousness. But it does not. Instead, it only alters the content and sophistication level of the consciousness. 4. Recent research increasingly finds that the cerebellum is more involved in higher cognitive operations than previously thought (Duncan 2003; Schahmann 1999; Schmahmann et al. 2001); this is consistent with enactivism because the cerebellum coordinates movement, and enactivism predicts that fairly subtle action imageries must be initiated (and then inhibited) to produce specific enough imagery of the corresponding action affordances vis a vis the environment. 5. Monkeys trained to play computer games with joy sticks can then have their brains wired directly to the joy stick, and thus play the game without using their hands to move the joy stick (Donoghue 2002; Hatsopoulos et al. 2001; Serruya et al. 2002; Wu et al. 1993). At first, an electrical brain implant picks up on the brain signal that the monkey sends in order to move its hand; then later the monkey merely imagines moving its hand, and the electrical brain implant continues to pick up on the same brain activity. This would not be possible except that the neurophysiological correlates of moving one’s hand are very similar to the correlates of imagining moving one’s hand. To imagine moving one’s hand is to continue sending the action command while at the same time inhibiting it, thus produc-





Curious Emotions

ing an action image grounded in efferent rather than afferent signals, and based on the inhibition of the action command through the action of neurotransmitters like acetylcholine (Hess & Donoghue 1999). Moreover, imagining moving one’s hand while not actually moving it involves more rather than less activity than just moving the hand. It involves all the activity required to move the hand (although perhaps with reduced intensity at the cellular level), plus the activity required to inhibit the command. In the experiment with the monkeys moving the joy sticks, the monkeys have already performed enough neural activity to send a signal to move the stick prior to the point where a decision would be made to inhibit or not inhibit the action command. This is also consistent with the fact that more intelligent animals (for example, humans) can much more easily imagine doing an action while not actually doing it, since the frontal area so much further developed in higher primates allows much easier inhibition of an action command once it is initiated. 6. The primacy of the efferent over the afferent in conscious processing, which was theoretically proposed by Dennett (1969: 80), can now be empirically demonstrated because of the unexpectedly long delay between the occipital 100P-200N processing of perceptual information and the parietal 300P event-related potential (ERP). The numbers in ERP measures refer to the number of milliseconds after presentation of a stimulus to the retina, while the “N” and “P” refer to positive or negative electrical potential. Thus, when an unexpected stimulus is presented, the ERP data indicate that occipital processing, and also V-4 temporal lobe activation, normally considered to be the “sensory areas” for perception, can be fully developed, yet until there is a positive ERP in the parietal area at about 300 ms, the subject has no perceptual consciousness of the object. There appears to be no perceptual consciousness of an unexpected stimulus until this parietal 300P occurs, even though the processing in sensory cortex is pretty much completed at this point (Aurell 1989; Runeson 1989; Bachmann 2000). As will be discussed later in this book, anything like an approximately 100–200 ms delay from occipital processing to perceptual consciousness (signalled by the parietal 300P) is much too long to be accounted for by the mere spreading of neural activation from the occipital to the parietal areas – which are immediately adjacent – and thus is consistent with the Ellis (1990, 1995) hypothesis that the efferent, anterior/frontal and subcortical processes involved in mental imagery and attention must be activated before sensory cortex activity can result in consciousness. Emotional and action-initiating processes are triggered as early as 18–20 ms,

Introduction

and involve a very primitive reaction to the object’s action affordances and emotional meanings, prior to any occipital processing. The parietal 300P simply signals that a resonance has developed between this anterior and subcortical efferent “looking for” activity and the corresponding posterior afferent activity in sensory cortex. The fact that without the 300P there is no perceptual consciousness confirms that afferent receiving and processing of information do not add up to consciousness. This is further confirmed by recent findings that there is a very early frontal activation (at about 100 ms) when an unanticipated stimulus is presented (Aftanas et al. 2001; Damasio et al. 2000; Federmeier & Kutas 2002), which again is consistent with the prediction in Ellis (1990, 1995) that action affordances must be activated (normally including some frontal motor/premotor and anterior cingulate activation) before there is consciousness. If reactivism were true, the occipital activation should be the cause of the perceptual consciousness, and then the frontal activation would follow rather than preceding it. The details of the timing of these ERPs will be discussed further as we proceed. The significance of these findings for enactive versus reactive approaches to perceptual consciousness can be summarized by means of the contrasting pathways illustrated in Figure 1. Each of the predictions just listed will be discussed in more detail in this book. There are many more predictions from enactivism that have not yet been confirmed, but the fact that so many already have is certainly enough to warrant interest. Part of the value of predictions is that they reveal something of the “cash value” of theories. This cash value can also be seen by drawing distinctions between one theory and another in terms of the way they interpret the same set of known data. One of the main practical advantages of an enactivist framework, I believe, is the way it accounts for the “higher” or more complex emotions, which have been a major stumbling block for reactivist accounts, particularly in its behaviorist and consummatory-drive reductionist versions. Chapter 1 will attempt to paint in broad brush-strokes some of the theoretical differences that result from taking an enactive starting point for understanding affects.





Curious Emotions ENACTIVE VS. REACTIVE BRAIN SEQUENCING

Parietal lobe

Frontal lobe Occipital lobe

(Object)

Thalamus

Figure 1. The DOTTED line indicates the sequence of spreading activation consistent with the stimulus-response paradigm, with perceptual consciousness a passive reaction in each successive area to an initial stimulus input. The SOLID line indicates the sequence consistent with an ‘enactive’ view, in which emotional interests (including curiosity) pre-select for conscious attention images relevant to those interests. Frontal (especially cingulate) and parietal areas activate image schemas in response to limbic emotional signals, and attention is directed to occipitally processed visual data only to the extent that they resonate with the motivated image schemas.

. The “curious” emotions On the enactive model, many of the affective responses that can be considered as unconditioned ones – not dependent on prior conditioning – would have been assumed to be conditioned responses on the reactive account. This is because enactivism assumes that a self-organizing system not only must act on its own; it must do so in complex ways, to maintain its needed level and particular patterns of complexity and energy expenditure. Curiosity, play,

Introduction

love, inspiration, loneliness, despair, existential anxiety, lust for adventure, the thrill of intellectual stimulation, the demand for social cooperation and freedom from oppression, the quest for an overall sense that life is worth living – these “higher” (i.e., more complex) human emotions do not show up on most short lists of “basic” or “primary” (unconditioned) emotions. In fact, they are usually thought of as derivative through learning and socialization from more basic ones, such as anger, fear, joy, and sadness. The “short list” approach, then, assumes that only simple emotions can be unconditioned, and that complexity results from conditioning. Most of these simple schematizations of basic emotions are derived either from philosophical analysis (for example, Griffiths 1997; Ben Ze’ev 2000) or from the assumptions of behavioristic psychology and the lingering effects of a century of behaviorism (for example, Ekman 1999; Plutchik 1980). The most prominent way of explaining the higher emotions as derivative from the “basic” ones is to posit that the basic ones are hardwired responses to specific types of stimuli, and then either operant or classical conditioning leads to stimulus generalization or transfer, with punishment or reward based on satiation of comsummatory drives serving as reinforcement (LeDoux 1996; Rolls 1999). Mixtures and blends of the resulting responses then supposedly account for the more subtle and sophisticated affective processes (Plutchik 1980). But a century of intensified work based on these assumptions has brought us no closer to understanding the most interesting aspects of the human emotional life – the higher, presumably “secondary” emotions – than we were before. Indeed, it might be argued that we have backslid, because the behaviorist and neo-behaviorist mode of analysis has suffered from careless interpretations of phenomenological experience, in a way that oversimplifies the intentionality of even the most “basic” feelings. If we turn to actual neuropsychological studies of the way higher emotions are orchestrated in the emotional brain, as in the work of Anderson (2000), Damasio (1999), Panksepp (1998), or Watt (2000), we find a much more complex picture. Love, curiosity, separation distress, play, the need for freedom of movement, just as much as anger or fear, are innately rooted in systems of neural connections, anatomical structures, and neurotransmitter activities not only in the human brain, but even in the brains of the lowest mammals. These systems are not merely overlays of culture or learned responses. They are not derivative through associative learning either from consummatory drives or from simpler responses to specific hardwired stimuli, as is usually presumed to be true for “primary” emotions such as anger and fear.





Curious Emotions

The term “consummatory” can be used in at least two senses. Emotion researchers speak of an “appetitive phase” and a “consummatory phase” of motivated activity. But another interrelated but different sense of “consummatory” – the one relevant for our purposes – is the one used in traditional learning theories, where consummatory drives are usually thought of as motivating behaviors needed for the organism’s energy-efficient homeostasis, such as eating, drinking, and the chemical reactions involved in sexual consummation. The point at which the energy-efficiency needs of the organism are met, where homeostasis is regained, is presumably the point at which the motivational need or desire is “consummated.” The issue at stake for our purposes is that, increasingly, it appears that many of the “curious emotions” just listed, including highly social ones, are not geared merely toward energy-efficient homeostasis. On the contrary, they seem to seek higher levels of complexity and energy at which homeostasis is still possible – yet they still seem to be primary rather than derivative. This explosion of new physiological findings about the emotional brain has coincided with the emergence in philosophy of mind and psychology of new ways to think about the mind-body relationship. One now hears more and more about enactive (Varela et al. 1993; Ellis 1995; Newton 1996), or “embodied” (Clark 1997; Gallagher & Marcel 1999; Lethin 2002), or “dynamical systems” (Freeman 1988; Juarrero 1999; Thelen & Smith 1994) or “self-organizational” (Ellis 2000a, b; 2001a, b, c; 2002a, b, c) accounts of the relationship between consciousness and cognition. In all these new perspectives, consciousness is part of the organism’s attempts to use its environment in the service of purposeful organismic action. This is a reversal of the usual way of thinking about causes and effects in human psychology. Throughout the twentieth century, the stimulus-response paradigm was dominant. Emphasis was on the role of the stimulus as causing the response, given the previous state of the organism (including the results of its conditioning history). Since consciousness apparently required widely distributed global brain activity, it was traditionally assumed that consciousness must be the final step in a chain of causes and effects, the final response to a stimulus, with informationprocessing going on in the intervening steps. This led to epiphenomenalism, micro-reductionism, and what Natsoulas (1993, 2000) calls “appendage” theories of consciousness: Consciousness was considered to be a mere side effect of the real causal processes, which were at the micro-level. This epiphenomenalism would actually entail a hidden dualism, since if consciousness really had no causal power, then it would not be identical with any physical process that does have causal power.

Introduction

Emotion too suffered from this epiphenomenalist causal story, often resulting in an utter reversal of cause and effect. Since emotions and affects are supposedly “responses” to environmental conditions and events, it was thought that inputs must first be processed, at least to a considerable extent, and that this processing must then cause the resulting emotional responses. For example, Rolls (1999) argues that we must first process a perceptual stimulus before an emotional “response” can take place, and that the perceptual processing will be substantially independent of and prior to any emotional influences. Rolls’s reasoning, as I shall discuss in more detail later, is that perception of a food stimulus is not impaired by lack of hunger in an experimental animal. Thus in Rolls’s view perception and other cognitive processes must precede any emotional response. The enactivist, self-organizational kinds of approaches just mentioned would reverse this sequence. Emotions are aspects of the organism’s ongoing self-organizational activity, and these emotional processes drive the processing of information rather than being merely responses to it. Objects and events do not so much cause affective responses as they are used by the organism for its own emotional purposes. The Rolls example just mentioned does not capture this point because the interest in seeking out potentially useful environmental affordances, such as food items, is primarily a result of a general emotional system – Panksepp’s “seeking” system – and operates independently of whether the organism happens to be hungry at the time (see especially Panksepp 1998: Chapter 8). Our interest is always geared up to look for useful affordances. But Rolls is correct to point out that we cannot assume a simple oneto-one correspondence between a perceptual object, such as a food stimulus, and a unitary affective quality that would motivate attention to that stimulus, such as hunger. The intentionality of emotion is much more complex than this. The motorist who cuts in front of me is not simply or even primarily the intentional object of that momentary feeling of anger; on the contrary, he is used by me to express emotions that were already important for my total selforganizational purposes, and may frequently involve larger life issues than the trivial but powerfully triggering event of being cut off in traffic. To understand the real intentionality of an emotion or feeling requires exploring the way these total life events contribute to the affect. Following convention, I use the term “feeling” for conscious or preconscious registering of the affective quality of an emotion – where “preconscious” refers to processes able quickly to become conscious – although there will also be completely non-conscious aspects of the emotion itself that involve many





Curious Emotions

physiological processes not accessible in terms of a felt quality. In spite of the intuitive simplicity of these distinctions, however, the relationship between conscious and unconscious emotion is a good deal more complicated than this conventional usage might lead one to believe, as we shall soon see. Let’s now take a closer look at how an enactivist approach to affect and emotion might be worked out, beginning with the way some of the key concepts can be set forth in a self-organizational framework. Before going further, we should devote a little attention to developing preliminary definitions of some basic concepts needed for a coherent enactive approach to the affective sphere. As noted earlier, the most fundamental concept is the distinction between “action” and “mere reaction.”

. Conceptualizing action versus reaction To define what is meant by “action” as opposed to “mere reaction” is a difficult philosophical problem in its own right, including the puzzle of how our emotions enable “mental causation” to occur, so that the whole organism can purposely use its parts rather than being a mere summation of their separate mechanical reactions. Chapters 2 and 3 will flesh out the differences between action and mere reaction in some detail, but at this point a preliminary definition is in order. The recently elaborated concept of complex dynamical systems (Bunge 1979; Juarrero 1999; Kauffman 1993; Kelso 1995; MacCormac & Stamenov 1996; Monod 1971; Newton 2000; Prigogine 1996; Thelen & Smith 1994) can provide the necessary tools for such an understanding. Complex dynamical systems, in the terminology of these theorists, are structured so as to preserve certain continuities of behavior across exchanges of energy and materials with their environment. The more they must seek out and appropriate the replacement components to keep the pattern of organization going, the more clearly they fall within the category of “living” systems (Bertalanffy 1933; Bunge 1979; Monod 1971). The activities of living systems, then, are not merely built up through the accumulation and summation of the causal powers of their micro-constituents, although their behavior is consistent with normal causation at the micro level. I shall argue that there is also a causal power on the part of the whole system, to regulate – and in fact even to find and purposefully make use of – micro-components on an as-needed basis. Emotions are expressions of this self-organizational tendency, and this is why emotions can cause actions holistically rather than in partes extra partes fashion.

Introduction

The sense of “action” to which I am referring here is simpler than the more elaborate notion of the planned, deliberate actions of a conscious agent. What is more basic than, but presupposed by, action planning is the idea of a system that moves its own parts for the purposes of the whole, rather than merely being moved by its parts. The system must have the ability to rearrange its own components to maintain certain structural relations for the whole organism, and in the sense of “action” to which I am referring this power will not be a mere epiphenomenon, but rather will play a genuinely causal role. There is no need to worry that physical systems will fail to be “causally closed” (Kim 1992, 1993, 1998) on this analysis. Self-organizational theory does not contradict causal closure at the micro-level, as I shall explain in detail later. Each step in a reaction is causally necessary and sufficient for the next step, granted certain background conditions. But it is the system as a whole that monitors and adjusts the background conditions as required to render a given micro-level event causally efficacious for the outcomes needed to achieve the system’s overall objective, which is to keep acting in its own definitive pattern. Such a pattern obviously must be very complex in order to incorporate the right kinds of subsystems and interrelations of homeostatic feedback loops to achieve this kind of regulation of the causal conditions governing its own components. They must be at least as complex as a living cell. And they must be even more complex if they are to achieve enough intentional understanding of environmental action affordances to interpret disruptions of their own selforganizational patterns as indicative of some aspect of their overall relation to their environment. To think that environmental events cause organisms to act would be an oversimplification on the self-organizational view. As Bickhard (2000) emphasizes, organisms are always already acting on their own, unless they are dead, and discrete environmental events can only “disturb” the ongoing pattern of self-initiated action, so that a readjustment must be found somewhere in the pattern’s total configuration that can reorganize the parts of the organism as well as the environment, in order to maintain the general contour of the whole pattern of activity as nearly as possible. Sometimes this may mean radically changing the environment (as in fight), sometimes it may mean repositioning the organism vis a vis the environment (as in flight or as in turning away in disgust), and sometimes it may mean reorganizing the organism to find different ways to achieve the desired homeostatic patterns (as in cooperative agreements and other complex methods of getting around obstacles to freedom, or as in adjusting to tyranny by cultivating a mood of quietistic resignation). Any com-





Curious Emotions

bination of behaviors able to sustain the desired homeostatic patterns is as “natural” as any other. Just as emotional responses depend on a total self-organizational context, so do the emotional meanings of stimuli. We should not think of an emotion as just another type of conscious qualia caused by certain kinds of inputs, alongside perception, thought, and the proprioception and interoception of bodily sensations. Emotion is unique in that its qualia usually do not take specific inputs as their intentional objects, since they are always felt in relation to the action affordances of objects, and these will be different depending on the object’s relations to the ultimate aim, which is self-organizational and can be realized by a variety of combinations of conditions, both environmental and bodily. The ultimate aim is the establishment and maintenance of the animal’s definitive sequence of self-organizational patterns relative to the environment; the disruption of the pattern is assessed at the level of the whole, not in an isolated part. I do not mean to imply that, experientially, intrinsic value is not located outside of the organism and its aims – a philosophical problem to which we will return later. My point for now is that, even though perception and thought may uniquely identify a likely candidate as the environmental stimulus that is largely responsible for a disruption, the felt quality of the emotion itself is not this specific. One instance of anger initially seems just like any other instance of anger, whether it is caused by the motorist who cut in front of me, or whether the true cause of the anger is a general frustration with the way my career and love life are going, with the motorist’s behavior serving only as an opportunity to symbolize these feelings. The felt quality of the emotion does not pinpoint what the real disturbance is in the first casual or superficial glance. The emotion reveals that something is “off ” in the organism-environment relationship, and it reveals some general features of this “offness” in terms of action affordances – that something is confining or constricting my activity. I shall discuss later the ways we can and typically do explore the hidden nuances of a felt quality to get more specific information out of it, but this specific information is not obviously announced by a superficial reading of what an emotion initially seems to be triggered by – again, because the organism’s aims are holistic and thus must be understood in terms of a larger picture. Many people may feel that some emotions or feelings are not oriented toward action. For example, in basking in the calm beauty of a sunset, I may simply want to sit passively and sink it in. But in the sense of “action” we are using here – unified action initiated by a self-organizing entity as opposed to mechanical reaction to an external event – the presumed passivity of the reaction to the sunset is misleading. We may not know about all the internal

Introduction

activities that must be taking place in order to maintain the homeostatic patterns that are needed to subserve the pleasurable experience of the sunset, but they are taking place nonetheless. Moreover, as I shall argue in detail in later chapters, I would not be looking at the sunset in the first place had I not chosen to direct my gaze in that direction. The experience of the sunset is being chosen and appropriated by the organism as a whole because it facilitates rather than thwarting the self-organizational structures of activity with which it resonates. If something were a little “off ” in any of the activities within the total homeostatic system, then different environmental conditions would be needed to facilitate regaining the proper balance. This need might be consciously experienced as boredom, or simply as an interest in directing attention to some other object or activity. This point remains applicable even in the aesthetic experience of artworks. In fact, the aims served by the experience of art are especially revealing about the relation between emotions and their intentional objects, because in the experience of art we are forced to take seriously the ambiguity of object affordances in terms of our emotional use of them, many of which are preconscious until we explicate them with the help of the perceptual object. To further clarify this distinction between conscious and preconscious intentionality will be the central focus of the first chapter.

. Plan of the book This book will focus on the interconnections between self-organization and the intentionality of conscious experience, and the ways in which emotion and the motivation for attention-direction can enter into this relation. In thinking about the intentionality of action, there should be careful consideration of the contributions of action theory, and the ways action can relate to the forming of the understanding of intentional contents (as in Clark 1997; Ellis & Newton 1998a, 2000a, b; Gallagher & Jeannerod 2002; Humphrey 2000; Juarrero 1999; Newton 1996). This kind of analysis seems crucial in laying the foundation for an intentionality of affective consciousness grounded in the intentionality of action. In my view, to make sense of these interconnections, there also needs to be discussion of the way the active rather than passive nature of emotional intentionality connects with the notion of self-organization. A good source for exploration of this dimension can be found in Stuart Kauffman’s work, and other authors who use self-organization and/or dynamical systems to ground a difference between purposeful action (as in biological organisms) and mere reaction (as in machines). These elements, combined with careful attention to





Curious Emotions

phenomenology (as it relates to neuroscience and analytic philosophy of mind) can form the basis of a systematic attempt to show how emotional consciousness is grounded in an organically self-organizing system that acts to preserve and express its patterns of organization. We then find that many emotions once considered as responses to stimuli, such as fear and anger, can now be thematized as the body’s attempt to get rid of or get around impediments to its already ongoing activity. If this dynamical framework goes all the way down to the most basic emotions, then the distinction between higher and lower emotions cannot be drawn by reducing the former to the latter by means of learning schemes or socialization processes. Instead, the higher emotions will be more complex “attractors,” or patterns of organization into which the system’s endogenous dynamical rhythms motivate it to shift, and in which it strives to remain as long as maintenance of the total system requires it. Nor can the emotions be neatly reduced to elaborately elongated schemes for the attainment of chemical homeostasis. It is true that dynamical systems seek homeostasis, but they also seek to maintain suitably complex and highenergy basins of attraction to fit the life trajectories that various life forms set for themselves, because of the specific ways in which they are self-organized (Juarerro 1999; Prigogine 1996). This interesting fact, that animals do not seem to want merely to satiate consummatory drives, but also seem to intrinsically value adventure, playfulness, and curiosity, will be one of the pivotal issues to be explored throughout this book. A theory as to why there would be such important non-consummatory motivations and emotions will be spelled out in some detail in Chapters 4 and 5. But certain groundwork must be established first. Chapter 1 will provide an overview of the way an enactive approach affects our conceptualization of the basic terms used in discussions of affect and emotion, and our understanding of the intentional meaning as well as the underlying neurophysiology of affective feelings. Since the action/reaction distinction is a pivotal point in this re-conceptualization, Chapters 2 and 3 will then develop an overall concept of self-organization that is capable of making sense of the notion of action in the most fundamental sense (as distinguished from mere reaction), and will show how a capacity for action in this sense allows holistic motivations to cause our bodies to move in executing useful actions. Chapter 2, specifically, will investigate the interrelations of emotion and the imagery associated with emotional intentionality, by taking its cue from the phenomena of inattentional blindness and selective attention and inattention. Emotion will emerge as a determinant rather than just a response to our consciousness of the world – a fundamental component of any conscious experience. Chapter 3 will examine the way gen-

Introduction

uine mental causation and self-initiated volition, as opposed to mere systems of piecemeal reactions, become possible from the self-organizational standpoint. After this groundwork has been established, Chapter 4 will then develop the idea of non-consummatory motives as specific emotional applications of dynamical systems theory, and will ground an understanding of the “higher” affects as related to the organism’s needs vis a vis three types of conditions: (1) “extropic” (i.e., non-consummatory); (2) homeostatic; and (3) boundary protection conditions, viewing all three types of conditions as necessary features of complex dynamical systems. Chapter 5 will show how this emotions system grounds the various senses of self. Chapter 6 will then combine this analysis with the previous discussions of non-consummatory affects, taking some clues from the arts and from psychotherapy, and exploring some “existential” issues that arise from awareness of the challenges facing the problems of the self-world relationship in general, which are of crucial concern to very intelligent beings. These issues include the problems of death, powerlessness, social alienation and oppression, and the need to experience the value of being per se with enough intensity to justify life’s struggles. Here the complicated relationship between an emotion, its intentional objects, and related imagistic triggers will become more clear. Finally, Chapter 7 will show how emotion serves to organize experience into the structure of a self in the sense of personality. Not only primitive agency is made comprehensible by the phenomenon of self-organization, since organisms act rather than merely react, but also it can be seen how reflection-capable subjectivity and overarching unity of the flow of experience within a coherent personality structure become possible through the pattern of motivated attention progressions in the stream of experiences.



Chapter 1

Preconscious emotional intentionality

While the felt quality of an emotion cannot be relied upon to pinpoint a specific environmental object as what it is “about,” it would also be a mistake to say that emotional qualia per se do not have intentional objects until we cognitively interpret the cause of the qualia. Salmela (2002) expresses the essential problem here very incisively, taking issue with the view of many cognitivists (for example, Ben Ze’ev 2000) that a feeling has no intentionality until we cognitively assess its cause. According to Ben Ze’ev, we know that we are in love because we feel certain sensation-like feelings such as butterflies in the stomach, and then cognitively associate a certain person with that feeling, noticing that we have the feeling whenever the person is present. Salmela’s objection is that a feeling of butterflies in the stomach would not be identified as love, or indeed as any emotional feeling, unless it already contained an intentional reference even at the pre-cognitive level. It could just as well be an upset stomach or a feeling of dizziness. We can identify it as an emotional feeling, rather than dizziness or an upset stomach, because we sense precognitively that it already refers to something beyond itself. We sense that it indicates something about our relation to some aspect of our environment. When we reflect on an emotional feeling, we do not add the intentionality to it as a side-effect of the reflection; we discover that it already had an intentional reference. There must already be an intentionality inherent in the feeling, by virtue of which our judgment as to which aspects of a situation are the referents of the feeling can ever hope to be either correct or incorrect. This point can also be seen in a more mundane, empirical way: We react to an emotional stimulus more quickly than we can even cognitively register the content of the object by means of perceptual processing. The hippocampalcerebellar-hypothalamic loops that instigate action begin as early as 18 ms after presentation of an unexpected stimulus (Haines et al. 1997), whereas the occipital lobe does not process the visual identity of the unanticipated object until at least 200 ms later (Aurell 1989). When we then reflect on why we feel the way we do, the feeling reveals itself as already having been intentionally directed to the action affordances of an object or environmental situation. For exam-



Curious Emotions

ple, an infant does not need to learn the fear of heights, because the edge of a precipice affords falling. The fear reaction is signalled by neurotransmitter activity (glutamate, monoamines, and other neuropeptides) in a circuit including the lateral amygdala to medial hypothalamus to PAG and pons, prior to perceptual and intellectual processing (Panksepp 1998; LeDoux 1996). Damasio (1999) also emphasizes that emotional responses occur earlier in the processing stream than perceptual and intellectual ones. It might be objected that the initial pre-cognitive response was only a mechanical effect, not an intentional one. But in order for the mechanical effect to take place, the organism had to already preconsciously understand something about the action affordances of the environmental conditions. Otherwise the action would be to no avail. While it is true that initial pre-cognitive actions toward objects are very general and amorphous, in order to maintain real-time flexibility that will increase efficacy, the fact remains that the actions initiated at the precognitive level are not completely off base. We sense threat preconsciously and tense up, prior to cognitive understanding (or misunderstanding) of the nature and source of the threat. The intentionality is there. When I “displace” feelings onto a motorist who cuts me off, I know on some level that many other frustrating issues in my life are lurking in the background, and this is why I tense up when the motorist cuts me off. The subsequent identification of the motorist as the cause and the intentional object of my anger is only a cognitive deception. In a sense, as Gendlin (1992a, b) says, the body knows better, and this is why beating up the motorist usually does not satisfy the emotion. The affective impact of environmental conditions (past, present, or anticipated) need not be slowed down by their subtlety or intricacy. The internal complexity of a land mine does not prevent it from going off immediately when tripped. A particular insult may be only the tip of the iceberg of the emotion’s intentionality, and to beat up the person who insulted me may therefore miss the point. Not only would the feeling not achieve its aim; I also might not achieve the limited aim of understanding what the emotion is trying to tell me about my situation, which is the intentionality of the feeling. In order to allow an appropriate role for preconscious emotional intentionality, we must distinguish emotions from motivations without thereby equating the emotions with conscious feelings. Up to this point, the terms “emotion” and “motivation” have occurred repeatedly, and we have acknowledged the possibility of both conscious and unconscious emotions. However, motivation should not be equated simply with unconscious emotion. The relationship between these concepts is considerably more complex than that, as the next section will discuss.

Chapter 1. Preconscious emotional intentionality

.

Motivation, conscious emotion, and unconscious emotion

An old riddle asks how many psychiatrists are needed to change a light bulb. The answer: only one, but the light bulb must genuinely want to change. A psychiatrist can motivate a patient to move, or to change, but not in the way that a light bulb is moved or changed. Simply to push against something with a mechanical force is not necessarily a motivation, no matter how complex the mechanism that transmits, processes, or transforms the force into the object’s resulting motion. What, then, is the difference between motivation and mere movement? The self-organizational approach provides a clear way to make this distinction. We can speak of a system’s being motivated only in cases where the self-organizing system’s self-initiated aims toward maintaining its ongoing patterns of activity are necessary to supply the force for the movement. Since a light bulb is not a self-organizing system, and has no self-initiated aims, the usual psychiatric methods cannot motivate it, although as a last resort the psychiatrist can physically move the bulb. When a patient with high blood pressure takes medication to control the problem, he does not motivate his blood vessels to relax, but only causes them to relax. But when someone’s heart beat speeds up while making a play in an athletic game, it does so because the player is motivated. The player in our example may not realize until after the play is over that he felt nervous that he might not have been able to complete the play successfully, and that this nervousness was reflected in the increase in his heart rate and blood pressure. Yet the nervousness is properly referred to as an emotion rather than just a motivation. An emotion, as we saw earlier, can be present even though we are not conscious of it. The difference between a motivation and an emotion cannot be simply that an emotion is a conscious motivation, or the conscious result of a motivation. In that case, there would be no such thing as an unconscious emotion. In fact, we cannot even say that motivations lack an intentional understanding of the organism-environment relationship at some basic level, whereas emotions have such an understanding. There is a non-conscious sense of “understanding” in which even a plant can acquire, process and act on information about its environment, so in this sense even a plant has some simple understanding of the action affordances of its environment. If the light is not in the direction in which the plant has been growing, it changes the direction of growth to achieve self-organizational aims. But there is something about plants that we want to say precludes attributing “emotions” to them. How, then, can





Curious Emotions

we define precisely the difference between a motivation, an emotion, and any other biochemical process? To some extent, the way we make this distinction might be an arbitrary terminological convention, but here is the way I shall distinguish them in this book: An emotion is a process through which an organism monitors, in a coherent, unified way, how well it is doing in achieving and maintaining the various motivated patterns in its overall state, and forms a sense of what needs to be done vis a vis the environment in order to achieve or maintain the particular self-organizational balance that is motivated. For example, the feeling of fear alerts the organism that its bodily boundaries are in danger of penetration or injury, while anger means that freedom may be constrained. The motivations at stake here are the needs for freedom, security, and boundary integrity; these motivations could be present in organisms without emotion or consciousness, such as amoebae. To attribute emotion to an amoeba would seem clearly to be an anthropomorphism, but amoebae do behave in motivated ways. By contrast, emotions such as fear, in organisms sophisticated enough to be capable of such emotions, help them to monitor how they are doing with respect to freedom and boundary needs. This knowledge can then be used to make more complex decisions than an amoeba could make. By virtue of even a vague knowledge of how we are doing with respect to freedom and boundary needs, we also can form some sense of what kind of action needs to be taken. If there is a cue suggesting that our movement is about to be impeded, or that our boundary integrity is in danger of being violated, then it prereflectively makes emotional “sense” to move in a different direction. How an organism learns to recognize such cues is a complex story depending on empirical details which should not be presupposed by the mere definition of the word “emotion.” Definitions should leave empirical questions as open as possible, so that different claims can compete while sharing common definitions of terms, facilitating communication and debate. Suffice it to say for now that emotions can alert the organism that something is either going well or not going so well with respect to the organism’s motivating needs and values (its “motivations”), and there are certain environmental affordances that can serve as cues indicating, at least tentatively, that something may be “off.” For example, suppose the cue in question is a set of images suggesting that we are at a great height. Before an infant even knows what it is like to fall from a great height, it will instantly and instinctually recoil from such height, if the cues are sufficiently registered (Panksepp 1998: 215). But given what we saw earlier about the timing of brain events, we can assume that the fact that height affords falling is registered more quickly than the cortex can process enough

Chapter 1. Preconscious emotional intentionality

perceptual data to consciously identify the perceptual details of what we are looking at – a chasm for example. As soon as the most vague, minimal cues as to the action affordances of the chasm reach the thalamus (where they will be minimally processed and relayed to both perceptual and emotional brain areas), the organism already senses that something is “off ” and begins gearing up for possibly vigorous action to shift its momentum away from the direction of the chasm. This sensing of the holistic “offness” of the organism-environment relationship vis à vis the organism’s motivations (in this case, the motivation to preserve boundary integrity and control of its own movements) is already an emotion – fear – prior to the completion of perceptual and intellectual processing. If organismic motivations are already present prior to forming emotions to indicate how well we are doing with respect to them, then it is necessary to distinguish the terms “motivation” and “emotion.” Some organisms can have motivations and act on them without emotions. Even amoebae can readjust their behavior to restore homeostasis, given the right environmental conditions. To be motivated requires only that a system be self-organizing and complex enough to regulate its patterns of organization, readjusting its behavior in relation to the environment. In organisms incapable of emotion, the needed readjustment is an automatic aspect of the self-organizational pattern, although the environmental adaptiveness of individual organisms at this level is minimal (more extensive adaptation depending, for them, on genetic mutations and gradual evolutionary changes). An emotion, on the other hand, is a higher order phenomenon. In this case the organism does not only behave automatically to restore homeostasis. To be sure, it may respond quickly, but it also maintains a lingering sense of what is “off ” in the organism-environment relationship (“off ” with respect to self-organizational aims). The availability of this information makes possible not only a long range decision as to what is the best course of action, but also a balancing and prioritizing of the various motivations within the system. The more complex the self-organizing system, the more different ways there will be to balance disparate motives. This follows from the fact that complex dynamical systems, by definition, can rearrange elements of their subsystems to maintain a higher level pattern. They can appropriate and replace components as needed, can readjust the background conditions needed for various micro-level causal sequences, and can use multiple shunt mechanisms and feedback loops to ensure massive overcausation of the multiply realizable homeostatic aim. So when the balance of the whole is disrupted, the system must monitor the state of the whole in order to decide which of the many sub-





Curious Emotions

systems can best be re-adjusted – for example, the fear system, leading to hiding or relocation of the organism, or the anger system, leading to a getting rid of environmental impediments. The definitions being used here allow for the possibility of motivation without emotion. This does not preclude, however, that emotions themselves can motivate. There is nothing incoherent or nonsensical about making the claim that an emotion such as compassion, having been felt as an emotion, can then motivate kind behavior. To say this is only to say that the organism then wants to act in a kind way, and further environmental events can then either impede or facilitate the kind behavior, leading to still further emotions which may then motivate still further actions. Given the above definition of emotion, most people would want to allow for unconscious as well as conscious emotion, although powerful emotions do normally have a strong tendency to break into conscious awareness. It makes sense to acknowledge unconscious emotion, not just because the organism behaves in ways that are analogous to conscious emotion, but because it actually has information about what needs to be changed in the organism-environment system to satisfy the motivations reflected in the emotion. The athletic player who unconsciously “tenses up” understands that there is danger of making an error, and understands this in relation to the central aim of making the correct play, without having to pay conscious attention to what he is feeling (or to be more precise, not-feeling). This distinction between conscious and unconscious emotion might sometimes even be a distinction of degrees. The player in the example may actually have been slightly conscious of his nervousness before becoming fully aware of it, even though most of his consciousness was taken up with other things. In the same way, in Armstrong’s (1973) famous example of the truck driver whose driving is all unconscious, there may actually be a minimal level of consciousness of the cognitive decisions involved in driving, as Faw (2002), Natsoulas (2001), and Newton (2001) have suggested. But there is more to an unconscious emotion than a mere motivation. The emotion includes a balancing of the competing demands of the subordinate motivations of a number of sub-systems, and an overall sense of how this balance can relate to the environment.

Chapter 1. Preconscious emotional intentionality

. The murkiness of emotional intentionality Many theories of the higher affects – i.e., affects that are more complex than the simple ones on the “short lists” – not only assume that specific triggering stimuli are both the causes and the intentional objects of emotional responses, but also postulate that the intentionality of a feeling or emotion can be equated with the subject’s cognitive identification of its cause – that this determination is added by means of a cognition after the proprioceptive or interoceptive quality of the feeling is present. According to these “cognitivist” theories, I would feel certain bodily sensations and then theorize as to what environmental circumstances might have caused me to feel that way (Ben Ze’ev 2000). It is assumed that, the more complicated the cognitive assessment becomes in this interpretive process, the further removed will be the affect from the “basic” emotions from which it is derivative. For example, on the cognitivist account, I may feel a subtle combination of fear, anger, and sadness, and cognitively figure out that what I am feeling is resentment against someone who sharply criticized my work. According to the cognitivist approach to higher affects, elaborate intellectual and rational processes are necessary to determine the intentional referents of higher affects, and this is what makes them different from the “basic” emotions, which are hardwired to accompany specific types of objects and do not require such complicated intellectual processes. But we have already begun to see that unconditioned emotions can reveal complex information about the organism-environment relationship prior to cognitive assessments, and in all cases the emotional response is already underway before we have even processed the perceptual event presumed to trigger it. Cognitivism would have to be modified in this and a number of other respects in order to remain consistent with recent findings about the mammalian emotional brain. In the cognitivist framework, identifying the causes of our own feelings is essentially like any other inductive reasoning process. But this view has some strange consequences. The usual intellectual process of first associating two simple events with each other and then reasoning inductively that one must be the cause of the other would seem to be an abnormal way to determine the intentional “aboutness” of emotions and feelings – as in the case of the alexithymic’s feelings toward his wife. There are two main problems with positing an inductive reasoning process as the way we interpret the intentionality of emotions and feelings. In the first place, the “aboutness” of emotions cannot be as simple as a one-to-one correspondence between an immediate object and an interoceptive qualia; if the aims of emotions are self-organizational, the emotion will have its mean-





Curious Emotions

ing only in terms of a larger context, and the aim could just as effectively be achieved by adjusting other elements of the larger self-organizational picture as by taking a specific action toward a specific object. Anger, for example, might be about deep and complicated life frustrations, not about the trivial insult that triggers it. Love can be expressed through altruistic behavior or writing music as much as by sexual activity. In the second place, in order to assess whether a given object qualifies as the intentional object of an emotion or feeling, we must be able to compare it somehow with the felt quality itself to determine whether it does or does not resonate. If the intentionality of the feeling were not already implicit in the felt quality itself, then there would be no basis for such a comparison, and we could never know how we feel. If emotion were not pre-cognitively intentional, then a feeling of love might just as well be dizziness or an upset stomach. A mere conjunction between such feelings and some perceptual object cannot support an intentional or “aboutness” relation. In concert with my basic principle that emotions aim toward holistic selforganizational purposes, rather than toward specific environmental events, I shall argue that the higher affects do not result from an intellectual and cortically driven interpretation of discrete environmental “causes” of our feelings. If the aim of an emotion is to re-establish a holistic organizational balance in a complicated and interrelated set of organismic activities, then as Gendlin (1992a, b) suggests, the organism must already have a felt sense of the direction in which it is trying to move before identifying any specific object as helpful or hurtful toward that end; and whether an object is helpful or hurtful will have to be assessed in terms of the body’s holistic sense of its total life situation, not just in terms of what is going on in the environment at the immediate present moment. Lacking such a felt sense in the emotion itself, it would be impossible to know whether a given environmental circumstance resonated with the feeling in the way needed to qualify as the feeling’s intentional object. Moreover, assessing what is wrong will not be nearly as simple or straightforward as identifying a perceptual object. If what immediately presents itself in perception is the motorist cutting in front of me, this does not mean that the motorist cutting in front of me is simply what the emotion is really about, or what causes it, or what the emotion aims primarily to change. In many cases, if not all, beating up the motorist will leave the underlying anger and frustration intact, because he is not what they are really “about.” In this more complicated kind of case, to determine what the emotion is really about will require that I delve into the emotion itself as directly felt, so as to see whether the motorist’s cutting in front of me can qualify as an adequate

Chapter 1. Preconscious emotional intentionality

explanation of it. But meanwhile, as Gendlin (1980 1992a, b, 1998) stresses, the felt quality of the emotion must have a fine-tuned specificity; otherwise I would not be able to identify its intentional object by reference to that specificity. Thus the aboutness of the emotion cannot result from attributing an object to it, but on the contrary we can know which objects are associated with the emotion only by reference to the intentional aboutness already present in the emotion. We must already have a sense of what the objectives are in relation to which an object or complex of environmental conditions presents itself as helpful or thwarting – if not, we could never figure out by means of cognitive activities what an emotion really wants to happen – i.e., what its aims are. Even in these more complicated cases, there are methods of getting at the murky holistic intentional referents of emotional feelings, many of which are initially unconscious. But before discussing this problem, it is vital to recognize some further conceptual distinctions, especially the distinctions between the aims, objects, triggers, and symbolization-vehicles of affective experiences.

. Aims, objects, triggers, and symbolization-vehicles On the self-organizational view, the precognitive emotion knows that it is about something, but the aboutness is vague and general; the felt sense has not yet pinpointed exactly what it is about, although it knows generally what kind of issue is involved. In the example involving the motorist, constriction of my projected patterns of activity is clearly a major issue; I am prevented from moving ahead in a chosen ongoing direction. But at the point when the motorist is perceived as the object, the cognitive assessment is no closer to identifying the real object than is the precognitive felt sense, although a careful following through with an appropriate self-questioning process can draw out the intentionality that is implicitly present in the felt sense. It will do so not through perception, logical reasoning, or any other “cognitive” process (in the corticocentric sense that cognitivism would suggest), but by listening more carefully to the felt sense and observing what kinds of images and symbols tend to pull it up more intensely. Gendlin (1992a) suggests that when we do carefully direct our attention to the felt sense itself rather than the objective event that triggered it, we can then notice that certain imaginings tend to relax or satisfy the feeling in a more holistic way than others. For example, if the image of telling off my boss and quitting my job suddenly produces a shift in the felt sense of the anger, and if at this point the image of the motorist no longer pulls up the intensity of





Curious Emotions

it, I then know that it is not really directed at the motorist, but instead the motorist’s behavior was only a convenient symbol for pulling up the feeling. This can be recognized still more decisively by the fact that, as soon as I realize that the feeling is really more about my job than about the motorist, I no longer even feel very angry at the motorist. The image of the motorist no longer even presents itself as a primary object of the feeling, and this image no longer pulls up the feeling very intensely. When I form the image of his cutting me off, I no longer want to fight or kill him. The images that initially pull up a feeling are not usually the same ones that finally satisfy it in the sense of relaxing its tension. This way of getting at the implicit intentionality of feelings will be explored more fully as we proceed. Gendlin (1980) calls it “focusing”: we pull an affective quality into better focus so as to get at its meanings. How would I know to entertain this image of telling off my boss and quitting my job to begin with? Again, this can be inferred from careful observation of the quality of the felt sense itself. For example, I may notice that the point at which the feeling becomes most sharply intensified is when I remember how the motorist arrogantly ignored me when I tried to block his move. It is a short step from this observation to realize that thinking of anyone arrogantly ignoring my efforts in this way pulls up the same felt quality with the same intensity – and that the most important example of a person who seems to do this is my boss. But the image of the motorist cutting his car in front of me is more vivid and simple than any physical image of my boss, so it works better at pulling up the feeling and calling it to my attention. It may even turn out, in some instances, that what is really “arrogantly ignoring” me is not the individual, my boss, but a complicated abstraction – my entire profession, including the way it is organized, the way it prevents the changes I want to promote, by means of ignoring new proposals and intricate organizational methods that ensure that such new proposals will be very difficult to make themselves heard. In that case, it may be that what my preconscious felt sense of the situation needed was a simpler example consisting of a symbolizing vehicle to help me make conscious contact with the preconscious felt sense – the vivid physical image of a specific individual motorist cutting in front of me and arrogantly ignoring my attempts to block him, rather than the complicated and amorphous “system” or “situation” that is the real problem. Further exploration, then, may reveal that the boss, like the motorist, is only a symbolization vehicle for explicating the sense of a much more complicated environmental situation in relation to which self-organizational aims will need to work themselves out, in some one of many possible ways. As the intentional meaning of the felt sense develops, the affective response to the trig-

Chapter 1. Preconscious emotional intentionality

gering event can also change; now it is not the motorist’s insulting behavior that bothers me, but the fact that he was oblivious to my very existence. I was too insignificant to qualify even for a personal insult. So far, I have argued that identifying the object or the cause of an affective feeling is no simple matter. Ironically, it is the cognitivists who seem most strongly to agree with this point, insisting that a feeling per se does not identify its cause, i.e., that we must add the causal interpretation in thought after we have felt the feeling. In some ways, this viewpoint is in agreement with the selforganizational point that qualia do not pinpoint specific objects, but instead refer to general action affordances of the environment in relation to the organism’s holistic condition. But the reason the qualia cannot pinpoint the object is not because cognitive processes are better able to do so; it is because the real object of the emotion is not as specific to begin with as the cognitive processes initially take them to be. Cognitively, it may seem that the motorist who cut me off caused my anger; in reality, his behavior was only a convenient vehicle (no pun intended) through which to express an anger that was already present on an unconscious basis, an anger whose aims and objects have nothing to do with the motorist who cut me off. The emotion would have been just the same had the motorist not cut me off. It would have found some other occasion to express itself. The motorist’s behavior is being used as a symbol, just as an artist might use the image of a dog in a painting as a symbol of the idea of fidelity. The important point is that removing cognition from emotion does not leave mere interoceptive sensation. If emotion were merely interoceptive sensation plus cognition, then the “what it’s like” content or qualia of emotions would be fully given in interoception. What is missing in such accounts, as Watt (2000) points out, is precisely the valence that emotions attribute to a given situation, which again is seldom reducible to the mere interoception and proprioception of pain stimuli coming from the viscera or other parts of the body. Most of us know the difference between a negative emotion and a stomach ache. If emotional feelings were merely perception plus interoception, then it would be impossible to distinguish between an emotional feeling and the delivery of a perceptual-like signal from the periphery or the viscera of the body – i.e., proprioceptive or interoceptive sensation. It is true that emotions typically involve interoception, but they also involve much more than that. It is possible through certain types of anaesthetics and damage to nervous tissues (for example, the nociceptors) to dissociate proprioceptive or interoceptive pain sensations from their evaluative valence, so that we know we are in pain, and clearly receive the signal telling us that our hand is burning – without feeling the pain as consciously hurting (Newton 1989b; Damasio 1999). What is most





Curious Emotions

interesting about this phenomenon is that we will spontaneously withdraw the hand, even though the pain does not consciously “hurt.” So there is an intentionality there in the pain, even though the type of nociceptors that deliver the evaluative component of the pain sensation may be incapacitated. Before pain achieves a conscious status, we have already initiated actions to deal with it. Many might try to write off this preconscious intentionality as a kind of pseudo-intentionality, since it might be explainable as a merely mechanical and meaningless causal response. But we cannot escape the fact that, when I feel afraid of a snake, before I realize cognitively that what has been presented is a snake or even a threatening object, and before the object is even visually processed, the emotion is already taking actions to avoid the snake, very similar to the actions that will be continued even after the image of the snake has become conscious. What is revealed in the phenomenological reflection on the fear of the snake is not that a mechanical interaction with the snake is taking place, about which I now begin to feel a certain emotion. What is revealed is that the emotion was already being felt, and is now moving from a prereflective to a reflective status. In the process, the what-content of its intentionality may become more sophisticated, or simply different, but it does not acquire intentionality per se at this point. It only modifies a pre-existing felt sense and intentional content of the experience. Damasio makes this notion of prereflective emotion do a great deal of work in his theory of consciousness. How can an emotion already have intentionality prior to perception and cognition? As Jarvilehto (1997) points out, a situation’s action affordances can be assessed, at least to some extent, independently of intellectual processing of inputs, because the success or failure of our own attempts to act in certain ways – which is determined essentially through efferent nervous activity (signals sent outward to the body’s action systems) – can yield information about the environment, and this can be done even without processing of perceptual inputs, which requires afferent nervous activity (signals received from nervous areas earlier in a processing stream). Jarvilehto’s argument is that even without the afferent input, merely efferent activity, essentially by itself (with the help only of efferent copy, or feedback indicating that the body’s own moving parts have received the efferent commands), can in principle reveal information about the environment; that is, it can reveal its action affordances, which can even be used to construct a working map of the environment. Not much perceptual detail is needed to reveal very general contours of environmental obstructions or facilitations. Moreover, as the ERP (event related potential) data discussed earlier clearly show, the activity that tells the organism generally what kind of action it wants to do are more quickly accessible than any

Chapter 1. Preconscious emotional intentionality

perception. The resulting knowledge of action affordances is not precise or certain, and in actually existing organisms it is always supplemented by perceptual and cognitive assessments. But if emotion is self-organizational, then its qualities will be felt more efferently than afferently. This is why the knowledge that something is “off ” in the organism-environment relation is vague and murky. It reveals a holistic imbalance without specifying exactly what is causing the imbalance. Nonetheless, it has an amorphous sense of what general type of action is likely to be relevant. This remains true despite the fact that such vaguely and intuitively determined actions sometimes must be taken swiftly and decisively. The intentionality of a feeling is complex, murky, and holistically contextual. Correlatively, intentionality, whether conscious or preconscious, must involve more than simply reacting causally to a stimulus. Husserl (1913) emphasized this distinction between psychophysical causes of experiences and intentional objects of experiences, as do contemporary consciousness theorists such as Gillett and McMillan (2001). In the thinking of the classical phenomenologists (Husserl 1913; Merleau-Ponty 1941, 1942), intentionality is an “aboutness” relation, where the subject is the individual whose mental act is taking place, and the object is what this act is focused on; yet the “what” that is focused on here may not be the same as the psychophysical cause of the experience. They thus emphasize that we should not confuse the psychophysical cause either with what the experience is “about”or with the perceptual object at that moment. This is especially the case in the realm of the intentionality of emotion, where the psychophysical cause may be only a trigger for an emotion, and not at all what the emotion is really “about” – as in the example of the motorist who cuts me off. The intentional object(s) of a feeling may be distinguishable from both perceptual and imaginative objects, as well as from the physiological or environmental causes of the feeling. What a feeling is really about (its intentional object or objects) can be completely different from the physical workings of the brain (its causes), but also will not necessarily be the same as the perceptual and imaginative objects that are concurrent with the feeling. For example, a jogger with high blood pressure may experience feelings of agitation caused by physiological changes, while his perceptual intentional object at that moment is a barking dog. He may then form the imaginative intentional object of a dog that once bit him. Here we have both a psychophysical cause (high blood pressure) and an imaginative intentional object (the past dog) neither of which is to be confused with the perceptual dog that is present (the dog that is barking now). Neither the imaginative nor the perceptual object may play much of a role in the real intentionality of the emotion; it may be, for example, that the





Curious Emotions

jogger was thinking a few moments earlier of an argument with his wife that occurred weeks before, which contributed to the raising of his blood pressure, and primed him to look for an object for irritable emotional expression. The barking dog at that point may not even be a primary cause of the emotion, but only a long-sought opportunity to act out an emotion that was already present, but at a preconscious level. Generally, the aim of any emotion is distinguishable from the psychophysical events that trigger it, because the aim can be achieved by means of a variety of adjustments, and is more general than any specific action in relation to a specific object; the aim, again, is to maintain a self-organizational pattern of self-initiated and ongoing activity, with the specific object functioning only as a helpful or thwarting element that may be appropriated for the organism’s purposes, or may be difficult to appropriate for those purposes. The jogger in our example has his motivational reasons for not wanting to re-open the argument with his wife or dwell on his lingering anger at her: all things considered, the obnoxious little barking dog serves as a better symbolization vehicle to help concretize the feelings. To be sure, the image of the barking dog plays a role in helping to get the feelings “pulled up” into consciousness, but it is not primarily what they are about. A more mundane example of this principle can be found in the experience of abnormal subjects, such as in bipolar syndrome. It is a truism that the bipolar will find something in the environment to regard as the cause and intentional object of depression or elation – for example, the death of a favorite movie star. Someone may think that this dynamic characterizes only abnormal cases, and that “normal” people really can easily and unambiguously identify the objects of their emotions with the perceptual objects that are simultaneous with the feeling of the emotions. In my view, that would be an abnormal case. We all frequently use current objects as opportunities to explore and unfold continuing emotional processes. What is atypical is the rare case where a simple object, such as a snake, elicits a hardwired emotional response, such as fear. Even in that case, we observe that a “swamp-wise,” seasoned snake handler feels no fear around snakes, although he behaves cautiously. This is not to deny that certain situations are initially hardwired to elicit certain emotions. We come into the world primed to look for a smiling face and to be pleased when we see one (Baresi & Moore 1996; Meltzoff & Gopnik 1993), and we initially fear heights (Panksepp 1998). These “initialized settings” in our emotional constitution are very helpful in the care of the young, and thus evolutionarily adaptive. In fact, some specific conditions, such as extreme height, will always signal unmitigated danger, and only in the most abnormal cases will

Chapter 1. Preconscious emotional intentionality

those initialized settings ever be overwritten. But, generally speaking, it is most adaptive, in an intelligent and discriminative being, for the initial settings to be flexible. Not only learning, but also the natural patterns of maturation will change them in accordance with what works best for the changing patterns of organization of the developing organism. But again, the intentional meaning signaled by an emotion relates to an aim that in principle is always multiplyrealizable for a self-organizational structure, and it would actually be maladaptive for such a structure to become locked into a one-to-one correspondence between a particular stimulus and a particular response. It has been suggested that successful jazz musicians have a much better recovery rate from heroin addiction than the average addict (Hentoff 1973). Hentoff ’s explanation, consistent with Rollo May’s (1962) conceptualization of anxiety, is that a person whose addiction is only a small part of an ongoing life project, which is experienced as meaningful enough to be worth the hard recovery work, will have a better chance of getting through the necessary anxiety adequately. The temptation of even the intense pleasure of heroin is experienced differently, depending on its context in a holistic selforganizing movement toward larger aims and purposes that are strongly desired by the organism. Correlatively, still another advantage of the self-organizational approach, which allows the organism to act rather than just react, is that one kind of action consists of changing the pattern of functioning to one that is better. Complex self-organizing beings can change their overall structure – in conscious beings, this would mean personality change (discussed later in this book) – because part of what self organization means is that the system can rearrange the background conditions needed to facilitate a certain pattern’s happening; if so, then the system can also decide to rearrange the background conditions in such a way that a different pattern can occur, and then can seek out and appropriate the micro-level causal elements needed to make the new pattern occur. The way the self-organizing system changes its patterns is the same as the way it maintains an existing one – by actively appropriating the needed causal substrata to make it happen. Of course, this requires breaking the inertia of the existing pattern, but we humans do this all the time: Later I shall discuss a fundamental tendency in complex self-organizing systems toward “extropy,” or higherenergy basins of attraction, which differ from (and sometimes conflict with) consummatory drive-reduction, and which ground the exploratory, curiosity, nurturing, play, and social interaction tendencies in complex organisms, engaged in for their own sake and not just as means toward consummatory drive reduction.





Curious Emotions

An example of self-initiated pattern change is the “Fartlek” method of training for running, in which we alternate running fast and slow, gradually increasing the distance of the fast segments. Most amateur runners incline toward just the opposite method – to start with the same distance they have always run, and gradually increase the speed, but without much success, because the body naturally lapses into the old patterns of slower running, since the body remains in its old metabolic patterns. What works better is to increase the speed for a shorter distance, and then gradually increase the distance. The body is always willing to temporarily vary the way it functions, so we can take advantage of this willingness by getting the body into the habit of doing something it previously did only sporadically, and it learns to execute the more elementary capabilities (e.g., metabolizing sugar, delivering oxygen to the tissues) at different times and in different contexts from before. As a result, the body learns to shift into a new pattern of metabolism to support the new demands placed on it. After training in this way, the metabolism is generally increased. By the same token, we can train the body to change its patterns so that we no longer have nicotine cravings, no longer need the afternoon “high” from eating sugar, etc. This interconnects with a favorite point of moral philosophers, for example Aristotle, who emphasize the importance of habit – that we can apply the principles of psychological habit change to ourselves to change our personalities. This of course trumps Kant’s “ought implies can” in the ethical domain, because we can learn to do things that we previously could not do, if we re-train ourselves. Such self-retraining occurs primarily in humans because humans have more vivid imagination for non-existent states of affairs. We can imagine our own selves as enacting patterns that they do not now enact, and then calculate how to get ourselves into those new patterns. Thus we tend to hold humans morally responsible in spite of their predetermined psychological dispositions, whereas it would be absurd to hold lower animals morally responsible in this way. Our predetermination prepares us for the ability to change when required. More concretely, action imagery can consist of patterns of action in which we have not yet engaged, just as sensory imagery can depict sensory patterns that we have not yet seen. Having imagined a pattern of activity that would be better for one reason or another, we can then imagine what internal and external conditions would have to be rearranged to get us into the imagined action pattern. Dynamical systems make room for this kind of system change and not only for system maintenance. In self-organizing systems, one small change, if it occurs in the right place, can shift the whole system into a completely different pattern. In conscious beings, for example, the small component in question

Chapter 1. Preconscious emotional intentionality

could be a tacit presupposition that previously had been unexamined. In such cases, the trigger for system change can often be that some small component of the pattern turns out not to be any longer workable, and as a result the whole pattern goes into crisis and is finally perceived to be inadequate as a whole. In dynamical systems, the very removal of a component of the earlier pattern can automatically shift a system into a different pattern; if the resulting pattern then is not a very desirable one from the standpoint of “meta-emotional” evaluation, then humans are able to use their vivid capacity for imagination to find a better version of the new pattern. A complex dynamical system is flexible, and this flexibility extends to its ability to use objects to symbolize complex emotional situations. This is why it is so important to distinguish between the intentional object of a feeling and a perceived or imagined object that serves only as a symbolization vehicle – but at the same time we must remember that the true intentional emotional object of a feeling should not be confused with its physical cause. While the cause of an emotion may not be the same thing as what the emotion is about, or its intentional object, there is still a sense in which what makes an emotion “fully intentional” is that we correctly or incorrectly associate it with certain environmental components of its causation, as Salmela (2002) suggests. Damasio (1999) points out that all consciousness involves imagery of some sort (not just visual or auditory, but also proprioceptive, interoceptive, and sensorimotor imagery, and not only vivid but also muddled and truncated imagery). So it may well be that attributing a causal story to the emotion plays the role of explicating the emotion in the sense that it provides imagery that can pull up and intensify the felt sense of the emotion. Consider the case where the imagery that we entertain in explicating an emotion is known not to be its cause – for example, when a certain piece of music evokes feelings that we learn, as we explore them, are about serious life issues such as death and powerlessness. The music is not what causes us to have these despairing feelings, and we know this, yet the music serves to allow us a very usable medium through which to explore and explicate those feelings. In another case, I may have a certain feeling that I can “pull up” into full awareness by thinking of the Dutch boy trying to hold back the water with his finger in the hole in the dyke. The image only serves as an effective symbol to help pull up the feeling – it is not what the feeling is about. Even if it serves as a metaphor for what the feeling is about, this is only because I have made it serve in this way; other images could also have been used. In such cases, the image is known not to be the cause of the feeling, but just the reverse: I have chosen the image





Curious Emotions

because it worked well in pulling up and allowing me to focus attention on that felt sense. This last point is not meant to imply that the selection of the objects that trigger a felt sense is merely arbitrary. On the contrary, most images cannot work to trigger a given feeling. But neither does this imply that an image that initially succeeds in triggering a specific affective quality is simply what the affect is “about,” regardless of its total contextualization in the aim-directed life of the organism in relation to complex and variable environmental constraints. How, then, are specific objective triggering images relevant in the intentionality of emotion? And more generally, what is the role of imagery in making emotions conscious? The next section will explore these questions.

. The roles of sensation, interoception, and sensorimotor action imagery If a triggering image does not necessarily need to represent what an emotion is about, then what is the role of imagery in making emotions fully conscious? In my view, the key question here is whether the imagery involved is merely sensory imagery, associated with a causal story, combined with proprioceptive/interoceptive imagery, revealing the tingles, chills, and stomach churnings of the emotion’s felt quality, or whether on the contrary emotions also require sensorimotor action imagery. Whether Damasio is right or not in claiming that imagery is necessary for consciousness, it seems obvious from the critique of cognitivism that to have emotions requires more than just interoception, and more than sensory imagery, and more than a combination of the two. This would suggest that there is another type of imagery, beyond either interoception or sensation, that is important for emotion. This other kind of imagery is sensorimotor imagery. It differs from proprioceptive and interoceptive imagery in a way that is crucially important for our purposes. Whereas proprioception and interoception are characterized by a passive receiving of information from the periphery or from the viscera, sensorimotor imagery requires that the organism, acting out of its self-organizational aims, initiate an action command and deliver this action command to other parts of its nervous system (Jeannerod 1997). When the action command is inhibited, sensorimotor imagery results. We do not have to act in order to have action imagery. On the contrary, as Jeannerod shows, when we actually perform an action, we do not pay much if any conscious attention to the action imagery. It is when we inhibit the

Chapter 1. Preconscious emotional intentionality

motor command that we are most fully conscious of the action image as a mental image. While action imagery is clearly distinguishable from sensory or interoceptive imagery – for example, because it involves a cortically inhibited efferent signal, whereas sensation and interoception involve afferent signals – it can also be argued that for sensation, proprioception, or interoception to rise to the level of consciousness requires that action imagery already have been present, at least in the preconscious sense that an inhibited action command has been sent. For example, tracking of subcortically originated neurotransmitter activity shows that this activity must precede perceptual or cognitive consciousness. We have already seen that event related potentials (ERPs) and other brain activity measures show that emotional brain areas are activated earlier than perceptual processing areas even in perceptual and imaginative consciousness (Aftanas et al. 2001; Aurell 1989; Damasio et al. 2000; Federmeier & Kutas 2002). Moreover, the perceptual processing does not lead to consciousness of the object unless the emotional areas have already been activated, and unless they in turn activate brain areas associated with attention, such as frontal and anterior cingulate areas (Bernstein et al. 2000). And, as already noted, contemporary work on the cerebellum (Schmahmann 1997; Schmahmann et al. 2001) shows that the cerebellum, traditionally associated with motor routines, is a necessary way-station in this process, so that even conscious perception and cognitive skills depend on certain kinds of action commands being elaborated by the cerebellum, even though those action commands will be inhibited and quickly modified during the course of a perceptual experience. This notion that emotion and motivation ground all other forms of consciousness, including perceptual consciousness, is one of the most controversial claims of contemporary enactivism, and the entire next chapter will be devoted to supporting it. What I am arguing for here is an extension of Newton’s (1996) “action theory of intentionality,” where the understanding of a perceptual object involves imagining, at least at a preconscious and habituated level, how I could act in relation to the object in various ways. I understand that a baseball is moving toward my head, in the first analysis, because I can imagine catching it, hitting it, or ducking out of the way of it. Only later do I notice that it has seams and is green and yellow. To imagine green or yellow, in turn, means to imagine myself performing certain actions (in this case, actions of the anterior cingulate and frontal activity that constitutes paying attention to color phenomena, along with the appropriate vibrations of the retina and firing of the frontal eye fields). I may not know that I am performing all these actions, but I am doing so nonetheless. In artificial situations that restrain these actions,





Curious Emotions

such as the experimental paradigms of the Mack and Rock (1998) inattentional blindness studies mentioned earlier, we can see that failure to initiate these action commands results in lack of perceptual consciousness. Informally, we can notice the same effect when we fail to pay attention to a TV commercial, and thus do not consciously register it even though we are looking in the direction of the TV. In short, we must actively do much more than merely open our eyes in order to be conscious of visual information. This point will be explored much further in the next chapter. Emotional intentionality differs from the intentionality of other forms of experience in part because the roles of cause and effect are just the reverse of those other forms. Emotion is a barometer of the body’s dynamical activity, not just through interoception/proprioception, but through whether or not actions can be performed consistently with the patterns of activity in which the organism wants to engage. We know whether we are performing an action effortfully or effortlessly partly through proprioception (afferent feedback), but we also know this difference through the quality of sensorimotor imagery. Emotions deliver causal impact to other systems, and imagery comes about by means of the emotions’ directing attention to the imagery. It is a mistake to think of emotions as on the receiving end of the inputs of an information processing machine. They are on the sending end of the information, and what they send partly determines the feedback that we get from the external input sources; they propose, and the external input sources dispose, not the other way around. At the same time, I shall argue that there must be sensory or interoceptive imagery (in addition to sensorimotor action commands) in order for an emotion to become fully conscious. The reason is that we often can act without any consciousness intervening at all. So in some way it seems that sensory, proprioceptive, and interoceptive imagery are necessary for the full consciousness of emotion, yet the emotion itself tends to select and direct attention toward this imagery, based on its own action commands. Emotions are already ongoing, at least at a preconscious level, before the imagery based on the input from the triggering environmental event is processed. The emotions result from the ongoing activity of a dynamical system, and this activity is not caused by inputs, but only disturbed by them. When we are not aware of what is disturbing our emotional activity, or have no imagistic hypotheses about it, our emotion remains at a preconscious level. But when we entertain relevant imagery, this allows us to pull up the emotion into full awareness. This claim that emotion directs and is presupposed by other conscious states is, as I have said, a controversial one. Let’s proceed, then, to a full defense

Chapter 1. Preconscious emotional intentionality

of it, which is presented in the next chapter. Then in subsequent chapters we can explore the way a dynamical systems analysis can affect our understanding of the workings of specific emotional systems.



Chapter 2

Motivated attention in action How emotion creates conscious intentionality

The enactive approach begins with the thesis that affective intentionality must be understood as grounded in the system’s motivations to act, which can be as complex as the dynamical patterns they are meant to maintain. If this thesis is true, it follows that emotions are not merely responses to given stimuli, but actively seek out usable environmental affordances; thus the same emotion can be realized in relation to a wide variety of alternative environmental conditions. Moreover, if the needs of a complex dynamical system determine the aims of the organism, the emotions will monitor how well the total organism is doing in relation to its entire environmental situation, including aspects that are not immediately present. This means that the intentional objects of emotions are not as specific as those of other conscious states, such as perception, imagination, and logical thought – and that the intentionality of emotion is murky, open-ended and ambiguous because the aims of emotions are holistic and multiply realizable. This would be a somewhat unsatisfying state of affairs for those who wish to study specific emotional processes unless we can establish a framework for the ways in which emotions and their intentional objects interrelate and interact. This chapter will argue in effect that the relation of an affect to its intentional “aboutness” is almost the reverse of what the stimulus-response and information-processing models might suggest. Emotions are not responses to stimuli, but instead are ongoing, holistically motivated processes that attempt to use environmental affordances to further their self-organizational aims, which are built up in complex ways from the various needs that arise for living organisms in maintaining their relatively stable patterns of activity. Affective processes select environmental objects because of their possible affordances, and then process them to greater or lesser extents as determined by the continuing analysis of their utility valences. This implies, ultimately, that motivation determines the focus of attention and must be in play in order for those attended contents to come into sharp consciousness. This is a controversial issue, but it is necessary to confront it



Curious Emotions

in order to understand the relationship between emotion and its intentional objects in the right way. I shall argue here that the entire brain is organized around the principle that motivation drives all other conscious processes rather than merely responding to their resulting contents. In fact, I want to argue that the essential difference between conscious and non-conscious information processing is that the former necessarily is “enactive,” whereas non-conscious information processing may often be largely “reactive.” And this, in conjunction with certain considerations about the way the brain works, implies the even more radical thesis that there can be no intentional consciousness without motivation. Let’s begin with the perceptual contents of consciousness. We saw earlier that Rolls insists that we must first process perceptual signals, at least to some extent, before we can respond emotionally to them, and that this early processing is not dependent on the organism’s motivational state. However, the Mack and Rock (1998) extensive studies of inattentional blindness suggest that there is no conscious perception unless attention is first directed toward a stimulus in order to see what is there. It is not merely that perceptual stimulation can cause attention, but just the reverse seems to be true: if an attentional act is not first executed, perceptual contents do not attain the status of consciousness, and we are not aware of what we have seen – nor is it processed as fully as it would be if we were conscious of it. Failure to see objects within the visual field because attention is otherwise occupied, or “inattentional blindness,” can therefore be a rich source for rethinking the relationships between attentional mechanisms and motivational processes, and can also be relevant to the most basic aspects of the attempts to develop theories of consciousness in general – i.e., to understand the differences between conscious and non-conscious information processing. The Mack and Rock inattentional blindness findings are highly consistent with theories in which the property of consciousness arises from endogenously generated, motivationally directed “actions,” defined in contrast to mere “reactions” to the corresponding information processing. Correlatively, the inattentional blindness findings are less reconcilable with other types of theories of consciousness, which I shall generically term “reactive” theories. When already involved in a visual task requiring attention, subjects fail to see briefly presented visual elements extraneous to the attentional task in contexts where they normally would be seen. In most of these cases (and this is an oft-neglected fact in discussions of inattentional blindness), Mack and Rock cannot even detect any better-than-chance “implicit perception” knowledge about the unseen object. The main exception to this rule involves certain

Chapter 2. Motivated attention in action

trials in which meaningful words are presented as the unattended stimulus; in those cases, but seemingly only in those cases, subjects have better than chance knowledge of the object presented although they do not consciously see it. Of course, the failure to detect implicit perception in most of the experimental contexts reported does not prove that there is none, but the fact that Mack and Rock explicitly look for an implicit perception effect in such contexts and yet fail to find it cannot be overlooked. The inattentional blindness effect is consistent across a variety of different research designs reported by Mack and Rock. In a typical experimental design, the subjects are asked to determine which arm of a cross is longer than the other, and then while they are involved in this task a small image is presented in various places within or near this zone of attention, for a short time in the neighborhood of 200–250 ms. In many of the experimental paradigms, most subjects fail to see the extraneous image, even though control subjects (whose attention is not otherwise occupied) consistently see it quite clearly. Except for the trials involving meaningful words as priming stimuli (which will be discussed below), “when asked to pick out the critical stimulus from an array of alternatives, their performance did not differ from chance” (Mack & Rock 1998: 13, italics added). I.e., subjects exhibiting inattentional blindness in most instances did not even show implicit knowledge of the unseen stimulus, at least as far as Mack and Rock could detect. Again, this does not necessarily mean that there is no implicit perception, but the explicit failure to detect it must be taken into account. One might ask whether the statement that perceptual consciousness requires attention is a mere tautology, since “in this book, the term attention is used to refer to the process that brings a stimulus into consciousness” (Mack & Rock 1998: 25). But what is being said is actually quite substantive and surprising: A fuller definition of attention is that “to look and try to find something is to attend to the array in which it might be present and to intend to see it” (Mack & Rock 1998: 5). Attention is an act that can lead to perception rather than just following from it. Thus “attention, when otherwise engaged, must be captured before perception can occur” (Mack & Rock 1998: 18, italics added). The surprising finding, then, is that we must always pay attention to something before we can consciously see it. The attention is a precondition for the conscious perception rather than merely a result of it. This does not mean that our attention cannot be involuntarily attracted to something; but there seem to be surprisingly few types of objects that can attract attention when otherwise occupied – for example, an extremely large object, a fairly large moving object, or a cartoon of a smiling face. Interestingly,





Curious Emotions

neither a similar frowning or neutral face, nor a small moving object attracts the subject’s attention when otherwise occupied. This would seem to imply that pre-perceptual categories and mechanisms are involved in controlling the direction of attention, no matter how seemingly passive it is. These findings are intriguingly consistent with an enactive theory of consciousness, in which emotional activation of subcortical and anterior brain areas is necessary before conscious information processing can occur. According to enactivists, when the motivation to pay attention is missing, there can be virtually complete occipital activation with no perceptual consciousness; the organism must motivatedly search its environment (as reflected in anterior and subcortical brain activity) before visual stimulation of occipital areas can lead to consciousness (Aftanas et al. 2001; Aurell 1984, 1989; Bernstein et al. 2000; Panksepp 1998; Watt 1998, 2000). On the enactive view, the element of consciousness, above and beyond mere nonconscious data processing, is contributed by brain areas associated with the organism’s motivational interest in processing certain kinds of environmental stimuli while ignoring others. The idea that attention is motivated, of course, is not new in psychology, nor is the view that limbic and subcortical emotional processes play a role in directing attention. What is uniquely a prediction of enactive theories is that no consciousness can occur without motivated attention – that cortical activity produces only nonconscious information processing which can become conscious only in the case where emotional processes lead to a motivated attentive act. Enactivism is thus in contrast to the thinking of many investigators, who assume that the consciousness of higher cognitive contents, such as perception and logical thinking, is subserved primarily by cortical brain areas, while affective consciousness is subserved by subcortical and limbic activity, and that motivational activity is not a precondition for conscious perception and thought, except in the somewhat trivial sense that the needed neurotransmitters and energy sources must be released into the cortex to produce arousal. On this view, the subcortex functions like a power switch for the cortex, and as long as the power is supplied to the cortex in one way or another (for example, it might be possible to experimentally deliver it by artificial means), no further subcortical activity would be necessary for consciousness. This would constitute an extremely modular approach, where different brain areas subserve different forms of consciousness. An extreme modularity of perceptual, intellectual and affective states of consciousness would thus contradict the enactivist account of the effects of subcortical and limbic functions on all forms of consciousness, including conscious perception and attention. The modu-

Chapter 2. Motivated attention in action

lar cognitivist model also overlooks evolutionary trends that would tend to set up a critical relationship between higher and lower systems, where the laterdeveloped systems would be an incremental addition to and modification of the earlier. Without an incremental evolutionary perspective, the brain may not be seen as a fully integrated system. Some authors explicitly hold an extreme version of this modular view with respect to the relations between emotional and perceptual processes (Gazzaniga 2000; Rolls 1999; LeDoux 1996), while others believe there is not yet enough evidence to decide whether the reactivist or the enactivist approach to cognitive and perceptual consciousness is warranted, arguing that this remains an open question (Faw 2000). Some go so far as to argue that not only are motivational brain processes not essential to consciousness, but just the reverse – that it is the cortical cognitive processes that supply the element of consciousness over and above the subcortical functions, which themselves are unconscious until the cortex plays its role (Jackendoff 1996; LeDoux 1996). This would be the extreme antithesis of the enactive view, if we assume that the subcortical and limbic systems are important monitors of the body’s selforganizational tendencies. We shall see increasingly throughout this book that self-organization is at the heart of the distinction between enactive and reactive mechanisms. An enactivist theory of the relationship between consciousness and the brain predicts the Mack and Rock findings, since enactivism implies that motivated attention, either voluntary or involuntary, is needed for consciousness. One reason such theories have been slow in gaining acceptance has been an absence of conclusive empirical evidence of the kind that Mack and Rock present. Contrary to enactive theories, it is often assumed that perceptual consciousness is passively caused, primarily in the cortex (and in corticothalamic loops) by input from the perceptual object rather than being caused by the organism’s own self-regulated activity, which enactivists believe determines the motivation for the direction of attention. The reactivist position would view the substrates of perceptual consciousness as the same brain processes that subserve the corresponding perceptual information processing – primarily cortical areas looping with the thalamus in response to perceptual input. While most investigators of the thalamus grant that the “thalamocortical loops” can do nothing absent brainstem influences, they see the brainstem role in “higher” cognitive, perceptual, and conscious functions primarily as one of arousing and activating the brain areas involved in producing these forms of consciousness. It is true, of course, that there can be selection and extensive processing of a food stimulus, for example, without this selection’s being driven at an early





Curious Emotions

processing stage by hunger, and that we can be conscious of the food stimulus in the absence of hunger (Rolls 1999). But this does not mean that motivational processes more broadly construed do not affect selection for conscious attention at an early stage, especially if we view Panksepp’s “seeking” system in the brain as a motivational system that can direct attention toward interesting, meaningful, or potentially useful objects. This point will be discussed further as we proceed. The possible connection between inattentional blindness and enactivism is suggestive because, if attention must be engaged before contents can register in consciousness, then there may well be a subcortical and limbic mechanism that drives attention independently of the processing of the input, and such a mechanism could well be a necessary precondition for consciousness, always resulting at least partly from an endogenous aspect of organismic functioning, connected with the regulatory actions of emotional brain areas. This is highly consistent with enactive theories of consciousness and, as we shall see in a moment, it is quite at odds with a reactive and extremely modular view of the various conscious and cognitive functions. The next section will define some key concepts for clarifying how the predictive implications of enactive theories are in sharp contrast to the predictions that would follow from other theories. Then later in the chapter, we can examine in some detail how these predictions of enactive theories are borne out par excellence by the inattentional blindness phenomenon as investigated by Mack and Rock, in combination with certain well established neurophysiological findings.

.

Linear versus dynamical causal sequences in the brain

Even if the Mack and Rock data are consistent with an enactive theory of consciousness, the more interesting question is whether these data would necessarily be inconsistent with other types of theories. To answer this question, we need to work from a clear definition of what is meant by “enactive” theories, such that it can be inferred exactly what empirical predictions follow from enactive theories, and moreover whether these predictions are positively inconsistent with the predictions of other theories (which we can simply call “reactive” theories). And this will require a somewhat specific explanation of the differences between the causal structures of enactive and reactive theories, as well as a clear physiological fleshing out of the kinds of brain mechanisms

Chapter 2. Motivated attention in action

that would reflect these different causal accounts, even though many of the details of these mechanisms await further neurophysiological research. When I use the term “enactive,” I do not mean to commit myself to all of the specific baggage of Varela et al. (1993), Humphrey (2000), Clark (1997), or any other particular enactive theorist. My purpose is to capture the sense in which certain theories posit that consciousness must result only from “actions” of an organism, and cannot result from mere “reactions.” This distinction could get us enmeshed in thorny philosophical discussions of the nature of causation, but the simplest way to ground the distinction is in terms of the causal structure of self-organizing systems. “Actions” in this sense are performed only by self-organizing entities, and serve the purpose, either directly or indirectly, of preserving the organism’s pattern of self-organization. Earlier we defined “self-organizing” patterns to mean patterns for which a system has a strong tendency to seek out and find causal substrata, so that if one set of environmental conditions is not conducive to that particular pattern of activity, the system is likely to seek out some other set of conditions that is conducive to it. A self-organizing system initiates actions geared toward using the environment as needed to keep the system going in its definitive patterns. For example, animals seek environmental elements needed to maintain homeostasis. Reactions, by contrast, are explainable without any reference to selforganizational structure, even if the reaction is taking place in a self-organizing entity. For example, a knee-jerk reflex is a reaction, because, even though it takes place in a self-organizing entity, the entity does not initiate the action. To be sure, the reflex also may be used by the organism for its generally selforganizational purposes, but the reflex itself is merely a reaction, not an action in the sense intended here. Actions are behaviors that require reference to the entity’s self-organizing structure for their causal explanation, such as in the regulation of bodily temperature, where many readjustments of many systems will be accomplished if necessary to achieve the aim. It will be useful for our purposes here to distinguish between a selforganizing causal structure and a “purely linear” causal structure. By a linear causal sequence, I mean one in which a certain antecedent is both necessary and sufficient for its consequent under certain stipulated background conditions, as in flipping a switch. It is true, as we shall see in a moment, that self-organizational causal structures are composed of many linear ones; but not all complex systems of linear causal sequences result in self-organizational structures. What precisely, then, is required for a self-organizational causal structure?





Curious Emotions

A process is explainable “purely” in terms of a certain combination of linear causal sequences if it does not need to posit any principle other than those to explain why the process has the properties that it has. The knee-jerk reaction would be an obvious example of such a purely linear causal reaction. By a “selforganizational” causal structure, by contrast, I mean one in which, if a certain linear causal sequence is not available to support the continued existence of the system’s definitive patterns of activity, then there is a strong tendency for the system to readjust some of its functions so that alternative causal mechanisms can be used for that purpose. A light switch is not a self-organizing mechanism because, if the appropriate switch is not flipped, the light cannot alter any of its other activities in order to ensure that some alternative mechanism will allow it to come on. Even a light with two alternative switches is not self-organizing in this sense, because, if neither switch is flipped, the system can neither cause a switch to be flipped by means of a readjustment of the light’s other activities, nor can the light rearrange its other activities to find another way to come on. Even a thermostat, as normally constructed in heating and cooling systems, is not self-organizing in this sense: While it is true that the system’s reaction to a given temperature will tend to feed back and maintain the same temperature, the overall structure of the system is not determined by a need to alter any of its activities in order to ensure continuity; the same preconditions are always required for a turning on (or turning off) of the circuit, and the definitive organization of the thermostat is guaranteed without its having to find an alternative way to turn the heat on or off in order to maintain that. By contrast, a living organism’s maintenance of a constant temperature is self-organizing, because when the temperature gets too hot or too cold, the organism must systematically readjust many other activities throughout the system in order to maintain the temperature, and if it does not do so, the survival of the definitive structure of the system as a whole is threatened. (Obviously, it is possible in principle to construct a self-organizing thermostat, because many living organisms are, among other things, thermostats. But thermostats as normally constructed are not self-organizing; it is not necessary for them to be self-organizing in order to produce the results they do.) It is not necessary for our purposes here to decide whether the selforganizational vs. “purely linear” distinction is a sharp distinction or can be one of degrees. One could perhaps think of borderline cases, as suggested by Kauffman (1993). These definitions also do not presuppose that a selforganizational system is not completely made up of linear mechanisms; but how the linear mechanisms fit together determines whether the system is selforganizing or not. If it is not self-organizing, we can speak of it as being “purely

Chapter 2. Motivated attention in action

linear” – i.e., the explanation of the properties of the system is exhausted by citing the linear sequences that make it up. In a self-organizing system, by contrast, not only does each causal outcome require antecedents that are necessary and sufficient for it under the given background conditions (i.e., linear causal sequences), but the system can also change background conditions as needed to find alternative antecedents for a needed outcome if this is necessary for the continuation of the definitive patterns that make up the entity. Any metabolic system in living organisms involves multiple shunt mechanisms that are classic examples of this kind of self-organization (Monod 1971; Kauffman 1993). While living organisms are classic examples of self-organizing systems, they are not the only examples. Ecosystems, social systems, and even crystal formations can be self-organizing. It is not necessary for our purposes here to go into the differences between living and nonliving self-organizing systems, although this topic has been discussed at length in many contexts (Kauffman 1993; Ellis 2000d; Ellis & Newton 2000a). The controversial question for our purposes is not so much whether such self-organizing systems exist, but rather whether consciousness, and specifically conscious attention, is dependent on the real-time action of a self-organizational mechanism, as opposed to resulting from a purely linear mechanism already in place within the self-organizing biological organism. Obviously, there is nothing about the notion of a purely linear causal sequence that prevents its outcome from requiring the participation of two or more separate events as its causal antecedents. For example, if both A and B are needed to cause C, the sequence “(A&B) → C” (where “→” indicates causation) is still linear, as long as C is not self-organizing. There can also be loops and shunt mechanisms in purely linear sequences. For example, consider the following causal sequence A → (B or C), B → D, C → E, (D or E) → F

There is nothing about this causal sequence that would preclude its being a purely linear one. ‘A → C → E’ is one line of causation, ‘A → B → D’ is another line, and ‘D or E → F’ is the last step in the linear chain of causation. There can also be loops in which causal outcomes feed back to enhance the antecedents for furthering the same type of reaction, and in which the enhancement of a signal can feed back to cause still more antecedents for still further reactions of the same kind – i.e., circular causation in which A → B → A → B etc.





Curious Emotions

This is still a purely linear causal sequence, in the sense we are using, because each step presupposes only the prior step as a necessary and sufficient antecedent, given the background conditions under which the causal relations obtain, and none of the steps necessarily must have a self-organizing structure in order for this type of loop to occur. For example, the fact that there are corticothalamic loops in the processing of sensory information prior to consciousness does not decide between an enactive and a reactive theory of consciousness. The corticothalamic loops may be explainable in purely linear terms, involving loops of linear reactions, without any reference to the self-organizing structure of the system. But according to an enactive theory, to cite purely linear causal reactions needed for the processing of perceptual input would not be enough to explain why the information achieves a conscious status without additional reference to the self-organizing structure of the system – for example by positing that the emotions, which flow from the organism’s self-organizing needs, must be part of the causal input into the determination of the pattern of information processing in any instance where that information is to achieve a conscious status (especially if a generalized “curiosity,” triggered more by certain classes of inputs than others but endogenous in origin, is a motivating emotional process, as Panksepp 1998: Chapter 8, believes it is). The enactive type of theory does not deny linear explanations of elements of a system, but rather embeds them into the larger self-organizational structure, which will use different combinations of linear mechanisms to maintain its overall pattern by routinely rearranging the background conditions necessary for a given linear causal sequence to obtain. If (as enactive theories posit) consciousness must always be an action (in the sense opposed to mere reaction), then each state of consciousness will be preceded by some preconscious self-organizing activity on the part of the organism. Given the purposeful nature of self-organizing processes in living systems posited by enactive theory, this self-organizing activity will attempt to make the resulting conscious state a useful one for the organism’s purposes, and thus will exert a selective effect. Specifically with regard to attention, an enactive theory will predict that self-organizing processes – which connect to cognitive function via subcortical emotional brain areas – will influence the direction of attention prior to conscious awareness of the attended object by means of a motivated direction of attention. As I have already mentioned, the notion of motivated selection for attention is by no means a new or radical idea; what is a specific claim of enactive theories is that no consciousness can occur without motivated attention.

Chapter 2. Motivated attention in action

A completely reactive theory, at the other extreme, will posit that conscious states, at least in principle, could exist outside the context of a self-organizing system, and without the actions of such a system (again in the sense opposed to mere reaction) influencing the quality, properties, content, or attentional focus of the conscious states in question; moreover, such self-organizational actions would not be needed in order for such states to be conscious. Since all mechanisms in a completely reactive theory are explainable by means of the interaction of the mechanical system’s hardwiring and the inputs it receives, it should be entirely consistent with reactivism that consciousness could occur as a reaction to the stimulus inputs without any motivatedly selective attention.

. Conflicting theories with conflicting empirical predictions The differences between the predictions of the two types of theories are complicated by the fact that, even though reactive theories do not require that consciousness be a property of self-organizing systems, they do not preclude this either. In fact, living organisms contain both active and reactive components, and it is quite natural to assume that, if cognitive functions require complexity, then the most complex interconnecting sets of mechanisms – mammalian brains with their billions of interconnected cells – would be the best candidate for a naturally occurring information processing machine, regardless of whether such processing resulted from the enactive or from the reactive aspects of the organism. It might be held, for example, that conscious information processing results from a complicated series of purely linear mechanisms wired up in a certain sequence, like an electronic motherboard. In fact, the causal structures of the two types of theories may at first seem deceptively similar. In the reactive model, I (input) causes R (the response that subserves consciousness), and may presuppose P (some physiological preconditions for R). That is, Given P, I → R

In the enactive model, P (rather than I) causes R, while the nature of R (i.e., some of its quality and content) is influenced in some ways by I. That is, Given I, P → R

The difference between these two causal accounts may seem subtle or even nebulous. But the kind of “P” at work in the first formula does not completely overlap with the “P” in the second. In the first formula, which describes the





Curious Emotions

reactive causal story, the “P” that is presupposed is much more cortical than in the second one, which reflects the enactive story. In the enactive formula, the “P” is a system of subcortical-cerebellar-limbic loops that are always in the process of gearing the organism up to act upon its environment, and selectively seeking out inputs that seem potentially relevant to those general action proclivities. If a sudden environmental change is relevant in an unexpected way, the motivational system must quickly readjust to gear up for a completely different kind of action, and thus must look for other perceptual cues than the ones it was initially looking for – leading to a shift in early selection for attention. But whether the stimulus fits what was anticipated or is radically different, the system that directs attention in potentially useful directions is largely activated before the stimulus is perceptually processed. So, in spite of a superficial similarity, the two types of theories require a significantly different causal structure to bring about consciousness. In enactive theories, consciousness can result only from an organism’s motivated action, whereas in reactive theories the fact that consciousness occurs in selforganizing creatures results only from the fact that such creatures are complex enough to subserve consciousness. On the enactive view, perceptual consciousness, like any other form of consciousness, is an action that the organism executes, not merely a chain reaction that it undergoes. Not even a chain reaction that includes feedback loops can be conscious unless it results from prior self-organizing actions. This essential difference gives rise to major differences in the empirical predictions for the two types of theories – predictions for which, as we shall see, the Mack and Rock work is extremely relevant. For the enactive type of theory, if an instance of information processing X can occur either consciously (XC) or unconsciously (XU), the difference between XC and XU cannot be explained purely in terms of the extent or intensity with which X is processed in the chain reaction components of the system – in the occipital lobe, for example. On the contrary, there must be some self-organizational process P1 which is present in the case of XC, and absent in the case of XU. One might express this by saying XU + P1 = XC

I.e., the difference between XU and XC is not merely that the gain of some physiological process P is turned up in the case of XC. Instead, there are unique physiological processes P1, in the case of XC, which are lacking in the case of XU. This is not to deny that there will also be some self-organizing preconditions, P2, for XU itself; nonetheless, there will be some additional self-organizing process, P1, that is necessary for XC, but not for XU. A reac-

Chapter 2. Motivated attention in action

tive theory would not predict this. On the contrary, in a reactive theory, the difference between XC and XU is not dependent on any difference between P1 and P2, but rather on a difference in the way X is processed in the two instances. In more concrete terms, according to an enactive approach, subcortical and limbic emotional processes, which are self-organizing, will necessarily be different in the cases of XU and XC, and one such set of differences is likely to be the subcortically orchestrated motivation to direct conscious attention in ways useful for the organism’s purposes. To the extent that attention is an organismically self-organizing action, it will determine a difference between a perceptual consciousness XC and a corresponding unconscious processing, XU. So an enactive account should predict that an input can cause occipital processing in the absence of these motivational mechanisms, but will not cause consciousness of the perceptual stimulus – i.e., will not cause XC. A fuller fleshing out of this prediction of enactive theories is that the manner and extent of the processing of X will also be partly determined by selforganizing actions of the brain’s emotional system in accordance with overall organismic purposes. Thus the same P1 that determines the difference between XC and XU will also serve as an early gating mechanism that will tend to favor potentially useful or emotionally interesting information for more extensive processing. This means that emotional brain processes will exercise an early selection function for incoming stimuli, while those same emotional brain processes, at a later point in their activity, also will determine whether the information that is processed acquires a conscious status. In effect, whether X will ever attain a conscious status will already have been largely determined at an early stage of processing. Extremely reactive theories of consciousness would predict just the opposite – that whether X is selected for conscious attention is determined only very late in processing, after a great deal of primary perceptual cortex processing and encoding has already occurred. Paradoxically, the Mack and Rock implicit perception findings themselves suggest that there is evidence for very late selection, at least for some types of stimuli (meaningful words). That subjects show an implicit knowledge of stimuli that never reached conscious attention seems to suggest that much processing has already occurred prior to attention. We shall return to this issue. But first, I should point out that all of this can be put more in experimental terms more familiar to the way the neuroscience literature would address the same problem. Rolls (1999), as mentioned earlier, shows that a specific motivation process such as hunger exerts only a late selection effect on the perceptual processing of food stimuli. But this is different from showing that no motivational processes





Curious Emotions

exert an early selection effect here. Obviously, physiological arousal is a motivational process that affects early processing, and as we saw earlier, Panksepp (1998) shows how arousal is affected by a general emotional “seeking” system in the brain. Arousal, of course, is not selective attention, but the seeking system does have its own selective tendencies with regard to environmental affordances. So the question is not whether brainstem emotional systems exert an early selection function or not. The question is whether it would be possible for there to be conscious attention to a perceptual object without such an early selection effect, and whether the early selection is merely a generalized arousal, or on the contrary plays a role in determining which objects are allowed into consciousness. On the enactive view, both early and late selection mechanisms will be operative, and both will be controlled by the same subcortical and limbic emotional processes. On the reactive view, by contrast, selection for conscious attention will be late, and this late selection will be determined by mechanisms that are primarily cortical (involving also corticothalamic loops), and which occur not as a prerequisite for the extensive processing of X, but as a consequence of it. Here again, reactive theories can grant that this cortical functioning might presuppose some self-organizational subcortical activity, such as a general cortical arousal state that is needed for either XC or XU; but the cortical process does not presuppose any special subcortical processs that determine the difference between XC and XU, and which thus would also exert an early selection effect. Early selection processes would include subcortical motivational aspects according to enactivism, whereas reactivism would reject this assumption, insisting that the same brain areas that process X are the ones that subserve the consciousness of X. One may wonder whether the category of “reactive theory” has now become an empty category or a straw man. Can the difference between P1 (the self-organizational preconditions for the conscious information processing) and P2 (the self-organizational preconditions for the corresponding unconscious information processing) really be that important? Remember that something counts as a P1 precisely if it is a prerequisite for XC (the consciousness of X) without being a prerequisite for XU (the unconscious processing of information X). If a P1 is subcortical and exerts an early selection effect on conscious attention, this would mitigate against an extremely modular view of the brain substrates of different kinds of conscious states. We have already seen examples of prominent neurophysiologists who take widely divergent views on this “modularity” problem, some clearly falling within the reactivist camp, others leaving the question open, and still others

Chapter 2. Motivated attention in action

taking an enactivist position. Although enactive approaches are currently becoming more fashionable, it would still be a mistake to think that the category of “reactive” theories has become an empty category or a straw man. There are two respects in which the distinction is still crucial: 1. Some theories explicitly endorse reactivism by insisting that consciousness in principle could be attributed to a suitably constructed digital or connectionist computer with the right kinds of feedback loops. Almost any AI system that attempts to build consciousness into a system that is not biological will be a reactivist system by the definition we are using here. (The single exception would be an AI system that uses components that are not “biological” in the sense of being composed of carbon, hydrogen, oxygen and nitrogen, but nonetheless are biological in the sense that they are organized in the way that biological systems are, with self-maintaining and self-organizing structures that act rather than merely reacting; for example, see Bickhard 2000.) There are also still many theories of consciousness that content themselves with feedback loops as a precondition of consciousness, rather than requiring that consciousness presupposes self-organization (for example, see Dennett 1991). A feedback loop can be a simple linear system, and need not be self-organizing. Moreover, there are theorists, like Jackendoff (1996), who insist that consciousness only passively reads off the information processing that has already occurred at an unconscious level by means of discrete transmissions of information that work in the same way that an artificial computing system would work. 2. Perhaps more prominently, there are many theories that, while not explicitly defending a reactivist position, do want to remain neutral as to whether information is processed reactively or enactively in conscious functions. Besides those who explicitly criticize or reject enactivist thinking (for example, Boyce 1998; Nunn 2001), there are those who formulate theories of consciousness in such a way as to allow consciousness equally well in reactive or enactive systems. This would include many of the most prominent theorists of consciousness, such as Dennett (1991) and Pylyshyn (1973). Even Searle did not take his enactivist turn until very recently (see Searle 2000). So part of my point here is to convince even such “agnostic” theorists that they no longer seriously need to hold open the possibility that a purely reactive system could have the property of consciousness. One might expect that at least theorists of emotion would subscribe to an enactivist position, since the emotions are a main link between the holistic homeostasis of the body and the workings of cognition. But even emotion





Curious Emotions

theories are often formulated as if emotional responses could be purely reactive rather than originally initiated by the active expression of the organism’s endogenous self-organizational purposes. A classic example is the difference between the way LeDoux (1996) and Panksepp (1998) understand the neurophysiological mechanisms underlying anger and fear. For Panksepp, anger is clearly enactive; an organism becomes angry because something gets in the way of its self-initiated patterns of activity; the binding of an infant’s limbs produces the purest possible symptoms of anger. By contrast, LeDoux speaks of the anger response as if it could be passive, i.e., caused by a purely linear series of channels through which certain inputs resulting from certain types of stimuli finally cause certain chemicals and electrical transmissions to be delivered to the amygdala. While Panksepp agrees that these chemicals and electrical effects are delivered pretty much as LeDoux claims (with modest differences not relevant to our purposes here), Panksepp does not believe that the immediate reactive response is the whole story; the delivery of the electro-chemical effect yields a given affective response only if embedded in an elaborate network of independent neurophysiological contextualizations: Consider the example of a man burglarizing a house who suddenly hears a noise outside: the same stimulus that triggers fear in such a context may not produce fear in the same subject while visiting that same house as a guest. The organism’s ongoing self-initiated activity is a precondition for stimuli to produce the affective responses they do. While Panksepp’s approach is highly consistent with enactivism, and largely inconsistent with reactivism, LeDoux’s approach is at least as compatible with reactivism as with enactivism, and in fact encourages researchers to emphasize a focus on the reactive aspects of emotional responses and to de-emphasize focusing on the enactive variables that also determine our responses. Part of the reason for this difference is that LeDoux takes a more modular view while Panksepp takes a more embedded view of the functions of these structures and connectivities. As a result, LeDoux treats the amygdala as the substrate of anger, whereas Panksepp insists that the PAG – an action-initiating brain area – actually plays a greater role in anger feelings, and that the role of the amygdala is to learn which stimuli to associate with those feelings. Thus Bickhard (2000) is not addressing straw men when he argues that it makes a difference whether we think of stimuli as causing a response where there was no response, or whether we think of them as modulating or modifying a response that was already taking place in the organism, as caused by its own endogenous self-organizational tendencies.

Chapter 2. Motivated attention in action

. The P300 ERP as an operational definition of perceptual consciousness A standard way of operationalizing the point at which perceptual information attains a conscious status is to look for a P300 parietal event-related potential (ERP) following presentation of a novel stimulus (for example, see Aurell 1989; Bachmann 2000; Damasio et al. 2000; Federmeier & Kutas 2002; and others to be cited below). As we saw earlier, there can be virtually complete occipital processing, but until and unless the P300 parietal ERP occurs, the subject is not conscious of seeing the presented stimulus. So a good way to operationally define the question as to whether enactive processes affect the structure of information processing prior to the point at which they become conscious would be to determine whether they affect the structure of processing in the brain activities that occur prior to the P300. Because the correlation between the P300 and the onset of perceptual consciousness of a novel stimulus has been somewhat reliably observed, I shall use it as an auxiliary assumption in assessing the implications of the inattentional blindness effect. I also assume that the emotional brain areas function to preserve the selforganizing interests of the organism. There is strong evidence that the PAG and midbrain respond to important homeostatic imbalances by altering neurotransmitter functions, especially neuropeptidurgic activity, and that the resulting distributions of neurotransmitter levels play an important role in readying the organism for both learned and instinctual action patterns whose tendency is to restore the initial imbalance (Panksepp 1998; Damasio 1999; Watt 2000; Faw 2000). In essence, the emotional brain works in the interest of maintaining the self-organizing integrity of the organism. While “emotion” and “motivation” are different, as noted earlier, there is an important sense in which the activity of the emotional brain motivates the organism to act in the interest of maintaining the integrity of self-organizing patterns. The emotional activities result from a self-organizational causal sequence and, according to enactive theories, are present only in organisms capable of action, not in those capable only of reaction.

. How the Mack and Rock data relate to the two types of hypotheses If the preconscious steps that process incoming information function independently of self-organizational effects of the emotional brain areas, then a purely reactive theory of consciousness might seem possible. Consciousness





Curious Emotions

would just result from a sufficient intensification of the preconscious mechanisms, which would be linear (which is not to deny that they may involve feedback loops and shunt mechanisms). On the other hand, if the action of the emotional brain areas does affect the ways in which information is processed (beyond serving as a mere power switch), then a purely reactive theory of consciousness would be impossible, because in that case consciousness could not occur without presupposing some functioning that is enactive rather than reactive, and which is uniquely necessary for consciousness – i.e., functioning that is geared toward achieving the emotional purposes of the organism. For example, if attention is controlled at an early stage of processing by emotional brain areas, then attentive consciousness itself presupposes some functioning that is enactive rather than reactive, and therefore attentive consciousness cannot be explained as the final outcome of a series of reactive steps in the processing of sensory input, unless those steps are also influenced by enactive processes prior to reaching the level of consciousness. Furthermore, if consciousness can occur purely as a result of those same enactive processes, without any perceptual input, then consciousness would be completely the result of this enactive process, and the reactive processes that are also present when we do receive perceptual input would have no part in determining whether the information processed achieves a conscious status or not; they would only affect the content of the information that gets processed, not whether or not it is conscious To many researchers and theorists, as we have already noted, this enactive story is questionable or even unacceptable, essentially because a reactive account would be consistent with the following seemingly plausible scenario: Visual input causes processing steps to occur first in the thalamus, then in the occipital lobe, and then a series of causal loops occur between the thalamus and the occipital lobe, where each of these loops is a series of purely linear (non-self-organizing) causal sequences. When the occipital processing has led to sufficiently enhanced neural firing, subsequent linear causal sequences lead to a pattern of parietal (and perhaps also frontal) firing (for example, the P300), and this later-stage firing is tantamount to consciousness of the relevant perceptual input. Consciousness is merely the last stage of a chain reaction. Such a story might even include the caveat that a certain level of general physiological arousal is presupposed as a background condition for the intensity of neural activation needed for consciousness. But, given physiological arousal as a minimal background condition that can generally be assumed in awake organisms, the consciousness of a perceptual datum, according to a reactive theory, would be completely caused by the sequence of causal steps that constitute processing of the perceptual input. Correlatively, essentially the same

Chapter 2. Motivated attention in action

information processing might sometimes even be possible in the absence of the level of arousal needed for consciousness, but in that case it would occur on a nonconscious basis. On this view, the toning of the cortex through general arousal would be analogous to the volume control of a stereo speaker. If the speaker volume is turned all the way down, this does not change the encoding pattern with which signals are processed in all the other stereo components before the final signal reaches the speaker. What happens in the speaker is still clearly a causal outcome of what happens in the other components – it is not one of the causal inputs into the other components’ behavior. Are the Mack and Rock results compatible with this type of theory? When combined with some well-known and fairly non-controversial physiological findings, I think they are not, at least if we take them as indicating what Mack and Rock themselves say they indicate – that attention must be directed before there can be perceptual consciousness, and that in many instances there is not even implicit knowledge of the unattended stimulus. If, as Mack and Rock put it, “attention . . . must be captured before perception can occur” (even if we take “perception” in this context to refer only to conscious perception), it follows that the only way the perception itself could be purely reactive would be if attention did not itself already presuppose an enactive causal process – i.e., if attention were not motivated by the homeostatic needs of the organism to maintain self-organizational balances. In more concrete physiological terms, perception could be reactive only if it were not the case that all conscious attention is causally affected by emotional brain activity (which we have assumed responds to self-organizing causal structures of the living organism). Assuming a parietal P300 event-related potential as an operational measure of conscious awareness of a newly presented stimulus, is it possible that perceptual processing prior to the P300 (as depicted in Figure 2) is not affected by the motivational purposes of the organism? If the processing prior to the P300 were not affected by motivational factors that control selection for attention, then there should always be implicit knowledge of any perceptual input that proceeds up to but does not include the P300, since in that case the information would already have been almost completely processed prior to the conscious awareness indicated by the P300. Whether the P300 (and correlative consciousness of the object) occurred or not would be decided too late to impact the extent to which the information is processed. Thus any information that is processed as far as the P300 should cause implicit knowledge regardless of whether the P300 occurs or not.





Curious Emotions SEQUENCE OF ACTIVATIONS IN RESPONSE TO UNEXPECTED STIMULUS P300 (Moment of Consciousness) Parietal lobe

200ms. (Disinhibition of Action Command) Motor cortex 20–200 ms (Inhibition of Motor Command = Sensorimotor Action Image)

N200 ERP (Mismatch Negativity) Occipital lobe

(Object)

P100 ERP (Perceptual processing) Cerebellum 20 ms ERP (Initiation of Motor Commands)

Thalamus

Hippocampus 18 ms ERP (Motivational Activation)

Figure 2. Note that ERP numbers indicate number of ms after input stimulates the retina. The figure indicates that emotional activation, initiation of tentative motor commands, and forming of action imagery all precede the perceptual consciousness of an unexpected stimulus that correlates with the P300 ERP.

Equivalently, we can ask whether or not motivational factors that control selection for attention affect the processing of information prior to the point when the subject becomes consciously aware of the specific perceptual input corresponding to the consciousness. If there were no such motivational effects, then again the processing should occur in about the same way prior to the consciousness of the object with or without such effects – and thus there should be

Chapter 2. Motivated attention in action

knowledge of the processed stimulus, if processed this far, regardless of whether we are conscious or unconscious of it. Mack and Rock do not find this implicit knowledge except under very special circumstances, and even then the implicit knowledge is very limited, and lends itself to an enactive explanation. For example, they find implicit knowledge with preconscious meaningful words, but not with nonsense syllables. This would be explainable on an enactive basis; meaningful words are more likely to be useful for the organism’s survival and well-being than nonsense syllables. Why did Mack and Rock not find implicit perception with other (less meaningful and less motivationally relevant) kinds of critical stimuli, such as geometrical shapes or nonsense syllables? Granted, it may be that the subjects did have implicit knowledge, and the Mack and Rock methods simply failed to confirm its presence. But even if this were the case, we would still have to ask: Why did the same tests that do reveal implicit knowledge for some types of stimuli fail to reveal similar implicit knowledge for other types? If a reactive theory were true, then the amount of implicit perception that occurs when we are not conscious of a stimulus would simply be a function of the amount of sensory processing that has occurred – occipital and other visual-cortex systems for visual stimuli, for example – and this in turn would be determined by the effects of the stimulus itself, in interaction with the brain mechanisms that are already in place. So meaningful images would not show greater implicit perception than meaningless ones, if the amount of sensory processing is the same. With the enactive theory, on the other hand, visual data would not be processed as much when not evaluated as potentially useful in some way for the organism’s purposes; so the brain mechanisms would have to do something active in order to select an input for attention at a relatively early point in the processing of the various inputs available, otherwise it would not be processed far enough for either attention or consciousness. An enactive theory can allow that there could also be some later selection from among those stimuli that had made it through the early selection process. To be sure, a reactive theorist could also insist that there can be early selection because the organism may be classically conditioned, for example, to be more receptive to some stimuli than others – in which case the organism would not have to do anything active in order to make itself receptive to selected stimuli. I.e., the fact that an acquired physiological state is a precondition for an input to cause a response is perfectly compatible with reactivism, because even in reactive theories, the physiological precondition could have gotten the way





Curious Emotions

it is in a variety of ways, including past learning. Such a precondition, however, would tend to be rigid and inflexible. But the response here is that, in an enactive theory, there can be virtually complete sensory area processing – for example, occipital processing – without consciousness, and furthermore the consciousness is not contributed by anything caused by the sensory processing itself. Given a certain level of sensory processing, if there is no P300 and beyond, then there is no consciousness, while at the same time the P300 is not caused by the sensory processing itself, because the time delay from occipital to parietal processing (about 1/4 second according to Aurell 1989; Bachmann 2000; and others cited earlier) is too great to be explained by a direct spreading of activation within such a close proximity. Instead, the enactivist explanation is that the P300 must be caused by independent subcortical activity in interaction with the sensory processing. Some processing occurs unconsciously prior to the P300, yet the P300 is not caused merely by that processing, but also is caused directly, in a separate causal stream, by endogenous subcortical motivational activity looping with anterior attention mechanisms. These include anterior cingulate as well as frontal activity (Balconi 2003; Damasio et al. 2000; Posner & Rothbart 1998). The more enactive a theory is, the more it holds that attention functions to select items that are useful for the organism’s purposes, guided by preconscious brain processes, and in ways that are subject to organismic flexibility rather than rigidly hardwired. Rigidly hardwired selective processes can easily be explained in reactive terms, but early selection mechanisms that are flexible and subject to purpose-directed organismic modulations are more difficult. The more reactive a theory is, the more it holds that attention is caused by the stimulus itself in connection with the organism’s hardwiring. Enactive theories allow both for very early and for late selection – the former because emotional processes must already be active in selection before much sensory processing has occurred; the latter because, even though emotional processes can be activated very quickly (18–20 ms), the emotionally initiated neurotransmitters (especially the neuropeptides) are very slow in spreading, so it is useful to be able to have information already processed and ready for later selection if needed. But without early selection, there would be too much processing going on, and the organism would be directionless in its seeking activities. By contrast, with the exception of some rigidly hardwired early selection, reactive theories favor late selection, because attention itself must be caused by the processing that has already occurred. Thus attention will occur late, when the processing is substantially complete. This view can be attenuated by intro-

Chapter 2. Motivated attention in action

ducing feedback loops into the chains of reactive processing, but in a reactivist view, if attention selection occurs at an early point in these loops, then the earlier attention mechanisms must be very rigid and not related to whether the processing is conscious. Conscious attention, by contrast, consequently can result only from late selection on the reactive view, since consciousness must be one of the later causal results of the processing: When we are conscious of data, we have already extensively processed them. What is novel about the Mack and Rock studies is precisely that, when we consciously direct our attention toward one task, irrelevant stimuli tend not to be processed readily unless they are (1) emotionally salient (for example, the smiling face); (2) meaningful (for example, the written words that yield implicit perception; the smiling face, too, may be a meaningful “icon,” at least in Western cultures); or (3) types of objects that the organism is strongly habituated or even hardwired to select in terms of very gross properties, such as loud noises, very large moving objects, sudden bright lights, etc. A reactive theory will not easily allow that mechanisms (1) and (2) can determine early selection, since this kind of selection requires some sort of teleological mechanism whereby the organism guides its own perceptual processes as dictated by emotional purposes, rather than having them occur in reaction to stimuli. It requires that a response have occurred prior to the stimulus, and that the response determines the stimulus at least as much as the other way around. What is especially important for enactive theories is that consciousness occurs only as a result of such a response that precedes the input corresponding to the intentional object of the consciousness

. The paradox of early and late selection Mack and Rock find it “paradoxical” that an unattended meaningful word, such as “snow,” is not seen when presented for 200 ms, yet in subsequent stem-completion tasks, implicit perception occurs, suggesting late selection for attention, after extensive unconscious processing. This is indeed a paradoxical finding, because the experiments using meaningless shapes or nonsense syllables suggest earlier selection, since in these cases the objects that are not attended to do not seem to be seen, even implicitly, as far as Mack and Rock can detect. In these cases, the subjects’ knowledge of the image presented “did not differ from chance” (Mack & Rock 1998: 13). The resolution of this paradox must involve acknowledging that there is both early and late selection for attention, as Posner and Rothbart (1992, 2000)





Curious Emotions

have suggested. The idea that there are both early and late selectional gating processes for stimuli relevant to the organism’s purposes is not a new one, and has been defended by Triesman (1964) and many others. What Posner and Rothbart propose is that anterior subcortical motivational and emotional mechanisms determine early selection; posterior attentional mechanisms then unconsciously enhance signals that have penetrated the early selection “gates”. Finally, if the data survive early selection gating and corticothalamic enhancement, then anterior-limbic looping processes (which are both subcortical and cortical) motivatedly activate image schemas (Johnson 1987; Newton 1996) that resonate with the posterior nonconscious processing; at that point, consciousness occurs. Now, there are arguments pro and con as to whether consciousness occurs at a particular point in time (for example, see Dennett 1991); but we do know, because of the ERP data cited above, that visual consciousness of a newly presented stimulus does not occur until a P300 parietal ERP is registered. “Image schemas” in this context refer to proprioceptive and sensorimotor routines based on action affordances suggested by an initial Gestalt impression presupposing only early thalamic awareness of the relevant perceptual input, prior to perceptual processing in sensory areas such as the occipital and temporal lobes, and assessed in terms of emotional feeling tones associated with the general impression of a given Gestalt through past learning. Since emotions and action commands can be unconscious when not associated with action imagery and/or perceptual/interoceptive imagery, image schemas do not seem to be conscious until they give rise to further processing, such as visual, auditory, or proprioceptive/sensorimotor imagery (Newton 1996). The more meaningful the data, from the standpoint of the organism’s overall motivational purposes at the given moment, the more the organism will be primed to allow the data through the “gating” mechanisms of the thalamus and anterior cingulate, which in turn are sensitive to the organism’s emotional needs through extensive connections to limbic and midbrain emotional areas. The fact that there can be extensive processing and encoding of data without attention corresponds to the physiological finding that there can be virtually complete occipital activation with no perceptual consciousness (Aurell 1989), and that the anterior cingulate must be activated in cases of voluntary attention, while both thalamic and limbic (including posterior cingulate) control of corticothalamic loops appear to be important for involuntary attention (Bernstein et al. 2000; Posner & Rothbart 1992, 1998). These physiological events would map well onto a phenomenological situation in which, if the given stimulus is anticipated as important for motivational purposes, then its

Chapter 2. Motivated attention in action

signal is more likely to be enhanced via corticothalamic loops as preselected by the thalamus in cooperation with the brainstem because of categories of utility dictated by the organism’s emotional purposes. Thus, for motivationally “primed” stimuli, there is extensive occipital or temporal-lobe processing even prior to the direction of conscious attention. But for meaningless or motivationally insignificant stimuli, this processing will be considerably less, because the preselective (i.e., early selection) mechanism of thalamic “gating” (or the “extended reticular thalamic activating system” gating discussed by Newman & Baars 1993; and Watt 2000) will control corticothalamic loops in such a way as to minimize the enhancement of those signals. This would also explain the anomaly we considered in Chapter 1 as to why, on presentation of a novel and unexpected stimulus, there is approximately a 1/4-second time delay between almost complete occipital activation and the more anterior cortical activation necessary for consciousness of the object, particularly the parietal P300 electrical potential that corresponds to perceptual consciousness (Aurell 1989; McHugh & Bahill 1985; Runeson 1974; Srebro 1985). As noted earlier, at the rate neural impulses normally spread in the brain, one would expect a neural impulse to be able to travel somewhere in the neighborhood of 10 to 20 meters in 1/4-second. Obviously, more is required than simply a spreading of activation from the occipital to the parietal lobe. Before we are conscious of a visual stimulus, other brain areas must be activated, including more anterior areas such as the anterior cingulate and the frontal and parietal lobes, and this activation must take place in response to input from emotional brain areas. In view of the 1/4-second delay, it seems very unlikely that these more frontal areas are activated in response to the occipital lobe’s activity at all. Instead, it seems likely that the organism must purposely activate the frontal and parietal areas to “look for” emotionally important categories of objects which the thalamus has already alerted the organism might be relevant, and this “looking for” activity must already have begun the forming of visual or conceptual image schemas (including proprioceptive and sensorimotor imagery associated with possible action affordances) prior to any occipital activity’s having any effect on our perceptual consciousness (since at this point the impulse has not yet spread from the occipital to the parietal and more frontal areas). For example, meaningful items in the environment (such as the smiling face) would attract more attention than meaningless ones because the organism is “pre-tuned” to be on the lookout for meaningful early clues that an object is worth paying attention to.





Curious Emotions

It seems therefore that the frontal-parietal-cingulate activation must take place at least as early as the perceptual consciousness signalled by the P300; the frontal-parietal-cingulate pattern must therefore play an important role in determining whether any given perceptual input will even register in consciousness, i.e., will be attended to. Moreover, the emotional areas that precipitate this frontal-parietal-cingulate activity appear to be essentially the same ones that determine whether a stimulus will make it through the preselective corticothalamic gating process that determines whether it will even be fully processed and encoded in the primary perceptual area (for example, the occipital lobe) to begin with. In effect, the perception of a stimulus can occur in conscious attention only if it can make it through two different selection processes – an early and a late one – and both these processes are controlled subcortically in terms of the motivational purposes of the organism. We have already seen that this is equally true in voluntary or involuntary attention, and even in cases of complete frontal lobectomy. In all these cases the emotional brain systems still play a selective gating role in accordance with general motivational purposes, by means of subcortical control of neurotransmitters and by the resonance between limbic (or ERTAS) and posterior cortical activity (Damasio 1994; Watt 1998, 2000; Faw 1997, 2000); otherwise, information processing would be no different in conscious than in unconscious instances, and all knowledge would be merely implicit knowledge. It would therefore appear likely that there is both early and late selection, which map into the brain areas just discussed to make us become conscious of a perceived object in three stages: (1) Preselective gating is determined by motivational categories of utility, mainly subcortical and anterior. (2) Corticothalamic looping enhances, processes, and encodes stimuli once they penetrate the initial gating. (3) Resonance is achieved between the anterior attentional mechanism and posterior processing, leading to consciousness; for a novel stimulus, this total sequence of events requires about 1/3-second to elapse (Aurell 1989) from the time a novel stimulus is first presented, and results in the parietal P300 ERP discussed earlier.

. Attention and conscious processing The third of the three stages just listed – resulting in consciousness and the parietal P300 potential – should not be regarded as a trivial ancillary as far as processing of information content is concerned. In the first place, the information processing is likely to be structured differently depending on whether it is

Chapter 2. Motivated attention in action

conscious or unconscious. In the former case, subcortical and anterior brain regions play a very active role in gating prior to the transformation and encoding of perceptual information. This is reflected in the Mack and Rock failure to detect implicit perception for meaningless and motivationally insignificant stimuli. And secondly, it should not be forgotten that, even in cases where priming has occurred, although there is nonconscious processing and retention of information, this processing and retention are not nearly as complete as they are for conscious perception; this difference is reflected in the Mack and Rock findings just as in any other blindsight or implicit perception experiment (for example, Weiskrantz 1986). Thus consciousness is not what Natsoulas (1993) has called a mere “appendage” to information processing that could just as well have occurred without consciousness. In physiological terms, this means that we should not regard the processing and encoding of information content as a completely posterior activity, and then regard the anterior direction of attention as merely bringing into consciousness the information that has already been completely processed. Information processing has both a different structure and a different efficiency depending on whether it has been done consciously or only unconsciously. Thus data of a kind that in principle could be relevant to the organism’s needs will be more likely to be processed, even below the conscious level (hence the implicit perception effect in the trials using meaningful words, which does not seem to similarly manifest itself in trials using meaningless images or nonsense syllables). At the same time, the organism will be more likely to direct conscious attention to kinds of data that are important for its purposes, unless distracted by attention to another task that, for the moment, is considered still more important. And attention will be “pulled” by items emotionally important to the organism’s overall purposes. Thus a newborn infant shows attention to a smiling face (Barresi & Moore 1996; Meltzoff & Gopnik 1993), just as the Mack and Rock subjects’ attention was pulled toward the image of a smiling face (but not by a neutral or scrambled one) even when their attention was previously occupied.

. Further implications for the problems of attention and consciousness There would seem to be several important philosophical and neuroscientific implications of these differences in the structuring of conscious and nonconscious processing. (1) A state of consciousness appears to involve global brain activity, and thus cannot be highly localized; if this is true, then it would fol-





Curious Emotions

low that logical thinking in humans, which is more effective when done on a conscious basis, cannot be highly localized (Ellis 1995). We have seen that even the intentional contents about which we reason require widely distributed activity if they are to be entertained in consciousness. Thus everyday logical inferences must be even more complex activities involving widely distributed brain areas, not merely highly localized transformations. (2) It would seem a safe prediction that consciousness will occur only in organisms capable of purpose-directed, emotionally motivated direction of attention, and this is a function of emotional brain areas; i.e., consciousness cannot be a function of a modular brain area or system that relates to the subcortex only as a “powerswitch” (Ellis & Newton 1998b, 2000a, b). (3) Resolution of the traditional mind-body problem will very likely hinge on understanding the causation of the motivational dimension of consciousness in self-organizational terms – for example, as in dynamical systems theory (Kauffman 1993; Mac Cormack & Stamenov 1996) – rather than in linear stimulus-response terms, yet at the same time, a dynamical systems account will not be likely to do the job adequately unless it addresses itself to the role of emotional and motivational processes. Thus, on the one hand, the phenomenological datum that a conscious decision to raise my hand plays a causal role can be made consistent with the fact that the physical movement of the arm is sufficiently explainable as resulting completely from physical causal antecedents (Ellis 1995, 1999a, b, 2000a, b; Wegner 2003). These two facts can be reconciled if the motivation to raise the hand can be framed as an endogenous, active process rather than a passive response to inputs, and the concept of self-organization makes the needed active/passive distinction possible, as we saw earlier. (4) Consciousness is not just an extra layer added onto a mechanism that could have occurred in the same way without consciousness. Instead, information is processed very differently depending on whether consciousness is involved – both because different brain mechanisms are needed for conscious processing, and because the information is processed differently when consciousness guides the direction of the process (Natsoulas 1993). We could go into more detail analyzing just how far the brain can go toward processing data as a result of “automatic pilot” corticothalamic loops, and at what point a conscious direction of attention is required for further processing. Bickle et al. (1999) and Bernstein et al. (2000) go further into the physiological aspects of this question. Their findings indicate that (1) both weak and strong visual stimuli are processed at least as far as occipital area V-1, although (2) the excitatory-inhibitory interactions will adjust the intensity with which the signal arrives even so far as V-1, with the more salient stimulus being en-

Chapter 2. Motivated attention in action

hanced up to a certain optimal point, and the less salient one dropping to a minimal level but not entirely dropping out. (3) This difference in strength primes V2-forward for further processing of the more salient stimulus. (4) Not only the anterior cingulate, but also brainstem areas (controlling acetylcholine, norepinephrine, serotonin, and histamine) affect how the thalamic reticular nucleus (TRN) functions, with its inhibitory effect. Those subcortical effects are also emotional effects on the way signals are selected for more or less processing (Panksepp 1998; Damasio 1999). However, the general question remains as to whether anterior brain areas are really necessary for involuntary attention. Faw’s (2000) response is that the degree of consciousness of whatever stimulus is there is determined at least as much by the level of subcortical (and emotionally determined) flooding with neurotransmitters, noradrenalin, etc., even without frontal control – a more basic level of cortical arousal. At the same time, Faw is careful to distinguish two different issues: (1) Degree of consciousness, which can generally be determined by cortical arousal at the more basic arousal level just mentioned; and (2) selective attention/inattention, which is a separate function in addition to mere cortical arousal. Faw’s distinction is similar to Damasio’s (1999) distinction between “core consciousness” and higher content-processing functions in that arousal relates to whether or not there is consciousness and wakefulness, whereas selective attention affects the content of the consciousness. Thus when the monkeys in the Gallant et al. (1998) studies casually look out at an emotionally neutral scene, their V-1 areas are found to be largely quiet with only a few “hot spots” in exploring the scene. Here we see the arousal versus selective attention distinction, with plenty of consciousness (from arousal) but not much selectivity (due to little difference in the salience or motivational value of stimuli). However, there are also two more primitive ways in which organismic purposes can “preselect” for attention prior to the signal enhancement that results from occipital-thalamic loops. One primitive way this preselection can occur is simply through association of stimuli with emotions through past learning, which would cause an emotionally important stimulus to trigger increased general arousal. The arousal would then excite the processing of that stimulus, which would lead to further inhibition of processing of other stimuli. If LeDoux (1996) is right, this would be largely an amygdala function. Still another primitive kind of attentive preselection mechanism in the service of motivational purposes is that, simply through natural selection, our organisms would be wired in such ways as to be always especially sensitive and alert toward the types of stimuli that are likely to be important for our sur-





Curious Emotions

vival – peripheral motion, intense signals, the unexpected, etc. There would not need to be any presently existing emotional mechanism for this; natural selection would have already taken care of it. Yet it would be a way in which we would be geared up to selectively attend to emotionally important stimuli. But the more interesting way in which subcortical emotional purposes direct attention occurs in the case where we have to selectively attend, inhibiting competing signals, because this is the case where consciousness comes into play. Even monkeys that are just generally looking out at a neutral scene are being motivated by curiosity (Panksepp 1998), and are tacitly giving that curiosity priority over weaker competing emotions that, if they became strong, would override the generalized curiosity and direct attention only to items that look specifically interesting, such as bananas. In that sense, they are making a kind of selection. Thus the organism is always sensitive to the possibility of having its attention redirected, even when no particular salient stimulus has presented itself – and if it were not, as the Mack and Rock findings suggest, there would be no consciousness of any object. On the one hand, Mack and Rock want to say that attention determines whether we see something at all, but on the other hand they say that the data are already extensively processed, and only then do we determine what gets attended to. The resolution of this paradox requires consideration of emotional brain functions, which shows that there is both early and late selection. All stimuli reaching the occipital lobe are processed somewhat extensively whether attended to or not, but at the same time, prior to this processing, there has already been considerable difference in the degree to which potentially motivationally important versus unimportant stimuli as well as salient versus non-salient stimuli are processed, and those differences are determined quite early, through TRN inhibition prior to complete V1 processing, and also through anterior cingulate direction of voluntary attention, which already precedes the differentiation in processing between the attention-grabbing and less attention-grabbing stimuli. In that sense, there seems to be both early and late selection for attention. The remainder of this book will explore more specifically the way selforganizational purposiveness expresses itself through the human motivational and emotional dimensions. The next two chapters will develop a selforganizational account of specific emotions and emotional systems. The first step is to examine more closely the way the demands of self-organization are manifested in phenomenal experience and in the physiology of emotion, because one of those demands is not generally recognized in the context of its applications to emotion and motivation – the demand for a suitably complex

Chapter 2. Motivated attention in action

and high-energy basin of attraction, which I call “extropy.” This tendency toward extropy expresses itself in terms of many primary yet non-consummatory emotions, such as curiosity and the needs for exploration, play, and social relationships.



Chapter 3

Non-consummatory motivations Extropy and “life wish” in the self-organization of emotion

One of the most exciting implications of understanding emotions in terms of the action capacities of complex dynamical systems is that, if such systems can reorganize their own micro-components and subsystems to maintain basins of attraction at preferred levels of energy and complexity, then it becomes comprehensible why non-consummatory motivations such as exploration, play and nurturance could result from a tendency of complex dynamical organisms toward higher-energy attractors, and are not as easily explainable as are consummatory motives purely in terms of energy-efficient homeostasis. In consummatory satiation, chemical needs such as in hunger or thirst are the driving forces, pushing toward consumption of chemical resources needed for homeostasis. But homeostasis can be achieved at both higher and lower energy levels. In this chapter, I shall define “extropy” as the tendency of some complex self-organizing systems to prefer higher- over lower-energy basins of attraction in some contexts. The lower-energy basins, however, can be convenient paths to energyefficient homeostasis within component subsystems, which is the driving force of consummatory needs. I shall argue that extropic motives differ from the purely homeostatic ones – i.e., those which seek a maximally energy-efficient homeostasis – in that the extropic ones resist the complacence that would follow from consummatory satiation if energy efficiency were the only ultimate aim. Non-consummatory motives, while maintaining homeostasis, can fundamentally conflict with consummatory ones by pushing toward higher-energy attractors in the interest of maintaining complex structural patterns for the overall system. Thus my argument is that extropy is an independent demand alongside homeostasis. Then in the next chapter, which builds on the concepts developed in this one, we can finally integrate these theoretical conclusions into more specific neurophysiological and phenomenological emotional investigations. This



Curious Emotions

analysis will show that there is actually a fairly large number of irreducible emotional-motivational aims, centering around three main types of needs: Extropic, homeostatic, and boundary-protection needs, conceptualized around these three basic dimensions of complex dynamical systems. Let’s begin by confronting an important reality of emotion studies, which inevitably must be multi-disciplinary. Intertheoretic reduction in the sciences, the view that psychological and biological processes are ultimately explainable in terms of atomic-level chemical and physical processes, is not as fashionable as it once was, but its main impact continues to be felt in the ever-present temptation toward what I shall call a “micro-reductionist” approach to motivation and emotion – the notion that all motivations are ultimately reducible to the effects of a thermodynamic energy-consumption system that could be equated on the psychological level with primary consummatory drives. Moreover, there are strong philosophical reasons for believing that the sciences collectively must tell a coherent, and thus a unitary story about reality; it can thus be argued, with Bickle (1992, 1998, 2002), that all sciences must be reducible to one basic science, which tells the story of how the most elementary phenomena – atoms, forces, etc. – interact. Such a view would suggest a reduction of all the phenomena of the brain sciences to basic chemistry and physics. Now if emotion studies are to have any hope of rooting themselves in a constructive multidisciplinarity, including the concerns of psychotherapy, philosophy, and the arts and humanities as well as the hard sciences, it is crucial to show that a solid grounding of psychology and biology in the more mechanistic formulas of chemistry does not entail a reduction of the total self-organizing system to a mere summation of the effects of its micro-constituents, yet also does not ignore the requirement for unity of all the sciences – i.e., does not deny a grounding of brain processes in basic chemistry. Self-organizational processes, as I have already argued, involve causal dynamics that are not confined to the effects of the micro-constituents of the complex system. Part of the importance of this point is that science does not demand rejection of humanistic and existential suggestions that some primary or intrinsic motives seek to avoid the complacence and stagnation in which all motivation would ultimately revolve around the achievement of a maximally energy-efficient homeostasis of micro-processes, and thus of mere comfort or satiety, with the latter explainable presumably in terms of the energy conservation demands of the organism. I shall refer to the postulated non-satiationdriven primary or unconditioned motives as “non-consummatory” ones. As we have noted, many emotion theorists have already rejected the notion that energy conservation is foundational for all other bodily functions.

Chapter 3. Non-consummatory motivations

Such non-consummatory emotions and motivations do not contradict the possibility of intertheoretic reduction in the sciences, from psychology all the way down to thermodynamic chemistry and subatomic physics, although they do require a theory of self-organization in complex systems for the comprehension of their physical basis. My thesis is that not only do complex selforganizing systems seek homeostasis (which already entails “negentropic” as well as entropic functions), but moreover they also sometimes seek “extropy,” which I define as a system’s positive preference for higher-energy over lowerenergy basins of attraction, even in many instances where either type of basin could serve the purposes of homeostatic energy-efficiency equally well. Exploratory, play, social bonding and nurturance subsystems in the brain are correlated on the physiological side with the “extropic” tendencies of complex systems, and on the phenomenological side with an endogenous and unconditioned tendency toward increased intellectual and emotional stimulation (up to some optimal point), and at the most complex human level a need for existential meaning, i.e., for the experience of the value of being per se, whose achievement in humans is sought through personal relationships, ethical values, and meaningful life activities including those that express creativity. The problem here is to reconcile viewpoints that have tended to be at crosspurposes. On the one hand, hard scientists want to ground the higher-level sciences of psychology and biology in chemistry, especially given the influence of the biomedical model in psychiatry; and this makes consummatory-drive reductionism attractive as a convenient way to connect thermodynamic energy systems with the biological systems that give rise to our emotions and motivations. On the other hand, there is the phenomenology of life experience, which seems to call for a richer system of motivations and emotions, to account for the preference for kinds of meaning that do not seem reducible to the satiation of the thermodynamic bodily system’s needs for energy consumption, such as getting something to eat or drink. I believe the notion of self-organization in complex systems can reconcile these two antitheses in such a way that the advantages of both can be preserved. On the humanistic side, human emotions need not be reduced to a complex system of means toward the end of satisfying simplistic consummatory needs. And on the hard-scientific side, the ideal of intertheoretic reduction can be maintained, at least as a way to open the possibility of dialogue between disciplines focusing on different levels of organization, so that different sciences and other disciplines concerned with human consciousness and emotion are not condemned to work in isolation or at cross purposes to each other.





Curious Emotions

.

Intertheoretic reduction and consummatory-drive reductionism

Human emotion and motivation have been viewed in notably divergent ways by the hard sciences and the humanities during the past century. That a placid, cud-chewing way of being is not the optimum for humans has been emphasized by literature and the arts, which have cross-fertilized with humanistic and self-actualization psychology and phenomenology, and are substantially congruent on this point with the ego psychologists, the object relations theorists, and the later Freud (beginning with Beyond the Pleasure Principle). According to Freud, there must be some instincts that tend to resist the attainment of satiety, because in his view consummatory satiation resulted literally from the energy-efficiency rather than the energy-expenditure needs of the organism, and in his later years he saw serious problems with reducing all motivation to the energy-efficiency needs. He therefore sought to conceptualize some drive or instinct whose tendencies would run contrary to the purely satiation-driven instincts, which result from the chemical tendency to move toward what Freud loosely called “inertia” (1925/1959: 68) – a maximally energy-efficient homeostasis. Part of the problem he saw with thinking of all behavior as in the service of consummatory satiation is that, in principle, if the energy-efficiency demands were the only motives, yet had been met, then the organism would not be motivated to do anything. Another aspect of the problem was that to make complete satiation the goal would lead toward a condition in which “the aim of all life is death” (Freud 1925/1959: 70). Freud means this aphorism literally. If a system were driven only toward the relative electrostatic stabilization of the molecules in successive steps of its various feedback loops – toward Freud’s “inertia,” a state of satiety and restfulness where the organism is not motivated by any further need – then the real motive of the satiation phase of nervous activity would be to maximize energyefficiency, i.e. to satisfy each molecule’s demands for electrostatic free energy reduction, insofar as possible given the initial structure of the system with its requirements for homeostatic balance in and between its interconnected chemical feedback loops. For example, when hydrogen reacts with chlorine to form HCl, the driving force is that the lone electron in the outer electron shell of the hydrogen atom tends to reduce its energy level by filling the lower-energy space available in the outer electron shell of the chlorine atom. By sharing the electron, both atoms reduce their energy level. Metabolic processes are built up from chains and loops of these electrostatically reductive reactions. The consummatory drives

Chapter 3. Non-consummatory motivations

thus seem to be motivated by a need to “consume” the chemicals and physical conditions needed to fuel this more basic free-energy reduction process. To be sure, it would then be up to the organism to avoid entropy, by replenishing the energy thus used up, spending part of it in activity designed to appropriate still more energy. But, as far as the satiation of consummatory drives is concerned, Freud realized that the most efficient way to achieve a completely sated condition with fulfillment of all chemical demands within the drive system would be suicide. A dead person’s drives have all been reduced, and all the chemical components that were pushing the system to reduce their free energy have done so quite effectively. Thus Freud, like many other humanists, looked for a way to conceive of some sort of emergent elan vital or life force, some principle of self-actualization or primary drive toward inherently self-motivated activity that would make it intrinsically rewarding for the organism to counteract the tendency toward submolecular free energy reduction, as opposed to relying on intrinsically unrewarding behaviors for this purpose. The later Freud thought he had found such a principle in the sexual instincts, although earlier he had tried to view sexual desire in purely consummatory terms. Freud was well aware that consummatory satiation does not lead to a reduction in the total amount of free energy within the organismic system, but just the opposite. It leads to an acquisition of energy to be stored for future use. He continually refers to the concept of homeostasis, the balance between energy-storage and energy-expenditure functions. But he also saw that homeostasis is usually meant as the most energy-efficient way possible to maintain balance between the forward and backward reactions of all its interrelated feedback loops – and this implies that the homeostatic balance is maintained with the lowest possible level of energy expenditure necessary to obtain the new energy needed to meet the organism’s energy-conservation needs. This continual push toward energy-efficiency in the sense of gravitating toward a minimal energy level is what led to Freud’s notion of the “death instinct,” which he thought needs a countervailing force – such as his new way of conceptualizing eros as a drive toward increasing the energetic activity of the organism – to postpone the final entropy of death. Unfortunately, Freud would have needed a more adequate theoretical foundation in terms of complex dynamical systems – and perhaps clearer philosophical analyses – to make this notion of conflict between “death instinct” and “life wish” into a coherent account. But the contrast between Freud’s “life wish” and the drive toward energy efficiency (which grounds the “death instinct”) can be clearly formulated if we imagine a dynamical system





Curious Emotions

exchanging energy with its environment while maintaining a continuity in its pattern of activities at a certain level of complexity. Homeostasis is the system’s way to maintain balance between inflow and outflow of energy. “Extropy,” on the other hand, is the system’s tendency to push both inflow and outflow to a higher level, while still keeping them in balance. By contrast to this notion of an elan vital that would strive to counterbalance the complacence of energy-efficient homeostasis, modern physiology and empirical psychology have tended toward an opposite view, which I shall call “micro-reductionism” (see also Ellis 1995). According to this viewpoint, the only ultimate driving force of organic processes is the same as for inorganic ones – the tendency to reduce the free energy of the most basic constituents of the system, its atoms, as much as possible given the initial structure of the system with its various feedback loops. This tendency is reflected in the electrostatic interactions of the molecules in the system, and its ultimate net result is a tendency toward the lowest energy level compatible with homeostasis – the push toward a maximally energy-efficient condition that corresponds to Freud’s “death instinct.” The inorganic chemistry principle that all atoms seek to reduce the free energy of their electrons is what would render possible intertheoretic reduction in the sciences (Bickle 1992), if indeed such an ideal is possible. But to make this free energy reduction principle the only explanatory principle – the approach that I am calling “micro-reductionism” – is to ignore the effects of self-organization in complex systems. We have already seen reasons to believe that self-organizational effects are among the most prominent features of biological organisms (Kauffman 1993; Monod 1971). This point will be developed still further as we proceed. Even the nervous system itself is often explained in micro-reductionist terms as governed by a free energy reduction principle. For example, the tissues of a frog or eel respond electrically (Restak 1984: 30ff.; Ornstein & Thompson 1984: 77ff.), showing that the interactions of their molecules are governed by the tendency of their atoms to gravitate toward electrostatic free energy reduction. Long chains of molecules acting to move closer to neutrality by passing off spare electrons in this way to neighboring molecules constitute nervous impulses, whose ultimate driving force is the submolecular tendency toward free energy reduction. The fact that neurotransmitters play a role in conveying the charges from one neuron to another by means of submolecular electrostatic chemical reactions does not alter this basic principle. It is against this micro-reductionism that I wish to argue here – not against intertheoretic reduction in the sciences per se. My point is that even the most rigorously scientific accounts of the biological organism lead to a richer sys-

Chapter 3. Non-consummatory motivations

tem of emotions and motivations than the consummatory-drive reductionism that, to many motivational theorists, would be suggested by a tendency toward maximal energy-efficiency of the organism’s non-consummatory behavior. I take behaviorism and other learning theories that posit consummatory needs as primary to be classic examples (Hull 1952; Spence 1956; Rolls 1999). These theories explain rewards that are not “consummatory” (in the sense of allowing the organism to consume materials and physical conditions needed for maximally energy-efficient homeostasis) by relegating the non-consummatory rewards to the status of secondary reinforcements learned by association with the primary consummatory ones. If the characterization I am sketching here of micro-reductionism sometimes sounds like a caricature, it is only because the most extreme versions of micro-reductionism bring out its motivational implications all the more sharply. Micro-reductionism favors the reduction of negentropic (or anti-entropic) processes to more basic entropic ones at a lower level of organization. “Entropy” is usually defined as a thermodynamic system’s tendency toward randomness; but this randomness itself is a corollary of free energy reduction. Thus entropy itself can be defined in terms of the tendency of electrostatic interactions to occur that bring the atoms and molecules of the system to their lowest possible energy level; the electrons of the atoms are brought to the innermost available electron shell, where the electron is required to expend less energy to remain, as in the hydrogen-chlorine reaction resulting in HCl. Not only does intertheoretic reduction strongly urge this view, but molecular biologists have in fact been fairly successful in explaining the interactions of molecules in terms of inorganic oxidation-reduction and other such electrostatic energy-reductive chemical reactions (see Baker & Allen 1968; Campbell 1987; Monod 1971). According to the micro-reductionist account, the satiation of consummatory drives is explainable in terms of such submolecular electrostatic free energy reduction mechanisms, even though the overall effect of these drives (e.g., hunger) is to cause the organism to act in ways that continually increase the energy available to the system. This may seem like simply an incoherent claim to those not accustomed to reducing organic processes all the way down to their lowest chemical level. It may seem that the paradigm reward activity is to eat, which increases rather than reducing the free energy within the system, by converting the energy into ATP. But the microreductionist explanation is that the ATP reaction itself is composed of successive steps of chemical reactions, and each of these steps is caused by the tendency of the atoms of the molecules to reduce their free energy, which in turn leads to each successive





Curious Emotions

step. The overall effect for a complex organism is not to incline it toward complete entropy, but only to incline it toward the lowest possible expenditure of energy necessary to maintain a homeostatic balance between its endergonic and exergonic reactions. In thermodynamic terms, whether energy is consumed or acquired in a thermodynamic reaction depends on whether the reaction is endergonic (resulting in an energy gain for the system) or exergonic (resulting in an energy loss or reduction of free energy). In the Krebs energy cycle that animals use, like every respiratory cycle, and every inhibitory feedback cycle in general, it is initially the tendency for unstable chemicals (e.g., ions) to seek reduction of the free energy of atoms by getting their electrons into more stable energy shells (e.g., neutralization of ions) – an exergonic reaction – that gets the process going. But then, because of the way the system is organized, this leads to further reactions whose end result is that more potential energy is harnessed for the system (“endergonically”) by converting ADP to ATP (whose phosphate bonds are very easily broken; this is why ATP has so much potential energy). Meanwhile, the resulting ionization of iron from FE+2 to FE+3 that occurs in this reaction makes for further instability, and the FE+3 seeks an additional electron to convert it back to FE+2, which it takes from other elements (an exergonic reaction), which in turn starts the energy cycle all over. To fuel the process, of course, the organism must continually engage in foodacquiring behavior. But this “negentropic” behavior’s purpose, presumably, is to feed the entropic phase of the cycle, whose rewarding quality can be explained through reference to the demand of the micro-components (atoms) for free-energy reduction. The net result of this whole cycle of activity is a continuing harnessing of energy captured from food sources in the form of ATP. Yet each step of the reaction is driven by the tendency for that step to reduce free energy by getting electrons into less energetic electron shells by means of electrostatic free energy reduction. And this implies that, in net terms, the acquisition of new energy to fuel the reactions will also be carried out as energy-efficiently as possible, i.e., at the lowest energy level compatible with the homeostatic balance between endergonic and exergonic reactions. Thus everything is derivative from the principle that “As electrons are transferred they fall from higher to lower energy levels. . . . The drop in the potential energy levels of the electrons is the energy-releasing mechanism” (Baker & Allen 1968: 110–111). Or, to quote another standard textbook, “As a reaction proceeds toward equilibrium, the free energy of the mixture of reactants and products decreases. . . . To move away from equilibrium is nonspontaneous; it is an endergonic process that can oc-

Chapter 3. Non-consummatory motivations

cur only when an outside energy source pushes the reaction ‘uphill”’ (Campbell 1987: 99–100). At the same time that the whole metabolic process is driven at each step of its reactions by a tendency to seek free energy reduction at that step, and attempts at each step to move closer to equilibrium, the process cannot be allowed to reach equilibrium, or else the whole system would collapse into metabolic stagnation, and thus die. This does not happen because the ATP and other terminal products are continually used up to fuel further organismic activities, other chemical reactions that are outside of the metabolic loop. It is assumed that natural selection has favored the survival of organisms that happened to be structured so that some of the energy diverted from the ATP cycle is used to fuel food-acquisition activity to replenish the cycle to prevent entropy or death. For example, Rolls (1999) relies almost completely on natural selection for this explanatory purpose. This micro-reductionist account is perfectly consistent with the fact that no bodily system can ever reach the state of equilibrium toward which the free energy reduction tendency would lead if unimpeded. On the micro-reductionist view, this is because there are loops of reactions in which one chain of steps, each pushing toward maximal free energy reduction in its components, is inhibited by reactions engendered by some of its own products. Also, movement toward equilibrium in one subsystem is opposed by that of other subsystems. The most energy-efficient level at which all these loops and subsystems can be held in balance is usually defined as homeostasis. On the micro-reductionist view, homeostasis results when the forward chemical reaction in a feedback loop is balanced with the reactions in the opposite direction, so that the loop cannot collapse into entropy at either end of the loop. The same result is obtained when opposing subsystems reach this balance. Opposing chains of reactions, whether in different parts of a feedback loop or in different subsystems, are halted in their movement toward entropy at just the right point to maintain a constant cycling of reactions. In order to maintain homeostasis in any system or in the organism as a whole, the organism must behave in such ways as to input the right reactants into the system at the right times. (The organism of course allows for some flexibility in meeting these needs by providing numerous shunt mechanisms in its feedback loops.) Behavior is therefore prompted by a deviation from homeostasis. If there were no possibility of any deviation from homeostasis, then on the micro-reductionist account there would be no behavior. So the fact that the organism must behave so as to continually achieve “negentropy” does not contradict the micro-reductionist’s motivational as-





Curious Emotions

sumption that the satiation or consummatory phase of the cycle corresponds, at the lowest level of organization, to reactions in which the micro-constituents in the system – the atoms – are reducing their free energy and moving toward entropy. There must be just enough negentropy (and no more than necessary) to maintain the most energy-efficient possible homeostasis. Microreductionists thus tend to assume that the non-consummatory phase of the life cycle is not primarily rewarding, but consists of acquisition of further energy sources to fuel the free energy reduction demands of the micro-constituents of the consummatory process. That animals’ natural tendencies toward play, exploration, and curiosity do not seem to fit this energy-conservation paradigm is merely a Kuhnian anomaly in need of further explanation for the micro-reductionist type of theory. It can be handled in several ways – by assuming that such activities, which do not lead directly to energy acquisition, do indirectly lead to it, and thus are “secondarily” reinforcing (Hull 1952); or that natural selection favored organisms with such behaviors, even though they are not intrinsically rewarding (LeDoux 1996); or that the system simply falls short of its attempts at energy-efficiency, so that the activities in question, which burn energy that was previously acquired, are sometimes necessary for the reduction of free energy since the previously acquired energy must be burned off in some way (Spence 1956). In sum, according to micro-reductionism, organisms are driven ultimately by a demand for free energy reduction at each micro-step in the satiation phase (which is entropic) within a cycling organism-environment system whose overall effect is to continue replenishing the energy available to the system; the system then again uses up that energy by tending toward entropic energy reduction of micro-components in the satiation phase of the cycle. Presumably, the consuming of substances and conditions leading to submolecular free-energy reduction within components of the system, and in net terms an energy-efficient level of chemical activity (and thus a lowest-possible energy level compatible with homeostasis) is intrinsically rewarding, and in conscious organisms is experienced as pleasurable. Many micro-reductionists are then tempted to take the next step and adopt a consummatory reductionism as their entire drive theory, in which the satiation of any primary (unconditioned) drive moves certain parts of the organism’s system (i.e., certain atoms and molecules) closer to the lowest possible energy level compatible with homeostasis, given the structural constraints of the system. On this extreme version of consummatory reductionism, the only primary motivations are the ones that aim to “consume” the resources needed to achieve the satiation of these

Chapter 3. Non-consummatory motivations

maximally energy-efficient homeostatic conditions. According to this view, except for acquired secondary reinforcements, the organism would not behave at all if the consummatory drives were completely satiated. It is this step that Freud was reluctant to take. This is why Freud believed that, in the absence of some contrary, non-consummatory kind of drive, all life would be dominated ultimately by a “death instinct.”

. The notion of “extropy”: A non-reductive force? There is an entire variable that this discussion of electrostatic reduction has not accounted for. The whole system can continue its energy-harnessing and energy-reducing cycle either very close to the equilibrium point (where very little energy would be harnessed, but also very little would be used), or it can veer very far from equilibrium, while continuing the same homeostatic cycle. How far the system tends to veer from equilibrium while maintaining homeostasis is determined not by the fact that each step seeks electrostatic reduction, but by a property of the overall structure of the system, with all of its interrelated feedback loops and embryological tendencies to develop and maintain such structures. It has now been suggested by some theorists of the self-organization of dynamical systems (e.g., Kauffman 1993; Alexander & Globus 1996; Anderson & Mandell 1996) that these structural properties can create “constraints” on the possibilities open in terms of electrostatic reactions at the lowest level of organization. The constraints do not change any of the principles of the way the electrostatic reactions occur, but simply constrain the initial arrangement of feedback loops that determine whether a given reaction will be inhibited, catalyzed, etc. at a given point in the cycle. The possibility that homeostasis can be achieved at either lower or higher energy levels can lead to a further principle (independent of the instrumental need for “negentropic” acquisitive behavior), which in this book I am calling “extropy” – a motivational tendency that operates in conjunction with electrostatic reduction within homeostasis, but which can sometimes conflict with the movement toward reduction of free energy at any given step of a reaction. Extropy would therefore conflict, in some instances, with the tendency to satiation of electrostatic needs toward which subsystems at lower levels of organization are trying to pull. A simple example might be the conflict between a runner’s interest in finishing a race and the fatigue, thirst, and breathlessness pulling him in the opposite direction.





Curious Emotions

“Extropy” as I am defining it here is not reducible to mere “negentropy.” Even consummatory satiation requires negentropic activity of the organism to replenish its consummatory energy supplies; otherwise, it would not get the consummatory satiation. Indeed, all organismic processes, whether consummatory or not, involve negentropy, which is completely explainable in terms of relationships between micro-components by means of traditional thermodynamic concepts. The negentropic activity is geared toward maintaining homeostasis between energy consumption and energy replenishment. But what I want to suggest is more – that some organismic processes in some organisms (e.g., as discussed below, in mammals at least) involve an additional principle, which is what I am calling “extropy.” The point is that homeostasis can be maintained at various alternative energy levels and degrees of complexity, which correspond to various alternative basins of attraction (some higher-energy and more complex than others). Extropy is the preference for the higher-energy, more complex ways of maintaining homeostasis over the lowerenergy, less complex ones. If there ever is such a tendency in complex systems, it would not necessarily be reducible to mere negentropy, because an organism can maintain homeostasis equally well at greater or lesser levels of extropy, and all of those ways of maintaining homeostasis involve negentropy. Consummatory as well as non-consummatory motivations involve negentropy, which is necessary for homeostasis. But as I shall argue later in agreement with Panksepp (1998), only behaviors like seeking, curiosity, play, and some aspects of social bonding are motivated by a tendency to prefer maintaining this homeostasis via higher energy basins of attraction by preference over lower energy ones – i.e., even when the behavior is not indirectly reducible to the consummatory needs. Consummatory and non-consummatory motivations thus can be thought of as different ways of maintaining homeostasis – the consummatory ones being motivated toward basins that are at the lowest possible level of energy necessary for replenishing the energy needed for the tissue satiation; the nonconsummatory ones are at basins that are at a higher energy level than necessary for that purpose. “Consummatory” motives can be thought of as motives whose gratification involves replenishment of energy for consumption within organismic tissues. Non-consummatory motives then would be the ones that involve energy expenditure not for the purpose of replenishing energy for consumption by organismic tissues. For example, animals do not play in order to replenish energy consumed by tissues within the organism in a way necessary for homeostasis (Panksepp 1998). They seem to be motivated to play, i.e., burn energy, essentially for the joy of doing it, even though it does not necessarily

Chapter 3. Non-consummatory motivations

lead to replenishment of energy consumed by bodily tissues (e.g., acquiring food). I.e., it is not necessary, for the purpose of maintaining homeostasis per se, to engage in all of the high-energy activities involved in play, pure exploration, etc. While it is true that natural selection makes it advantageous for animals to have seeking and play systems that in turn make consummatory satiation indirectly more likely, the fact remains that the individual organism is not motivated toward consummatory satiation alone. Cats, for example, often would rather hunt than eat. To be sure, an extropic emotion may be triggered by a homeostatic imbalance. The hypothalamic-brainstem system serves to monitor such imbalances, which serve as signals that the organism’s total self-organizing system may be deviating from the preferred overall patterns, including the preferred degree of extropy. For example, when the body has been sleeping (i.e., settling into a very low-energy basin of attraction), norepinephrine, dopamine, and serotonin imbalances will eventually be triggered by a brainstem mechanism (Panksepp 1998: 109ff.) and will signal the brain that some action should be taken to get the body aroused (i.e., into a higher-energy basin). This signal occurs in the form of a homeostatic imbalance, but the reason the imbalance occurs in the first place is that the whole bodily system has settled into a low-energy basin for too long to allow its definitive holistic pattern to be maintained without a change of energy level; the holistic need to shift into the higher-energy basin then leads to the homeostatic imbalance in the subsystem of the brainstem. The brainstem subsystem is really acting as a barometer for the total dynamical system’s need to maintain its structure by reorganizing its micro-components into a higher energy basin. Some organisms are structured so as to crave more extropy than others, even though homeostasis could be achieved with less extropy. A three-month infant will begin to explore its environment in some minimal ways even though its needs for comfort, security, and nurturance could well be met simply by continuing to eat and sleep in the same pattern that served it so well for the first three months (Thelen & Smith 1994). During these first three months, the homeostatic needs were served at a very low level of extropy. But at a certain point, maturation leads to new patterns of brain organization in which the thalamus is triggered to sense that something is wrong if there is not enough exploratory activity. Here again, although the signals that indicate the disturbance are the immediate results of homeostatic imbalances within a subsystem – the hypothalamic-brainstem subsystem, which senses the chemical results of too little brain activity – the craving for more brain activity would not need to be there in order to serve the purposes of homeostasis. The infant





Curious Emotions

has matured into a structure in which the same homeostasis that could have been served by less exploratory activity is now served by more of it. Another way of looking at this point is to remember that the demands for extropy do not contradict the organic- and ultimately inorganic-chemistry principles that govern the micro-components of the system. The system is merely structured in such a way that it often prevents the fulfillment of energy reduction for the micro-components except by means of a settling into different basins of attraction at different energy levels, and systems that are structured in very complex ways make it impossible to achieve homeostasis in a subsystem without seeking a higher degree of extropy than the one that happens to obtain at the time. So the failure of homeostasis in one subsystem – for instance, the hypothalamic-brainstem neurotransmitter activities that regulate sleep and wakefulness – may signal a failure of a suitable degree of extropy for the entire organism. As another example, a person on vacation might find it quite pleasant to just sit around and eat cookies for the first few days. At some point, the hypothalamic-brainstem subsystem will alert the organism that something is off: there is too much energy being consumed and not enough complexity of activity. Here again, the signal for this condition is that increased dopamine, norepinephrine, and other activity-inducing neurotransmitters are released. The organism wants to exist at a higher degree of complexity, which requires a higher energy level. But the interesting point for our purposes is that some types of animals will reach this point earlier than others, because their organisms are structured in such a way as to demand extropy more frequently. Cats can sit around for hours doing nothing – a feat that is very difficult for a fully awake human. It is not merely that every organism wants extropy simply as a means to the end of homeostasis; rather, some organisms want extropy for its own sake more than others, even though homeostasis can be achieved at many different levels of extropy. So the fact that a failure of extropy may be signaled by a homeostatic imbalance should not be taken as implying that extropy is only a means toward the ultimate end of homeostasis. Extropy is the demand to find homeostasis by means of various different basins of attraction reflecting various different energy levels. It is also true that natural selection favored organisms with exploratory systems and other extropic tendencies, for at least three main reasons: (1) Such organisms are more likely to find food in the course of their extropic exploratory activity. (2) Such organisms are likely to be more intelligent because of their curiosity and seeking orientation. And (3) they are likely to learn

Chapter 3. Non-consummatory motivations

things incidentally in the course of their play and exploration that will help them survive. But the fact that extropically-orientated behavior may have provided a selection advantage does not erase the fact that, for the individual organism, extropic behavior (play, seeking, etc.) is not derivative from consummatory motivations by means of reinforcement of primary consummatory drives, which would make the extropic behavior into a mere secondary reinforcement. On the contrary, the extropic behaviors seem to be primarily reinforcing in organisms at our evolutionary level. I.e., they are reinforcing regardless of whether they are necessary for us to maintain energy-efficient homeostasis, or even the best way to do so. They are one way of maintaining homeostasis, while at the same time satisfying a primary motivation to maintain the homeostasis at a suitably complex and high-energy basin of attraction. It is tempting to tacitly assume that negentropy is the highest level of organismic activity that is compatible with the lowest level of scientific explanation, i.e., the inorganic-chemistry/subatomic-physics level, and therefore that anything that does not fit that level of explanation is unscientific. Since the existence of negentropy in the behavior of organisms can be explained by positing that it consists of merely instrumental behaviors in the service of primarily consummatory motives, micro-reductionists of the Twentieth Century tended to assume that the notion of a primarily non-consummatory motive is incompatible with a scientific understanding of how our systems work. They assumed that all behavior is explainable as derivative directly or indirectly from consummatory motives, and that no other type of primary motivation is consistent with our understanding of the lowest level of scientific explanation. Thus those who are prone toward the ideal of intertheoretic reduction in the sciences often want to think that all apparently non-consummatory reinforcements are only secondary reinforcements, learned through reinforcement of primary drives, which are consummatory. It is true, of course, that one way to respond to micro-reductionism here would be simply to deny that intertheoretic reduction in the sciences is possible. But my point is that it is not necessary to deny this in order to establish that some behavior is primarily non-consummatory. One advantage of leaving open the question of intertheoretic reduction is that, when accounts at one level of science can be made at least compatible with what is going on at another level (even though this is not always possible – e.g., quantum theory does not seem compatible with relativity theory), our understanding at both levels is often enhanced. The explosion of biochemical manipulation of behavior and affect through drugs is an example of such intertheoretic bridging, although of





Curious Emotions

course the appropriateness of the practical application of such treatments must always be carefully evaluated. The maintenance of homeostasis can be accomplished at higher or lower energy levels, and some organisms are organized in such a way that they prefer to do it in more complex patterns, and thus at the higher energy levels, a good percentage of the time. But even in these organisms, there are ups and downs, where we sometimes feel a need to pull ourselves out of the lethargy of too much dull comfort, even though the dull comfort itself might be quite consummatorily satisfying. But this point also needs to be compatible with what we think when we put on our “scientific” hats. Within the respiratory energy cycle, there can be lengthy digressions (as in primates) where huge amounts of energy are diverted to other purposes, and then replaced by making more ATP later; or there can be fewer digressions (as in lower animals and plants), where less energy is diverted to other purposes, and so less energy needs to be replaced by making more ATP. Thus animals, for example, need to use aerobic respiration to fuel all these additional energy digressions, whereas plants can use anaerobic respiration, which yields less energy, but is perfectly adequate for plants, because they need less energy for other purposes. Plants remain closer to equilibrium than animals, and we primates deviate further away from it than lower animals. What extropy is explicitly opposed to is not homeostasis, but simply stasis per se. It could thus be argued that extropy requires considerable complexity in a system to counteract a tendency to settle into conditions too close to equilibrium. Many theorists would be content to grant that extropy is opposed to energy conservation, not because there is an independent extropic tendency in very complex dynamical systems, but simply because natural selection, operating separately from the mechanisms and motives of any individual organism, created conditions favorable to the survival of beings whose subsystems just happen to inhibit each other’s basic aims, so that such systems would be forced to subsist in a state very far from equilibrium. For example, such a view might agree with Schopenhauer (1956) that it would be better from the standpoint of the individual organism if sexual satisfaction could be attained without the complexities of courtship, bonding, and commitment to offspring, but natural selection has engineered our bodies in such a way that we cannot get the personal satisfaction desired except by making certain individual sacrifices in the interest of the species. I.e., we are structured so that we instinctually feel compelled to do things contrary to our individual best interests in order to get what we want – the satiation of our consummatory needs.

Chapter 3. Non-consummatory motivations

But this kind of “just so” explanation leaves many unanswered questions about how such systems could ever come about – what its tangible mechanism is. In a similar way, we might explain how a certain baseball pitcher throws an extremely effective knuckle ball by pointing out that he would not have survived in the Major Leagues without a good knuckle ball. At a certain level of theorizing, this is true, but it does not explain the real mechanism of the knuckle ball. It is at this more tangible level of explanation that the theory of complex dynamical systems is needed for an understanding of extropic motivational tendencies that are not reducible to energy-efficient homeostasis alone. Such an independent principle would seem to call for a further dimension of motivational explanation besides consummatory reductionism. Even though consummatory reductionism allows that there can be an overall gain of systemic energy through endergonic reactions and feedback loops, and that complete equilibrium is always unattainable (until the system dies), it still holds that at each step the system is only doing whatever it can to move as close to equilibrium as it can – somewhat like a donkey chasing an unattainable carrot that has been permanently dangled in front of its face. On this microreductionist account, systems would presumably push, in net terms, toward the lowest energy level compatible with homeostasis. If there is an extropic principle, it would oppose or counteract the tendency to reduce free energy, by allowing the system to be driven not only by components outside of the system as they pursue their own exergonic aims, but also by the extropic structure of the system itself, which causes it to go out and actively seek to appropriate components that can serve to maintain it. This notion of extropy opens questions that we must now try to answer. How can it be linked to motivation and emotion at the phenomenological level? How can it be grounded in specific neurophysiological findings, and at the same time integrated into a plausible and applicable scientific theory of self-organization in complex systems? To move first toward an answer to the first of these questions, let’s return momentarily to Freud’s “death instinct” problematic.

. The humanistic notion of “life wish” Freud from 1925 sought a principle very similar to what we are calling “extropy,” and expressed increasing doubts about the possibility of explaining all drives as derivative from the tendency toward free energy reduction in the elec-





Curious Emotions

trostatic behavior of the chemical components of the organism. Freud regarded the sexual instincts as anti-reductive in the sense just mentioned: The sexual instincts. . . operate against the purpose of the other instincts, which leads, by reason of their function, to death; and this fact indicates that there is an opposition between them and the other instincts. . . . One group of instincts rushes forward so as to reach the final aim of life [i.e., death] as swiftly as possible; but when a particular state in the advance has been reached, the other group jerks back to a certain point to make a fresh start and so prolong the journey. (Freud 1925/1959: 74–75)

The behavioral and motivational tendencies of which Freud speaks here (sexual “instincts”) presumably would occur at a higher level of organization than the chemical reactions that make up the system. In general, some such motivational principle would be consistent with the notion of “extropy,” even if the “sexual instincts” turn out to be the wrong place to look for it. Even though each reaction seeks to minimize free energy at that particular step, the complexity of the system’s organization can sometimes oppose this aim in net terms, and can seek to settle into more complex and thus higher energy basins of attraction by preference over simpler and lower-energy ones. The organism can then prefer energizing, exploratory, or playful behaviors over those that would lead to consummatory satiation and rest; and it often shows this preference even when we are tired or hungry, as during the grueling last quarter of a football game. If the very structure of a suitably complex system favors extropy in this sense, then exploration, play, nurturance, and other non-consummatory pleasures are primary reinforcers, not derivative through learning from the purely consummatory drives or from association of such behaviors with pleasures and pains at a simpler level of organization. Otto Rank (1924) and Rollo May (1969) developed this view more fully in terms of what Rank referred to as the “life wish” and the “death wish,” with a correlative “life fear” and “death fear,” whose conflicting tendencies make for the more existentially meaningful aspects of the emotional life, such as a desire to explore, to create novelty, and to actualize the more interesting potentialities of the self even if at the expense of considerable discomfort or pain. Abraham Maslow (1974) consolidated and popularized what we might term this “anti-consummatory-reductionist” way of thinking by positing two entirely different motivational tendencies, the “deficiency” motivations and the “actualization” motivations – the latter resulting from the expression of selforganizing tendencies in an explicitly neo-Aristotelian sense. Maslow credits Aristotle with the insight that “pleasure” is not always a matter of consum-

Chapter 3. Non-consummatory motivations

ing something (filling a “deficit”) or achieving any end state, but rather can be enjoyed as an active and self-motivated process of expressing the natural tendencies of one’s own pattern of being, even if at the expense of the reduction of consummatory drives. Maslow’s antithesis builds on Robert White’s definition of a drive as a “physiological deficit external to the nervous system, and perceived by the nervous system as noxious” (White 1959: 298). For Maslow, not all motivations are driven by “physiological deficits external to the nervous system” nor by the fear of any form of “noxious” outcome. White’s view may be still more extreme: His argument is that the notion of a “non-consummatory” drive makes no sense by this definition of “drive”; if a drive refers to a deficit external to the nervous system, then all drive-reductive motivations are “deficiency”-driven in Maslow’s sense. Thus the notion of a self-actualization motivation could consist only of a non-drive-reductive tendency per se – in a sense, an even more radical notion than Maslow’s. White’s definition of “drive” highlights the tendency of consummatory reductionism to think of motivations and emotions as arising from the organism’s needs. The needs in turn are for ways to rectify a bodily situation that has moved too far from equilibrium. Without food, for example, there is nothing to fuel the Krebs energy cycle, and as a result, its initial reactants and final products are too far from equilibrium. The “deficit” involved is a deficit of something that could help the system move closer to equilibrium, even though complete equilibrium can never be reached, since some of the terminal products of the energy cycle would be directed for other purposes than to inhibit its initial step. The humanistic attitude toward consciousness and emotion has regarded drive-reductivism as inadequate to address the complex moral issues and questions about ultimate meaning that conscious beings face – not because there is so much more yet to be learned about the various ways that “basic” emotions can be mixed and modified, but because if the “basic” emotions themselves are consummatory-drive reductive in the sense just discussed, then in principle no non-reductive tendency can ever be derived from them. Humanism also has tended to shy away from equating consciousness in general with its physical substrata, because folk psychology, like phenomenology, suggests that empirical observations alone cannot tell us what it is like to experience another’s consciousness directly (Husserl 1913; Ellis 1986). This principle was later discovered by analytic philosophers of mind under the title “the knowledge argument” (Jackson 1986): If consciousness were literally the same as its physical substrata, then complete knowledge of the substrata should constitute





Curious Emotions

complete knowledge of what the consciousness feels like to its subject; but it does not. How to resolve this problem remains an open question for physicalism, and further reinforces the humanists’ alienation from physicalist accounts of the emotions, if “physical” is taken to mean “empirically observable.” More generally, existential and phenomenological approaches are intent on avoiding Husserl’s (1913) “natural attitude,” which was discussed in our introduction. Importantly, Husserl distinguished intentional objects in consciousness from physical objects in the world that we may theorize to be the causes of the brain events that can affect our intentional consciousness. The objects of which we are conscious – i.e., phenomena as they appear to us – are not necessarily the same as the physical objects as they would exist independently of us, and it is quite arguable that they are never the same (Hutto 2000); yet, according to Husserl, naturalistic physiology and psychology tend to conflate physical and intentional objects, speaking as if it were self-evident that a stimulus “caused” the perceptual consciousness of it, and as if an emotional stimulus simply “caused” an emotional response. The resulting theory made human beings into robots or zombies (Chalmers 1995) whose responses are merely passive reactions to an input without any causal power attributed to the intervention of subjectively experienced intentional meanings (Jackendoff 1996; Ellis 1996a). Existentially oriented philosophers and psychologists therefore tended to gravitate to the view that physiological psychology could never adequately connect with the higher moral and intellectual lives of people, and would inevitably produce pictures of human nature too simplistic and brutish to satisfy the philosophical concern with meaning in life, or to accurately reflect the subtlety and richness of emotional meanings. In effect, we now have two opposite views of human motivation. The “drive reductive” view, for which physiology and behavioral science have amassed a vast array of evidence, holds that the aims of life are reducible to a roundabout way for the micro-elements of the system to reduce their energy level even though they must inadvertently re-energize the system as a whole in order to do so, and this reduction to the most energy-conservative means of achieving homeostasis is experienced as pleasure or the reduction of pain. What we think of as high-minded ideals are really only twitches of a satiation-driven nervous system. In the consummatory-reductionist scheme, any increase of physiological arousal that is not somehow in the service of drive reduction must be contrary to the real intentional aims of the system, since the latter are only epiphenomena of the push toward electrostatic energy reduction in the ultimate micro-components of the system.

Chapter 3. Non-consummatory motivations

The “self-actualization” view, on the contrary, is championed by ordinary folk psychology, phenomenology, artists, writers, literary critics, existential philosophers, and many practicing clinical psychologists. It holds that we have a fundamental interest in acting in such ways as to pull ourselves out of the dull comforts of consummatory satiation, and that a complete reduction to complacence would lead to an existentially meaningless and amoral somnambulism, a “wasteland” suitable only for T. S. Eliot’s “hollow men.” Is there a way to reconcile these conflicting viewpoints within a scientifically respectable yet existentially meaningful framework?

. A possible synthesis The gap between these scientific and humanist traditions has recently narrowed. Philosophy of mind, for its part, has begun to realize that information processing is not the same thing as consciousness (Chalmers 1995), and that emotional purpose and agency are essential to “enacting” consciousness (rather than merely “reacting” to input) by driving information processing in terms of the self-organizational structure of the living organism (Varela, Thompson, & Rosch 1993; Ellis 1995, 1999a, 1999b; Newton 1996; Damasio 1999; and others to be discussed below). Very new physiological evidence that does not reveal the emotions as geared merely toward consummatory satiation (Freeman 1987; Panksepp 1998; Watt 1998; Damasio 1999) and new philosophical movements within the philosophy of mind, such as the “enactive” movement, make possible a reconciliation between the long-diverging paths of physiology and phenomenology. The key to bridging the gap is a concept of self-organization. In a complex self-organizing system, the overall process is organized in such a way as to appropriate and replace its own needed substrata rather than merely being caused by their discrete interactions. This view has roots in Merleau-Ponty’s (1942) notion of psychophysical forms as well as the thinking of developmental biologists in the early twentieth century attempting to understand why an organism imposes its own patterns of organization on the material that serves as its component parts (e.g., see Bertalanffy 1933/1962). Kauffman (1993) and other recent dynamical systems theorists have synthesized this tradition into a coherent theoretical framework. Any self-organizing system has a number of basins of attraction toward which it tends, and which one it chooses is determined by the best way to maintain coherence within the system given environmental perturbations. I have argued here, with Kauffman, that in some instances, given a sufficiently complex sys-



 Curious Emotions

tem, a higher-energy basin is preferable over a lower one – thus giving the system a tendency not only toward homeostasis, but also toward extropy; if the energy efficiency of homeostasis within a subsystem, or the system as a whole, is allowed to go too far, the system will not deviate far enough from equilibrium at either end of its feedback loops, and the result is death and disintegration of the system (Kauffman 1993; Newton 2000). The reason is that, if the pattern of activity that is constantly appropriating and replacing its components can occur only at fairly high energy levels, then these are the energy levels that the system will continually readjust its micro-components and subsystems to maintain. The dynamical systems way of thinking does not disregard the push toward free energy reduction, but adds a self-organization principle to it, as an additional fact that must be accounted for. Discrete causal sequences (Kim’s “causal closure”) presuppose background conditions, and a self-organizing system is able to rearrange the background conditions so that the overall pattern can be subserved by alternative causal sequences at the micro-level (a kind of multiple realizability where the choice of realizers is determined by the structure of the larger system). This way of thinking can make sense of Jackson’s above-mentioned “knowledge argument,” because a self-organizing system acts on its own component materials and on the environment rather than merely passively reacting to them. It is easy to see why one person cannot enact another’s self-organizing patterns, and therefore cannot feel the other’s sensations or other conscious states even though these may all be observable physical processes. When the scientist Mary observes another person’s brain processes, she reacts to them consciously by enacting her own brain processes. So the subjective experience that the objective observations reveal to her is what it is like to enact her own processes – not what it is like to enact the other’s. Even though this enactment may be completely physical, what it is like is not empirically observable by someone who merely reacts to it. The self-organizing process in this case is not reduced to the sum of the externally observable components, because what it is like to enact such a process may not be observable from an external perspective, given that it organizes its empirically observable components and not just the other way around (Ellis 1999a, b, 2000a, b, d). The “what it’s like” component is the way things can appear only to the system initiating the action commands that are presupposed by any consciousness. The observing scientist also initiates action commands, which subserve her own consciousness of what is happening, but these are not the same action commands being initiated by the subject of the experiment.

Chapter 3. Non-consummatory motivations 

Obviously, not all physical phenomena can be observed from just any perspective. For example, the location and velocity of an electron cannot be observed from the same position. A rainbow appears in full color from one perspective, but only as particles of moisture reflecting white light from another. It may be that even though the components of a conscious system may be externally observable, certain aspects of their complex interrelations can be observed only from within the system. Even empirical observation depends ultimately on the subjective enactment of an experiencer, which itself does not seem to be observable from an external perspective. Yet the subjective enactment can also have some causal relevance, because it is a structural pattern that acts to appropriate its own future micro-components. In sum, a complex self-organizing system appropriates and replaces component elements, not just for the purpose of maintaining the complex homeostasis of its various chemical processes, but in some instances also for the purpose of gravitating toward higher energy basins of attraction by preference over lower-energy basins. Again, this is because maintaining the pattern across multiply realizable component-readjustments entails maintaining the energy levels necessary to that pattern. This makes possible a convergence between physiology and views of motivation that are not consummatory-drive oriented, such as those of Rank (1924) and Maslow (1974), who are at pains not to reduce human beings to complacent, robot-like creatures with no fundamental interests beyond avoiding hunger, thirst, and other discomforts. The next chapter will discuss the convergences that are now possible between physiology and these more humanistic and existential approaches to emotion, given that many motivations are primarily non-consummatory and not driven toward satiation of needs definable purely at the micro-level.

Chapter 4

Homeostasis, extropy, and boundary needs as grounding specific emotions

The previous chapter developed the notion of extropy as a tendency in very complex organisms toward settling into higher-energy by preference over lower-energy basins of attraction. We saw that there are many advantages in viewing this extropic tendency as independent of the push toward energyefficient homeostasis, and that it could help clarify the distinction between consummatory and non-consummatory motivations. To distinguish between satiation of simple homeostatic needs on the one hand, and on the other hand the self-energization and the non-consummatory energy-expenditure that occur in general arousal, curiosity, exploration, nurturance, play, and many aspects of mating activities, can help immensely in understanding the “higher” human emotions – many of which are in the list just cited. We shall see that the higher emotions are not learned through reinforcement of the consummatory ones, nor merely indirect means to consummatory gratification, and are not merely the results of a socialization process. On the contrary, they are just as fundamental in the constitution of animals at a high degree of complexity as are the consummatory motives. In this chapter, we shall not only flesh out the neurophysiological and phenomenological emotional correlates of the requirements for extropy and homeostasis in complex animal systems, but also add an additional requirement – the need for boundary protection. A self-organizing system needs to maintain itself within secure boundaries, although the boundaries are osmotic and interactive with the environment (De Preester 2002). In the self-organizational framework, emotions are identified in terms of the aims and activities of the organism rather than in terms of specific perceptual or proprioceptive triggering stimuli. But environmental conditions can be either facilitative or thwarting toward self-organizational aims. Internal conditions too can be either facilitative or thwarting. For example, physical fatigue does not facilitate exploratory behavior, but thwarts it. So, given the three types of aims in the most complex organisms – homeostasis, extropy, and bound-

 Curious Emotions

ary protection – and given that any of these three aims can be facilitated or thwarted by either internal or external conditions, we can roughly classify all primary emotions in terms of regularly occurring affective qualities that go with the thwarting or facilitation of each of the three types of motives by either internal or external constraints. For example, anger directs itself primarily toward external conditions that thwart boundary protection needs and that relate indirectly to boundary protection. Love is directed toward specific others who facilitate the expression of the social bonding, play and/or nurturance motives, which are extropic. Separation distress is directed toward environmental conditions – loss of experienced connection or contact – that deprive us of the expression of these motives, thus thwarting them. For each of the three basic requirements of the system, both internal and external conditions can make positive and negative contributions. So any list of “basic” emotions must include more than the usual four or six. Any variation in any of the facilitative or thwarting aspects of either internal or external conditions for any of the three basic types of self-organizational needs will stimulate an unconditioned (thus “basic”) emotional response. We must begin by laying some basic neurophysiological and phenomenological groundwork. First, we need to see how the notion of “extropy” in complex systems can be correlated with empirical evidence about human and higher-animal emotional systems.

.

Physiological evidence for non-consummatory motivation

Recent physiological and neuropsychological accounts of emotion are more sophisticated than the simplistic hedonistic-behaviorist theories of the past. Physiologists of emotion like Panksepp (1998), Watt (1998, 2000), Freeman (1987) and Damasio (1999) speak of many physiological emotional tendencies that have nothing to do with the motivation to reduce consummatory impulses – emotions geared toward play, exploration, bonding, empathy, and other non-consummatory activities that cannot be derived from learning based on reinforcement in terms of consummatory drives. “Consummatory” in this context means oriented toward filling a “deficit” that stands in the way of energy-efficient homeostasis through submolecular energy-reduction. In non-consummatory behavior, there is no prior deficit, but simply the desire to engage in patterns of activity consistent with a complex system’s tendency to maintain an optimal level of both homeostasis and “extropy.”

Chapter 4. Homeostasis, extropy, and boundary needs 

Freud’s concern with this point parallels Kauffman’s later distinction between open thermodynamic systems at the edge of chaos on the one hand, and closed and “frozen” systems on the other: A closed or frozen system moves quickly toward death. As Newton (2000) explains, In closed systems, isolated from the environment, the group [of entities] eventually succumbs to entropy: the interactions are random and the group is in a state of stable equilibrium – a state that does not change with time. In open systems, by contrast, there are three general possibilities with regard to order. First, the system could become chaotic: the interactions within the system could occur with increasing randomness. The system would eventually reach total entropy, or equilibrium with the environment, and disintegrate because of environmental invasions (such as when a leaf decomposes in a compost heap). When an organism is at equilibrium with the surroundings, it is dead. The second possibility is that the system could be frozen in a single state in which interactions cease altogether (a diamond is such a system). Third, the system could reach a sustained, nonequilibrium state of homeostasis, in which stability of the system as a whole is maintained by means of continual adjustments to environmental perturbations. (Newton 2000: 92–93)

Freud, however, if he spoke the language of dynamical systems, would want to go a step further. He would ask whether all nonequilibrium systems maintain homeostasis with the lowest possible energy expenditure compatible with homeostasis, or whether some systems prefer higher-energy basins of attraction over lower-energy ones that are equally capable of maintaining homeostasis. On my interpretation of Freud’s later view, he would prefer the latter alternative. Let’s look at some neurophysiological data that are consistent with this idea. Panksepp’s research on the physiology of emotion suggests that the mammalian organism is organized to self-trigger its own energization mechanisms. “The basic waking mechanisms of the brain do not require sensory input from the body to sustain arousal” (131). Specifically, “[NE, DA, and Serotonin] neurons contain internal pacemaker mechanisms to maintain spontaneous activity requiring no incoming influences. . . . Clearly these brain systems control holistic aspects of brain functioning rather than discrete behavioral processes” (109–110). Thus emotions in Panksepp’s view intend holistic balance, not discrete chemical reactions or rigid behavioral outcomes. The behavior of laboratory animals may sometimes seem rigid, but only when an animal’s choices have been artificially constrained to the point of excluding all but one useful behavior for the organism’s overall purposes. When Pavlov’s dogs’ options were continually constrained to this extent, they became “neurotic” (Pavlov’s

 Curious Emotions

word) and thus “no longer useful for our purposes” (Pavlov 1929). According to Panksepp, “Psychiatric disorders result from neurochemical imbalances (i.e., lack of regulation) among many transmitter systems as opposed to a pathology in a single one, so there may be many ways to restore overall balance” (117). Panksepp also regards “seeking” as endogenous and spontaneous, not derivative from learning and reinforcement. “The extended lateral hypothalamic (LH) corridor [with ascending DA circuits and descending glutamaturgic circuits]. . . responds unconditionally [i.e., without any previous learning or conditioning] to homeostatic imbalances. . . . This harmoniously operating neuroemotional system drives and energizes many mental complexities that humans experience as persistent feelings of interest, curiosity, sensation seeking, and, in the presence of a sufficiently complex cortex, the search for higher meaning” (145). Thus there is physiological evidence for the emergence of non-consummatory activity as an intrinsically (as opposed to instrumentally) valued activity of the organism. Panksepp goes on to emphasize that “this is a diametric reversal of traditional behaviorist thinking. . . . The affective state [that serves as reinforcer in some instances]. . . is not simply ‘pleasure’ but a highly energized state of psychic power and engagement with the world” (147–150). While “pleasure” and “reinforcement” are not synonymous for behaviorists, any behaviorist learning theory must presuppose a distinction between consummatory and instrumental behavior; the non-behaviorist element in Panksepp’s system is that there are entire categories and systems of complex behavior that do not involve, even indirectly, any “consummatory” reward in the sense we have been discussing. It is true that Panksepp does not distinguish the homeostatic motives from an explicitly conceptualized extropic dimension thematized in terms of dynamical systems theory. But Panksepp’s opposition to behaviorism calls for such a distinction, because without it, the distinction between consummatory and non-consummatory motivations itself remains unclear. This problem will be discussed more completely later in this section. The empirical research of Panksepp and his associates into the brain correlates of mammalian emotions relates patterns of mammalian emotional behavior to widely-distributed yet clearly distinguishable brain systems, based on multiple measures of brain activity, assessments of neurotransmitter activity, and neurological symptoms. The theoretical organization of the findings postulates seven distinct brain systems, subserving seven different motivationalbehavioral tendencies in the mammalian organism. Each brain system emphasizes a different combination of brain regions and a different combination of neurotransmitter pathways, and is initiated by brainstem and limbic mon-

Chapter 4. Homeostasis, extropy, and boundary needs 

itoring of the organism’s deviation from homeostasis. These systems can be initiated also by sudden hypothalamic and thalamic deviations that can be caused by amygdala, hippocampus, basal forebrain, and brainstem responses to both conditioned and unconditioned stimuli. At some risk of oversimplification, the seven motivational systems can be summarized as follows (I follow Panksepp’s convention of designating the systems with uppercase letters): 1. The seeking system (which Panksepp earlier had called the expectancy system) energizes the body toward exploratory and generally appetitive behaviors; this brain system emphasizes a circuit from the ventral tegmental area (VTA) to lateral hypothalamus, and heavily relies on dopamine and glutamate. 2. The rage system, from the medial amygdala to bed nucleus of stria terminalis and medial hypothalamus to periaqueductal gray (PAG), emphasizing Substance P and ACh activity. 3. The fear system, from central and lateral amygdala to medial hypothalamus to PAG and pons, with neurotransmitter activity emphasizing glutamate, monoamines and many neuropeptides (e.g., DBI, CRF, and CCK). 4. The nurturance/sexuality system, from the bed nucleus of stria terminalis to medial amygdala to preoptic hypothalamus to PAG, using steroids, vasopressin, oxytocin, and prolactin. 5. The separation distress/social bonding system, from the bed nucleus of stria terminalis to dorsomedial thalamus to preoptic hypothalamus to PAG, also including in primates the anterior cingulate and anterior thalamus, relying on opioids, CRF, and acetylcholine, as well as many of the same neurotransmitters used in the nurturance/sexuality system. 6. The play system, from midbrain central gray to posterior thalamus, relying extensively on opioids and acetylcholine. 7. The consummatory pleasure system, from PAG to hypothalamus, relying more extensively on opioids, as regulated by GABA, than do the non-consummaotry pleasures of the seeking and play systems. It is important to note that the seven basic brain systems are relatively independent of each other, in the sense that none of them is learned through conditioning by association with reinforcement of any of the others; thus only the last of the seven conforms to the assumption of early drive theorists that all behavior must be learned by association with consummatory reinforcement or association with physical pain. The seeking system seems especially resistant to that assumption, and is reminiscent of the “exploratory” drives and drives toward “mastery” proposed by revisionist drive theorists who wanted to reject

 Curious Emotions

the assumption that the organism would remain completely passive if not stimulated toward the acquisition of pleasure and satiation. There is a considerable likelihood that the seeking system is foundational for all the other systems. Panksepp’s way of approaching the causal mechanisms underlying the basic emotional systems is a “neuroevolutionary” one rather than a behaviorist one; i.e., mammalian behavior patterns evolved because they tended to be conducive to the survival of the species in the long run, even when some of the tendencies do not increase the pleasure or happiness of the individual organism. Accordingly, Panksepp’s analysis is strikingly at odds with several assumptions that many physiologists of emotion, still influenced by the legacy of behaviorism, tend to make without argument or evidence. He explicitly rejects the assumption that all behavioral tendencies are derivative through learning from the consummatory drives or association with pain. According to Panksepp, much activity is primitively non-consummatory (in the sense we have been using), and is instinctually and innately driven by the play, seeking, social bonding, and nurturance systems. Thus we cannot assume that all motivational tendencies are somehow geared toward directly or indirectly maximizing the pleasure of the behaving organism. We behave in certain patterns, not necessarily because doing so makes us happy, but because the natural selection process caused our brains to be organized in certain ways. Panksepp in no way means to deny, however, that failure to conform to instinctual motivational tendencies will cause emotional pain. For example, failure to explore the environment will cause a feeling of frustration or painful restlessness. But exploration itself is not derivative from any consummatory drive or from indirect associations through learning with simple physical pain. Rolls (1999) argues on the contrary that natural selection would favor a situation where behaviors that favor survival and reproduction would also increase pleasure or reduce pain for the organism. But this argument is circular. If we were to assume that the only behaviors in which organisms are motivated to engage are the ones that increase their level of pleasure, then Rolls would be correct in reasoning that natural selection would favor linking pleasure with survival value in all instances, so that the same behaviors that enhance survival would also enhance pleasure. On the other hand, suppose we do not assume a priori that organisms are always and only motivated to do things that enhance their level of pleasure. Then there would be no reason to think that linking pleasure with survival value would be the only way to motivate organisms to engage in behaviors that enhance their survival. In that case, some behaviors with survival value would be motivated by a desire for pleasure, and some would be motivated by other instinctually based tendencies that do not necessarily lead

Chapter 4. Homeostasis, extropy, and boundary needs 

to pleasure. This question as to whether there are genuinely non-hedonistic motivations will be considered in more detail in a later chapter. The difference between consummatory drives and non-consummatory (Panksepp uses the term “appetitive”) tendencies is that the two kinds of motives use different brain systems and different neurotransmitters. For example, consummatory pleasure relies at least as heavily on the use of morphine-related brain chemicals as it does on dopamine (DA), whereas seeking and play rely much more on DA, norepinephrine (NE) and acetylcholine (ACh) activity. Furthermore, stimulation of different parts of the consummatory and appetitive components of the hypothalamus will lead to activation of different kinds of brain chemistries, with correspondingly different kinds of gratifying behavioral tendencies and different extended brain system activations. The consummatory-pleasure nuclei of the hypothalamus (which rats self-stimulate in some of Panksepp’s research) is more morphine-reliant and vegetative/satiation-oriented; the completely different non-consummatory-pleasure, or active-engagement area of the hypothalamus (which also induces rats to self-stimulate) is more DA- NE- and ACh-reliant and activity/energization-oriented. The fact that some behaviors, such as seeking, play, and nurturance, are engaged in for their own sake leads Panksepp to reject the common assumption that emotional responses are simply caused by perceived stimuli. In Panksepp’s system, the mammalian organism is inherently motivated to engage in certain patterns of behavior, independently of the effects of any stimuli; the stimuli that do present themselves can then either facilitate or get in the way of the behavior patterns that have already been self-initiated. Many brain systems are activated by the PAG or the closely associated VTA (ventral tegmental area) of the upper brain stem; self-initiated action circuits of many kinds can be traced back to this area, and correlatively it spreads tentacle-like efferents to other brain systems. The classic symptoms of anger, for example, are produced by binding the arms of an infant, because the infant is already motivated to move in patterns that require freedom for arm movements. The initiation of this movement involves PAG and pontine activity. Part of the novelty of this approach is therefore that it sees behavior as self-motivated and self-energizing, rather than as passively responding to inputs. As another example, the brain has its own basic waking mechanism and does not require sensory input from the body to sustain arousal (Panksepp 1998: 131). This way of assessing the causal mechanisms of emotional behavior allows the organism to be seen as the agent of its own actions rather than as a pas-

 Curious Emotions

sively reacting machine. Panksepp’s model suggests a self-organizing view of information processing in living beings rather than one in which each input is merely transformed into an output in piecemeal and automaton-like fashion. The movements of the organism are initiated, coordinated, energized, and integrated by the whole animal working as the agent of its activity. There is an obvious objection that could be raised against this way of thinking. Much hinges on the distinction between consummatory and nonconsummatory motives, since a central claim is that non-consummatory behavioral tendencies such as the play, nurturance, and seeking systems are not derived through learning from indirect association with any consummatory reward. However, sophisticated behaviorist theories, which Panksepp explicitly rejects, could explain this kind of finding by granting that exploratory behavior, play, nurturance, etc. can function as primary reinforcers in a behaviorist learning system. Now traditional drive theorists can posit that there are primary drives toward exploration, mastery, etc.; but deprivation of the environmental conditions needed to satisfy these drives can be conceived of as needs; and it could therefore be argued that all behavior is still in the interest of fulfilling a need or correcting some stimulus deficit. From this point of view, it may not be so clear how Panksepp’s causal story is really different from a traditional consummatory drive theory (where even exploration is an attempt to “consume” the conditions needed to reduce the drive) – or indeed from a behaviorist learning theory in which the range of primary reinforcers is left suitably diverse. Panksepp could respond that the difference between consummatory and non-consummatory behavior is apparent from the difference in the brain systems and the neurotransmitters that subserve them. We get consummatory pleasure when we stimulate a certain part of the hypothalamus, whereas other parts produce the different pleasurable reactions associated with arousal, through DA and NE activity. But a sophisticated behaviorist can easily grant that there are different brain systems and chemical bases for different primary reinforcement consummations, while still viewing them all as consummatory. A “consummatory drive” consists simply of any chemical deficit that needs to be filled, and the filling of the deficit is pleasure. But if we accept the independence of the extropic tendency of the most complex animal systems, we can then sharpen the distinction between Panksepp’s system and those behaviorist learning theories in which complex, self-motivated behaviors are merely secondary reinforcers, derivative through learning from consummatory rewards. The difference between these two approaches has important implications for the way we understand the causal

Chapter 4. Homeostasis, extropy, and boundary needs

structures of mammalian and especially human behavior. For example, it seems obvious that not everything we humans are motivated to do is rewarding in the sense that it necessarily maximizes our consummatory pleasure, even indirectly. Also, there is a difference between the kinds of energy systems involved in a placid, merely satiation-seeking organism, and one that inherently seeks adventure, newness of experience, and complexity even when this tendency requires moving to a higher-energy biochemical pattern that can throw off the satiety of dull comfort. There are important psychological implications, as Freud discovered in his later work. To many drive-reductively oriented psychologists and physiologists, especially those committed to intertheoretic reduction in the sciences, it may seem strange that we would be motivated to seek extropy at all. The reason is that all motivated systems have self-organizing structures and are complex enough to prefer higher-level basins of attraction, as aspects of the definitive pattern to be maintained. The fact is that, even if each of our smallest subsystems (e.g., molecules) seeks the lowest energy level compatible with homeostasis, the whole still must be self-maintaining, so that for one system to gravitate toward energy-efficient homeostasis often prevents another from achieving it, and therefore the system as a whole is organized so as to prevent any subsystem from completely achieving it. This was the way Rank (1924) and the later Freud set up their “conflict” theories of motivation – i.e., theories in which conflict between the organism’s own basic motivations are inevitable. In such conflict theories, what keeps the organism going is that it cannot achieve one set of aims without creating new demands in another. Perhaps a more up to date formulation would be that, within the context of a self-organizing supersystem, there are phases that the supersystem often shifts into such that the subsystems can reach homeostasis only at a fairly high energy level, and therefore the system has to prevent any of them from settling into complete energy-efficiency. Such a highly energized state might be viewed as a “bifurcation” of the two lower-level attractors, in order to integrate their conflicting needs into a more coherent synthesis, able to avoid thwarting the demands of one attractor in order to sustain another. If everything were dictated by the tendencies that the most fundamental processes normally exhibit when not subsumed within a complex dynamical system – the inorganic chemical reactions that make up the organic reactions (e.g., chlorine reacting with hydrogen) – then those systems at that inorganic level would seek the lowest energy level compatible with homeostasis. But they could actually reach equilibrium only if operating outside the context of a self-organizing system (Kauffman 1993).





Curious Emotions

The conflict between homeostasis and the need for extropy is not a priori unavoidable, as it would be if entropy and homeostasis were equivalent terms. But the two tendencies will often tend to come into conflict, because the most energy-efficient ways to achieve homeostasis so often tend to be the ones that are most directly micro-reductive, and that enhance a subsystem’s proximity to equilibrium – eating, drinking, resting, sexual orgasm, etc. The play of young animals, for example, consumes considerably more energy than would be consumed by eating, yet with substantially less payoff in terms of any consummatory drive reduction with which the animal could associate the play behavior due to past learning. The desire to play is instinctual, and not contingent on past rewards. It is well established, for instance, that chickens literally would rather peck than eat (Wolfe & Kaplon 1941) – i.e., they prefer food that requires some effortful searching and pecking over food that is delivered to them with no effort required on their own part. It has also been known for some time that rats will cross an electrical grid for the reward of being allowed to explore new territory (Dashiell 1925; Nissen 1930), that monkeys will solve complex problems in exchange for being allowed to look out the window (Harlow 1950; Montgomery 1955), and that rats will run on a treadmill for the sole reward of being allowed to press a bar (Kagan & Berkun 1954). It is true, of course, that instinctual hardwiring toward playful behavior has a general selective advantage, since many survival skills are developed through play; but the animal does not know this in advance, and thus its playful behavior is not contingent on the rewards that much later will follow very indirectly from the playful learning. The animal values the play intrinsically – for its own sake – rather than instrumentally. One might say that rats at play, or in their seeking or nurturant modes, operate from a type of enjoyment that is more Aristotelian than modern: For Aristotle, pleasure is taken in the process of actively exercising one’s capacities, rather than in the achievement of a static end state. For the athlete, for example, the value of winning is partly instrumental – to provide a game structure to make possible the playing of the game, and also to earn the privilege of playing again tomorrow. The athlete not only plays in order to win, but also wins in order to play. The playing activity is valued intrinsically, for its own sake. The same is true for experimental rats, in spite of their considerable domestication. It is often assumed that emotions are simply reactions to discrete stimuli – that an insult causes anger, which in some sense basically intends behavior automatically geared toward the infliction of injury (unless repressed or sublimated). On the self-organization view, by contrast, the aim of all emotions is

Chapter 4. Homeostasis, extropy, and boundary needs

to maintain or restore homeostasis at a suitable level of extropy, and the relevance of discrete stimuli is merely that they perturb the organism’s already ongoing intentions. The behavior of the organism is self-moving, and objects either facilitate or get in the way of that ongoing movement, with consequent adjustments intended to preserve homeostasis within extropy. But it is incumbent on the self-organization theorist to work out the behavioral details of this approach for specific responses like fear and anger. In order to do so, we must make the notion of self-organization itself more physiologically concrete. When the brainstem emotional systems, with the help of the thalamus and hypothalamus, register that the complex self-organizing system of the body has either deviated from homeostasis (for example as in hunger or thirst), or has settled into a state too close to equilibrium (as when the organism has been sleeping for too long), the neurotransmitter nuclei in the midbrain, pons, medulla, and other emotional areas release neurotransmitters that are then propagated along neural fibers to the cerebellum, motor cortex, hypothalamus and thalamus to gear up the brain to command the body to take action (Faw 2000; Damasio 1999). Still other brain areas – the amygdala, cerebellar cortex, thalamocortical loops, and supplementary motor area – are used to compute what kind of action is likely to be effective in restoring the desired balance (Schmahmann 1997). For example, disgust motivates the organism to distance itself from the object with the chemical properties that disturb its own homeostatic chemical patterns. Any sudden environmental change will trigger at least some of these action-oriented emotional responses to a greater or lesser extent (Faw 2000). But most sentient organisms are structured so that specific kinds of changes will trigger them more than others. In very sophisticated animals, such as mammals, new stimuli can be conditioned to trigger these responses to an extreme extent, and the organism can pre-tune itself to be either hypervigilant or relatively oblivious to the same stimulus at different times, depending on subtle environmental contextualizations. For example, the same noise that will trigger an intense fear response for someone burglarizing a house may not trigger any fear for that same person if he is merely a guest in that same house (see Ellis 2000d). LeDoux (1996) shows that, when a sudden novel stimulus is presented, the amygdala is the first brain area to get input screened by the thalamus (see also Damasio 1999; Goleman 1995). Like the amygdala, the hippocampus is also activated very early after presentation of a novel stimulus, showing enough activity to produce an event related potential at least as early as 18 ms after presentation of even an unanticipated stimulus (Faw 2000). The cerebellum also





Curious Emotions

has been shown to be activated at about 20 ms (Woodruff-Pak 1997). This subcortical activity occurs long before the visual cortex can determine the identity or specific features of the unanticipated stimulus. Visual processing of a novel stimulus occurs somewhere between 100 and 200 ms after presentation, and we are not conscious of such an object until about 300 ms (Coles et al. 1990; Aurell 1989). Long before this cortical processing occurs, the amygdala produces a startle response to anything that does not feel either familiar or like “good” novelty. Notice the important role of the “feeling of familiarity” here. I argued in Questioning Consciousness (1995: Chapter 6) that a feeling of familiarity is essential to explicit recognition or recall memory. The amygdala is immediately adjacent to the hippocampus, which is especially sensitive to whether things feel familiar. So a mature, psychologically healthy person’s amygdala will not be startled by something that feels familiar or like good novelty, because the hippocampus always already knows what feels familiar, and the amygdala is primed only to produce the startle response to what does not feel familiar (or like good novelty). But consider a person without good hippocampus training – e.g., someone whose parents’ behavior was so completely unpredictable that they acted playful and loving one minute and brutally angry the next, for no discernable reason: this person’s amygdala will continually produce a startle response to things that are perfectly familiar, because even familiar things are unpredictable. In this person’s infancy, the familiar was almost as likely to pose a sudden threat as was the unfamiliar (Zachar 2000). The person’s amygdala therefore cannot be discriminating and make use of the hippocampus’s ability to read situations that feel either familiar or like good novelty. We see, then, that when and how a basic anger or fear response will be triggered by a given stimulus is always already contextualized in terms of the dynamical interrelations of various organismic systems operating within a total project of maintaining homeostasis within extropy. All of the recent physiological observations discussed in this section lend themselves to a self-organizational description based on the often conflicting tendencies toward an energy-efficient homeostasis on the one hand and extropy on the other. In the normal course of experience, contrary to the presumptions of many behavioristically influenced and consummatory-drive oriented theorists, environmental novelty is often craved because of our desire for extropy. But as Freud and Rank were ahead of their time in noticing, this craving for novelty, adventure, unpredictability, and higher-energy forms of experiencing often conflicts with our desire for homeostatic energy-efficiency,

Chapter 4. Homeostasis, extropy, and boundary needs

satiation, predictability, and Freud’s “inertia,” toward which our consummatory drives toward dull comfort would impel us if not counterbalanced.

. Novelty, constraints to freedom, and the action-consciousness connection Any newly presented stimulus consists of a change, yet since the organism desires to maintain continuity in its successful patterns of homeostasis, it seems reasonable to assume that the basic physiological emotional tendencies are always addressed essentially to the issue of novelty in one way or another. Since any change could pose a threat to homeostasis, fear and anger are directed negatively at sudden new stimuli. But feelings toward novelty can be positive as well as negative. Panksepp’s fundamental “seeking” tendency positively craves novelty. These emotional interests toward novelty work in conflicting ways. Since our chemical self-organizing systems are motivated toward preserving homeostasis but without lapsing too much into inertia, it follows that we want to seek out novelty as a way to increase extropy (see Ellis 1986); but at the same time, novelty can be a threat to the integrity of the system, especially vis a vis maintaining homeostasis. Here again, extropy and homeostasis, though not always in conflict, are different motivational tendencies, because homeostasis can be achieved at either a higher- or a lower-energy basin of attraction. Since the hippocampus is especially sensitive to familiarity versus novelty, and is adjacent to the amygdala, it makes sense that boundary protection activities, as in anger and fear, will be triggered by unusual or suddenly changing environmental conditions (which lack a feeling of familiarity), and that the resulting emotional systems will be activated very quickly. Since the amygdala and hippocampus are the earliest areas to be alerted by the thalamus receiving new input, the organism will be pre-tuned to look for subtle patterns of unfamiliarity and novelty, which are picked up sensitively with the help of the hippocampus, long before perceptual processing is developed occipitally. The organism can initiate action commands in relation to patterns of familiarity or novelty (for example, potentially threatening affordances) before the physical properties of an object can be analyzed (LeDoux 1996). But there are also good novelties that are sought (curiosity, exploration), and the hippocampus, thalamus, and amygdala surely will be much involved in those cases as well. This prediction is consistent with Panksepp’s (1998) finding that the play system uses many of the same brain areas as the fear and anger systems, but





Curious Emotions

with subtle pretuning for patterns of novelty with negative affect and very different actions. The brain systems in which homeostatic and extropic chemical issues for the organism as a whole are translated into neural (electrochemical) ones – primarily, the brain-stem/midbrain/limbic system – facilitate the emergence of self-organization, if sufficiently complex, into consciousness; this process requires efferent commands directed toward action affordances, as we saw in Chapter 2. According to Jeannerod (1994), action commands when inhibited give rise to action imagery. Action itself may also require “action imagery” just prior to initiating the movement (Libet 2000), but we are relatively more conscious of those images when the movements are not actually executed. Since understanding an object in consciousness requires at least implicitly understanding its action affordances (Newton 1996), this kind of action imagery is implicitly present in all consciousness, and ultimately must be emotionally motivated by the organism’s general desire for homeostasis at a suitably extropic energy level. Consider verbal thoughts, for example, which according to Joseph (1982) and Tucker (1981) are a truncated form of imagined speech, with the action commands to the Broca speech area inhibited by prefrontal activity. If Broca’s area is damaged, the person may still be able to imagine speaking, and thus think silent verbal thoughts, but if Wernicke’s area is damaged, the action command will already have been garbled by the time it reaches Broca’s area. This notion that all consciousness presupposes action commands will be further developed in the next chapter. Besides extropy and homeostasis, there is a third fundamental concern for any motivated being. Any self-organizing entity, if it is to remain an entity, must protect the boundaries within which it is free to act in its preferred patterns. Fear and anger responses are not simply blindly and automatically caused by the presentation of a stimulus; they are caused by the organism’s attempts to continue its desired activity patterns into the future, as modified by perturbations in the environmental conditions needed to do so. We have already seen that the classic paradigm of anger, as measured by all its various physiological indices, is produced when the arms and legs of an infant are bound (Panksepp 1998). The purpose of the organism in this case is not to inflict injury or destruction, but simply to break free. The movements toward which the organism is energized consist of pushing, lashing out, or flailing around. The same behaviors are observed any time the infant’s free, self-directed movement is opposed or impeded. It is interesting to note that this same response gives infants the ability to swim (Dalton 2000); the inability to breathe, perhaps the most confining of all feelings, triggers the flailing behavior. As a young animal

Chapter 4. Homeostasis, extropy, and boundary needs

develops, this general lashing out may lead to destruction or warding off of a predator. Later, the movements involved in breaking free may take on symbolic functions, so that an adult may express anger by behaving as if in the position of a confined infant. It is especially important to realize that the predatory behavior of animals does not develop as a modification of this infantile anger response, but rather from two separate instinctual systems, the seeking and the play systems (Panksepp 1998). An animal hunting is not distracted by angry symptoms, but is focused, controlled, gracefully poised, and in a state of appetitive enjoyment reminiscent of its play with other young animals of its own species. When play or hunting leads the animal into situations that threaten or confine it, on the other hand, then anger may be evoked as a secondary appendage to the initial playful and seeking motives. In many instances, the prey is unthreatening enough that no such anger responses need be evoked. At the beginning of the hunting behavior, at least, the animal’s motives are similar to those in play with other young animals of its own species – strengthened of course by hunger but not requiring hunger per se – until the final kill, which behaviorally already resembles an eating or consuming behavior. Given these boundary-protection needs, sudden environmental change not only can be positive, negative, or neutral depending on contextualization, but it can also be both intrinsically rewarding and threatening, because the organism has potentially conflicting aims – maintaining complexity at relatively high levels of energy, which systematically are preferred over lower levels of complexity, but at the same time maintaining homeostasis and protecting the boundaries of its inner-directed activity against injury or confinement – conditions that interfere with the organism’s ongoing activity toward homeostasis at a suitable level of extropy. Chronic “thrill seekers” are especially aware of these conflicting emotional responses to shocking, sudden changes. Moreover, each of the conflicting tendencies can be either facilitated or opposed by environmental conditions, which can threaten the boundaries of activities in the service of both homeostasis and extropy, or can offer useful affordances for them. Extropy can be facilitated by interesting environmental conditions free of constraint, or it can be opposed by conditions that do constrain the desired pattern of activity. And the need for homeostasis can be opposed by environmental insult to the integrity of the organism, especially by violating its boundaries to cause injury or illness; or this same need can be facilitated by the environment when it offers sustenance and conditions conducive to rest.





Curious Emotions

The interactions of these organismic and environmental variables means that a sufficiently complex living organism (one that is a highly complex dynamical system) will tend to exhibit at least three fundamental motivational needs, based on three inescapable overall purposes of the system: (1) extropy, (2) protection of the boundaries for its activities, and (3) homeostasis. It is difficult to separate (2) and (3) conceptually, since in a sense there must first be boundaries before homeostasis becomes an issue. For this reason, boundary issues have a special status: The “inside/outside” distinction does not yet apply to it, since without boundaries there is no meaningful inside or outside. So, while extropy and homeostasis can be either facilitated or thwarted by either internal or external conditions, boundary needs must be met before this internal/external distinction is possible. Nonetheless, it seems inescapable that organisms do have boundary protection needs, and even after those have been met there are still homeostatic and extropic concerns to be addressed in the interest of self-organizational balance and maintenance of sufficient complexity. Correlatively with these three purposes, there is a natural tendency of the organism to seek novelty and complexity, as a way of affirming the intrinsic value of its activity itself (not just instrumental value toward accomplishing other aims, such as tension reduction). But the organism also tends to avoid having freedom constricted/imprisoned (which involves our agency and makes us angry if too constrained – the object is to break out of the constriction). Constriction prevents us from freely acting to seek expression of active tendencies. Thus anger is a boundary-protection issue. Protecting the boundaries of the organism leads also to fear or suspicion of novelty. Physiologically as well as phenomenologically, there are very close interrelations between fear and anger (Panksepp 1998). Independently of these extropic and boundary issues, there is simultaneously also a desire for consummatory pleasure/comfort which, as we have seen, is also equilibrium-driven and homeostatic (by contrast to playful enjoyment of our activity per se, which is extropic). The three types of self-organizational demands – extropy, boundary protection, and homeostasis – are interdependent and not neatly separable. Each of these overall tendencies can further be divided according to (a) internal and (b) external conditions that affect their expression, and these effects can be either positive or negative, depending on whether the affordances presented by the relevant conditions (internal or external) are facilitative or thwarting. Based on this division, we can construct a list of categories of affects that reflect motivations that will influence the behavior of any highly complex dynamical organism, as summarized in Table 1.

Chapter 4. Homeostasis, extropy, and boundary needs

Table 1. Affects in conscious organisms relative to self organizational needs Safety, Confidence, Comfort (if boundaries fluid yet secure) BOUNARY NEEDS

vs. Anxiety, Anger, Fear, Traumatic Pain (if boundaries not fluid and secure)

Grounds a relative distinction of INTERNAL CONDITIONS

EXTROPIC NEEDS

PURELY HOMEOSTATIC NEEDS

Seeking, Inspiration Curiosity (if int. conditions good) vs. Lethargy (if int. conditions bad) [dependent on earlier int./ext. relations] Satiety (if int. conditions good) vs. Appetitive Desire – Hunger, Thirst, Illness Symptoms, etc. (if int. conditions bad) [dependent on earlier int./ext. relations]

vs.

EXTERNAL CONDITIONS Play, Bonding, Exploration (if int./ext. relations good) vs. Restlessness, Ennui, Separation Distress (if int./ext. relations bad)

Pleasurable Stimulation (if int./ext. relations good) vs. Disgust, Dissatisfaction (if int./ext. relations bad)

The table indicates that when internal conditions are facilitative of meeting extropy needs, the resulting motivational tendencies are seeking, curiosity, or feelings of inspiration. When internal conditions are thwarting toward extropy needs, the result is lethargy. When external conditions are facilitative with respect to extropy, the results are bonding, play, lust, nurturance, etc. And so forth for the other needs, conditions, and motivational tendencies. Of course, we must always remember that internal conditions always already have been affected by earlier external ones, and avoid the temptation toward an oversimplified internal/external antithesis. From a phenomenological point of view, anyone can introspect what it is like to be on a continuum (1) somewhere between inspiration/curiosity and lethargy; (2) between playfulness and restlessness, and/or between some form of empathy and loneliness; (3) between courage/confidence on the one hand, and anxiety on the other; (4) between a feeling of safety on the one



 Curious Emotions

hand, and anger or fear on the other; (5) between wellness/satiation on the one hand and sickness/consummatory dissatisfaction on the other; and (6) between peripheral pleasure and peripheral pain. Feelings like sickness, wellness, peripheral pleasure and peripheral pain often are not thought of as emotions, and sometimes debates rage as to whether such qualia should be classified as emotions (for example, Aydede 2001; Ben Ze’ev 2002; Goldstein 2002). Some people prefer to separate all the felt senses directly associated with homeostatic needs (for instance, hunger, pain, physical pleasure) from the truly “emotional” qualia. For similar reasons, when Watt (for example, 2000) summarizes Panksepp’s emotional brain systems, he lists only six of the brain systems – omitting the seventh system that was included in my listing earlier, the consummatory pleasure system. But technically, if these valenced qualities can be monitored at a conscious level, and have self-organizational aims, they too are aspects of “emotional” systems, strictly speaking, and even those who do not define them as emotions do treat them at least as important motivations; in fact, pleasure and pain are the paradigmatic motivators in most psychological research involving reinforcement. The main reason for discounting pleasure and pain as emotions is that their qualia are often highly localizable, are usually peripheral, and often seem to lack intentional objects external to the perceiving subject. Even when they are not peripheral, they still may not be counted as “emotions” because they are such obviously “physical” feelings, lacking in intentional reference beyond the body itself, whereas emotional qualia at least seem to their subjects to be about external events. So while peripheral pains are too localizably received from specific areas to count as emotions, general feelings of illness are lacking in external intentional referents. Generalized “emotional” pain, however, would seem to present the same problems. And we have argued that the intentionality of emotions is usually not immediately obvious, or to be equated with a simple perceived object or event concurrent with the emotion and which happens to trigger the emotion. The trigger can be quite distinguishable from what the emotion is “about.” What really identifies emotions is not that they have immediately obvious objects, but that they have aims in relation to environmental affordances, many of which are not immediately obvious. Perhaps the best argument against regarding pain and pleasure as emotions is that they seem completely afferent. But in fact, the afferent delivery of pain signals through nociceptors is only the best understood aspect of pain. As Damasio (1999), Newton (2000) and Watt (2000) suggest, the less understood aspects involve how the brain evaluates the signals in terms of their good or bad valences. Here again, if there were no more than afferent nervous activity

Chapter 4. Homeostasis, extropy, and boundary needs

going on, we would neither be conscious of pain nor use it as a guide to action, as I have argued throughout. This point will be taken up in more detail in Chapter 5. In any event, if someone prefers to eliminate such qualia as peripheral pain and pleasure, general wellness, and illness from inclusion under the term “emotion,” I have no quarrel with such exclusions. There are different systems of definitions, and the important point for our purposes here is to describe and understand the dynamics of emotional and motivational phenomena. The mechanisms of all the qualia listed in the table are quite similar, and obviously have motivational significance. For any of these qualia to be felt requires contextualization in a self-organizational system, as I have already stressed. And, like other emotions, feelings of sickness, wellness, pain, lethargy, etc., help the organism to monitor how well it is doing in relation to its ultimate aim, which is to maintain the system’s ongoing pattern across replacements of microcomponents and other adjustments to its environment. All these processes are oriented toward organismic action. When we classify the most universal emotions of complex animals in terms of an explicitly self-organizational perspective, it is the aims and activities of the organism, rather than any presumed constant association with particular environmental objects, that delineates each kind of emotional system. The hardwiring of endogenous emotional systems is not linked to specific objects, except insofar as specific objects are especially prone to trigger organismic action affordances in terms of the aims already underway because of the organism’s own patterns and structures of action. Fear, for example, may be triggered by a variety of objects, and there is no need to learn to be afraid of one object because it has been associated through classical conditioning with some other object that already elicited fear, due to association with a still earlier fear-evoking object, tracing all the way back to original unconditioned fearful objects such as heights or snakes – or to objects that once caused an unconditioned response to sheer physical pain. Any kind of object that suddenly gets in the way of the organism’s ongoing projected aims by constricting or threatening to constrict the organism’s boundaries can trigger either anger or fear, depending on what kind of action the organism initially imagines taking to remove the threat or remove itself from the threat. From this perspective, even fear and anger are expressions of the selforganizing activity of an agent-directed organism, rather than passive, mechanical reactions caused by a stimulus. The organism must be already in a motivated and self-directed movement of activity in order for this freedom of movement then to be thwarted or constricted. Novelty is therefore both good



 Curious Emotions

and bad for the organism. “Bad” novelty is, initially, any novelty that interferes with the ongoing self-organizing activity – i.e., that does not present good affordances for it. Then, very quickly, the novel stimulus either becomes confining, in such a way that it affords “pushing against” or breaking out of the confinement (hence anger and frustration); or it remains “uncanny,” just as if it were still completely novel and unexpected, as if we were seeing it afresh, and we remain immobilized (hence fear, terror, or anxiety). Phylogenetically, this tendency originally came about because, in a dangerous situation where any move could kill the organism, it is best for it not to move, until it figures out what is going on. Thus both fear and anger are preceded by a more fundamental “startle” response (Panksepp 2000). Some situations stick us in that “uncanny” mode, because we continue to be unable to determine what action to take. So then our own immobility is what feels confining. The confinement of our own fear may later lead to an anger which then breaks us out of the confinement and leads to energized action. Thus a major difference between fear and anger is that anger means feeling confined by something external and wanting to break free, whereas fear is feeling confined by our own immobility in contending with a threatening novelty. The difference is in whether the confinement is self-initiated or object-initiated, although both may be responses to a similar object. In general, bad novelty feels confining, in one way or another. Whether in fear or in anger, confinement constricts a self-organizing freedom of movement, which is a precondition for either response. It is important not to oversimplify here. We all too often assume that those emotions that are most universally experienced are the “basic” ones, and that all the others are “derivative” from them. But this is not a logical inference. The properties shared by all members of a certain class are not the ones from which we can derive other properties those members may possess. All baseballs have stitches and are spherical, but it does not follow that all other properties of baseballs are derivative from these. There are many and various different emotions that feel confining when our freedom to realize complex action potentials are not afforded by the environment, and these include grief and separation distress. In these cases, the experienced lack (Lacan’s “missing object”) constricts us. For example, many people in the grips of grief feel almost as if they cannot breathe. Thus it is not counterintuitive that Panksepp classifies separation distress as a separate “panic” system in the brain. When an important human relationship facilitates a subject’s being in a more interesting, higher-energy basin of attraction, the withdrawal of that love object also

Chapter 4. Homeostasis, extropy, and boundary needs

means constriction of the organism’s normal set of self-organizing activities, especially at the level of brain functioning. One of the effects of love is that it enables the subject to value being per se more intensely. One of the reasons why empathy with a conscious being whose structure is interesting is a positively valued experience is that it inspires the subject with a feeling that being a conscious being per se is a fun and valuable thing to do. Infants whose exploratory drives have not yet developed are utterly dependent on empathy with a caretaker for this kind of value experience, and therefore if deprived of empathic interaction can die of marasmus even though their consummatory needs are met (Spitz & Wolf 1946). Loss of a love object withdraws this “being-valuing” action affordance; the subject can no longer as intensely value its own activity, and seeks a less active, lower-energy basin of attraction, lapsing into lethargy and depression. In the case of older children and adults who have already developed other extropic dimensions such as exploration and play, separation distress will be experienced as a negative affective situation, and the desire to live, still pronounced in other subsystems, feels constricted by this very lethargy, so that those subsystems force the experience of restless anxiety and a feeling of painful tension and confinement (for further discussion of this point, see Ellis 1996b: 81–95). Even though bad novelty can trigger fear, anger, or separation distress, it is equally true that good novelty can feel liberating – it presents affordances consistent with the body’s ongoing desire to keep flowing in natural patterns, but to increase the energy level while doing so. Thus anger and fear, just as much as joy or play, intend the organism’s self-initiated activity toward maintaining a self-organizing pattern, and are not just causal reactions to external stimuli. This is an important point for integrating physiology with existential emotional issues, because it allows anger and fear to be conceived of in terms of the organism’s already ongoing activity, rather than as causal reactions to an external stimulus. In this way, we can make phenomenological sense of what it is that the organism wants, as correlated with its basic selforganizational structure, rather than simply regarding anger and fear as either arbitrary causal reactions, or as tendencies derived through learning from consummatory drives. The intentional aim of every emotion is to continue the pattern of self-directed activity whose purpose is to maintain the organism’s overall active pattern of being.



 Curious Emotions

. The importance of extropy needs in higher mammals The higher we go on the evolutionary scale (and also the more consciousness there is), the more the organism wants to deviate far from equilibrium and find more extropic basins of attraction that can also facilitate homeostasis. Each organism has many basins it can settle into, and it was advantageous for a hunting/gathering creature to develop a tendency to prefer the higher energy (seeking and playful) ones. This is the genetic hand we are dealt with regard to physiological emotional tendencies. We have a natural tendency to prefer extropy, but that tendency often can conflict with the preference for homeostasis, and the basic physiological emotional tendencies are complex ways of addressing these internally-conflicting concerns. Of course, an inherent desire to explore, a sense of general curiosity, and a need for intellectual stimulation, even when such activities do not lead to consummatory reward, are going to be selectively favored for survival in a hunting and gathering gene pool. An important existential parameter addressed by this concept is also the need not just to achieve whatever it is we value, but also to affirm the value of what we value – to feel that it is valuable. If we did not feel inspired to exercise our capacity for the realization of our activity at more extropic basins of attraction, we would lapse into lower energy ones. Thus the feeling of inspiration – the Aristotelian valuing of our own activity for its own sake, for the sheer interest of it – correlates with the release of DA and NE, which are neurotransmitters that facilitate efferent action commands and make us feel more conscious, more “alive” and, quite literally, more awake. Hence the power of caffeine addiction, which directly stimulates the release of NE, and indirectly of DA. In order to exist at our most extropic basins of attraction, we must feel that doing so is worthwhile, i.e., we must feel “inspired.” Lists of “basic emotions” that do not include the need for something like a feeling of inspiration are therefore too impoverished to be applicable to the motivational structure of highly complex self-organizing animals. Even a dog or cat can feel inspired or, on the contrary, generally lethargic and uninterested in its environment. In an extreme case, we have already discussed the fact that marasmus infants can die even though all their consummatory needs are met (Spitz & Wolf 1946). There are entire categories of human activity whose purpose is not to attain that which we value, but to intensify our sense that whatever it is we value is valuable, through a feeling of inspiration. Religion, love, and the arts are examples of such “value-affirming” as opposed to “value-achieving” activities. Love, for example – the valuing of another being as an intrinsic rather than

Chapter 4. Homeostasis, extropy, and boundary needs

merely instrumental value – is an existential need, called for by the interactive structure of our being, as well as a basic physiological tendency of the brain. Love in its various forms presents another being who is novel as opposed to boring (in terms of moment to moment lack of predictability compared with one’s own predictability) and liberating as opposed to confining, yet not threatening the way some novel stimuli are. We vicariously experience the other as someone whose being-structure would be interesting to enact, and we try to do so as nearly as possible through empathy. Correlatively, in love we experience the other as extremely valuable, so that an existential sense that the overall struggle is worthwhile can be served. The behaviors of love are often meant to enhance the experience of this inspiration rather than only to achieve the outcomes posited as valuable by the value feelings that are intensified in the process (for a more elaborate discussion of this point, see Ellis 1996b).

. Existential requirements for an adequate dynamical theory of emotion We have seen that there are three aim-oriented issues involved in being a complex self-organizing being, all stemming from the fact that our basic tendency, in order to be what we are, is to act out our own unique pattern of being in ways that maintain both extropy and homeostasis: 1. We need extropy. This means that we need to have novelty, because novelty affords extropic activity. It also implies that we need to find ways to further concretely embody our conscious states so as to amplify them, which entails symbolization. I.e., we feel conscious states more intensely when we express them concretely. For example, uttering eulogies intensifies grief, caressing a loved one or singing love songs intensifies the feeling of love, and clenching the fists intensifies anger. Similarly, we grasp our general felt sense of a situation more vividly when we try to put it into words. This need for symbolization in turn leads to speech, art forms, and relations to others, so that we can have media through which to symbolize and thus amplify the activities of our being (see Ellis 1986, 1996b, 2000d). 2. We need to protect the boundaries of our being, since to be is to be within certain boundaries. This means that we become startled and thus freeze at the unfamiliar or uncanny. We then need to further process the unfamiliar so that it ceases to be uncanny, which entails imagery of action affordances in relation to the threatening object, and in many instances symbolization



 Curious Emotions

activities which can further refine the threatening situation’s meaning. This often may include expressing our symbolizations to others, in the interest of further clarification and amplification of our consciousness of what is going on; and of course we finally need to either get away from or destroy the uncanny thing, which fundamentally threatens our freedom of action, and thus requires a self-energizing to break out of the confinement either of our own immobilizing fear, or of the dominance of the predator (as in anger). 3. We need to reduce our chemical imbalances, so as to achieve homeostasis. This implies that we need predictability/familiarity in the environment, and not just novelty/adventure. This need in turn entails a need for further clarification of what it is that we need and want, which again leads to imagery/symbolization, and to cooperative arrangements with others to meet the needs, since cooperation is one of the main ways mammals are adapted to meet them. This process then often leads still further to feelings of being motivated to form social bonds. Correlatively with (1), we need to experience the project of achieving all these goals as one that has value per se, rather than as meaningless and pointless. I.e., we need to feel that being what we are (or being anything) is worthwhile or has value – that the entire project of being is worth the effort. This implies the need for an object that affords play (extropic homeostasis) – and the only completely adequate object for this purpose is another conscious being. We then need imagistic symbolization to amplify our experience of the value of the playaffording object, i.e., the other person (which can mean expressing our feeling of the value to others, or just to ourselves); and we need to act in relation to the valued object, in such a way as to amplify our experience of its value – which leads to friendship, respect, loyalty, and many other social emotions. In this way, the entire realm of object relations, in the sense used by object relations theorists, can be grounded not only in existential phenomenology, but also in physiology and a basic understanding of complex dynamical systems. Phenomenological experiencing of emotion can be correlated, given a detailed enough account, with what complex self-organizing (homeostasiswithin-extropy-craving) beings need, in chemical terms. The need for extropy correlates with what Panksepp (1998) calls the brain’s seeking system (curiosity, boredom, vigilance, enthusiasm for activity), which also interrelates with the play system, and also the wakefulness and arousal system. These tendencies cause the total self-organizing system to try to use norepinephrine, dopamine and other such chemicals to amplify the complexity and activity level of the sys-

Chapter 4. Homeostasis, extropy, and boundary needs 

tem. Hence the need for social relations, because other conscious beings afford novelty and unpredictability within safety. The need for homeostasis correlates with the consummatory pleasure– pain system, which when combined with the other needs leads to social integration of pleasure–pain issues through the brain’s play, lust and nurturance systems (and in less direct ways to other social emotions). Political, economic and professional activities are complex attempts to harmonize all these disparate motives. Thwarting of the freedom of movement in the resulting activities may again result in frustration of the more basic aim to engage in patterns of activity consistent with the organism’s whole self-organizational structure, which can trigger very basic feelings such as fear or anger, but now in relation to complex social conditions such as political or economic oppression. The need for secure boundaries and freedom of activity within those boundaries correlates with the startle response to “bad” novelty, which leads to enaction of either the fear or anger systems as ways to self-energize and reassert freedom of action (unconfined by the bad novel stimulus, whose effect is to constrict and prevent free activity). Given the complex physical, economic and professional relations just mentioned, the anger and fear systems can be triggered by events that are much simpler than their true intentional referents, which must be contextualized in the realm of these more complex political and economic arrangements. Events such as motorists cutting in front of us are routine examples that occur on an everyday basis. The need to experience the project of being as worth the effort (the need for inspiration) is a further elaboration of the arousal and seeking systems, and correlates with Panksepp’s separation distress (or what Panksepp also calls the “panic”) system, which is the negative manifestation of a love system, also including lust and nurturance (utilizing release of oxytocin in the brain for females, and oxytocin, vasopressin, and prolactin for males), as well as friendship and play (combining oxytocin and prolactin with energizing DA, NE, and other neurotransmitters). Ultimately, we need others in relation to whom we can fully express what we are – which is just as fundamental and existential a need as eating or protecting one’s boundaries. The failure to achieve this aim can result in depression, which correlates with chemical imbalances of the total system, but ultimately stems from a failure to energize the system at a high enough level of extropy, since the value of doing so needs to be felt in order to motivate it. Thus dogs act listless and disoriented when they fail to get the kind of play, social interaction, and even obedience training that they want (being such an extremely cooperative

 Curious Emotions

species). Death through marasmus in human infants is the clearest instance of total failure of this system. Although we speak of self-organizing systems as having a “tendency” to appropriate whatever physical elements are needed to keep the system going in its preferred patterns, and readjusting to a different basin of attraction to compensate for changes in the environment, we should of course remember that maintaining a complex self-organizing system is not at all easy, and requires the cooperation of several different extraneous mechanisms, such as the gifts we inherit through long and hard natural selection, as well as conscious planning and good luck in terms of environmental conditions – and even then, of course, the project is ultimately doomed to eventual failure. Even a tiny disruption of the evolutionary legacy, such as a stroke, can make the project impossible – and probably in almost all instances the best we can do is to come close to achieving all the elements for the total extropic-homeostatic balance that is desired. The complexities resulting from anticipation of this ultimate failure in death affects the most intelligent organisms in such an important way that the need to appreciate the value of being per se, as instantiated in beloved others and experienced through positive relations with others, becomes still further pronounced than it would be in an animal unintelligent enough simply to live in the present moment without concern about death and other conditions of finitude. This leads to a further intensification of many kinds of love experiences in humans. Not only is love not merely derivative from more elementary needs such as sexual pleasure, but in fact sexual lust can be a relatively mild impulse compared with the impassioned turbulence of love when intensified by the larger existential need to affirm the value of being through a sense of inspiration, which is further exaggerated by existential understanding of the limits of human life in all dimensions of finitude, including the finitude of the human life span (Ellis 1996b).

. Toward an integrated physiological and phenomenological account The tendency of the most complex self-organizing systems toward extropy is a good candidate for explaining the “seeking” and “play” systems – so important in the conflict between consummatory drive reductionism and the “higher” or “existential” emotions – from a self-organizing perspective. We have seen that novelty is a preperceptual category that is always recognized by the emotional limbic system prior to perceptual processing (LeDoux 1996; Damasio 1999). The first instantaneous response to real novelty – novelty of

Chapter 4. Homeostasis, extropy, and boundary needs 

a kind for which we are completely unprepared by the immediately preceding context – is a startle reflex, which has basic emotional-brain ramifications. Then, as the novel stimulus has time to be processed further, if it continues to be novel in a bad way (in a way that does not present self-actualizing affordances), the startle response quickly evolves into a mild fear response, i.e., freezing. Natural selection serves well enough as an explanation for this form of organization: We are better off doing nothing when any wrong move could mean death. Subsequently, the freezing becomes a confinement, and confinement then provokes either flight, or a kind of energized rage response, either of which breaks us out of the frozen startle response and equips us for action. On the other hand, if the novel stimulus presents good novelty (self-actualizing action affordances), then the seeking and play systems go into action. We can thus correlate existential issues with Panksepp’s seeking, play, rage, and fear systems. With regard to his lust, nurturance and separation distress systems, all three of these brain systems activate social emotions involving empathy (and thus utilization of oxytocin and prolactin circuits in the brain) although separation distress does so in a negative way. These systems too can be given existential accounts. In order to actualize ourselves, we need to be in interaction with others – this is the nature of mammals (and of intensely conscious beings as conceived of in existential terms). When not in interaction with others, either directly or indirectly, it may become impossible for us to be what we are; our self-organizing systems cannot maintain themselves, and we feel the resulting failure of both homeostasis and extropy. “Lust” and “nurturance” are similar to each other both physiologically (Panksepp 1998) and phenomenologically (Ellis 1996b); they involve actualizing one’s being through empathy with another, and experiencing the value of being as instantiated in the other through empathy. Sex too (which is not really necessary for reproduction, in lower species) evolved hand in hand with these tendencies toward empathy. This does not explain, of course, why sex gives pleasure – that explanation would be more elementary, having to do with achieving chemical homeostasis. What the need for higher value experience explains is why we are the kind of beings that associate sexual reproduction with empathy in the first place – why we connect the sexual pleasure with the perception of other beings of the opposite sex (in a way that empathizes with them) rather than, say, thinking about algebraic equations or of nothing at all while ejaculating. (The evolutionary reason, of course, is that beings that thought about algebraic equations while ejaculating didn’t reproduce themselves as well, and thus didn’t compete well with those who thought empathically about other beings while doing it). With-

 Curious Emotions

out this empathic dimension, we have acute distress, because we either cannot be what we are, or we cannot see the full value of being what we are (or of being anything at all) – which is a frequent precursor to depression (Panksepp 1998: 274). Even the courtship behavior of sub-mammalian species, such as birds, often exhibits elaborate imitative routines very suggestive of empathy. Understanding why we need others to actualize our being requires understanding the structure of the self or personality, which is a higher-order pattern of progressing from one conscious state to another: A personality can be identified by the pattern in the directions of those changes, which is a style of being, rather than by a mere listing of which conscious states are experienced; we might all experience them all, at various times (Ellis 1986). So to be what we are, we have to relate to others who will allow the right kind of pattern of progression from one state to the next, in addition to allowing progressions that generally afford extropy within homeostasis. For each individual, different patterns will be required to achieve these ends, since the ends themselves have to do with complex structures of dynamic change. In sum, we have now seen considerable evidence that contemporary neurophysiology of the emotions does not require a simplistic egoistic-hedonist and consummatory-drive-reductivist view of the emotional life, but is generally consistent with the demands of existentialists and self-actualization psychologists. The key to reconciling physiology with these higher-order ways of seeing the meaning of emotions is to approach the physiology underlying consciousness itself in terms of complex self-organizing systems that seek extropic basins of attraction as well as homeostasis.

Chapter 5

Varieties of extended self and personality

As I type at the computer, I am aware that the word “I” in this self-reference can mean a variety of different things. It can mean the conscious awareness or “subjectivity” in which the texture of the computer keys is felt and the meanings of ideas are entertained. As Kant pointed out, the existence of this subjective “I” does not require that I continually reflect on myself as the subjective dimension of every objective content (every object of perception, thought, etc.). It is enough that the subjective dimension is always potentially available to reflection. But even this one sense of “I” has many variations, depending on which philosophers one reads. Sartre (1937) famously disagreed with Kant and Husserl as to whether there is such an entity as “I,” other than just another constructed object within the field of other objects of consciousness. In a different sense, “I” can also mean the primitively “agentic” ability to move my own fingers in a purposeful way, for example as I type these words on the keyboard (Gallagher & Marcel 1999; Marcel 2003). One advantage of this sense of “I” is that it can refer to my self while asleep, since even then the organism continues to engage in self-organizational activity toward the holistic purpose of balance and integration within the complex pattern of activity that is being maintained across replacements and readjustments of its substrata. But to say that this sense of “I” is a meaningful one that really refers to something is not necessarily to deny the existence of the self in the other sense – the “subject of experience” sense that Kant and Husserl talk about. Instead, it raises the interesting question as to whether both senses of “I” refer to different aspects or capacities of the same single entity or not. I shall return to this question later in this chapter. Besides these two senses, there is also the “I” in the sense of my body as an object – for example, the fingers that I can see as they type and the face I can see in the mirror. This concretely instantiated object, my body, which can be studied in various ways by scientists, is a normal referent of the term “I” that is not rendered nonsensical by the fact that, in the first two senses mentioned above, I could still be “myself ” even if my hands were amputated or if my face were reconstructed so that I could serve as a Sadaam Hussein body-double.

 Curious Emotions

It is an open question whether, as Bergson affirmed and Aristotle denied, I would still be the same “I,” in some sense, even if all the micro-components of my system were replaced. In fact, we must question to what extent I remain the same “I” from one moment to the next, as not only the contents of my thoughts and the molecules of my body are replaced, but also the patterns of activity in which they engage. This question leads to still another commonly used sense of “I” – the concept of that which provides the continuity from one moment to another. Against the temptation to define a person simply as the Humean “bundle” of thoughts and sensations that occur in one body, there is the problem that, if everyone could entertain pretty much the same thoughts and feelings at various times, then they would all have the same personality, or would be duplicates of the same person. But this seems counterintuitive. Instead, there seems to be a personality that endures across the changes of momentary states, and does not even depend on whether or not I entertain one or another particular momentary state. In some sense, I would still be the same person even if I had not experienced the pain of hitting my foot on a chair this morning. This “person” or “personality” sense of “self ” seems to be predicated more on the pattern that influences the direction of change from one state to another than it does on what the particular states are. I shall call this dimension the “person,” to distinguish it from the above senses. But one of the challenges facing this notion of “person” is whether in fact there is any such thing, and I believe the self-organizational approach to emotion is in a good position to elaborate how such a notion can be coherent. Another important distinction whose neglect could lead to a great deal of confusion is the difference in meaning between “self ” and “sense of self.” That is, having a self may be different from sensing this self. For example, some phylogenetically low-ranking animals may have selves, at least in some of the above senses, but not sense that they have them. By distinguishing the meanings of these terms, I do not mean to foreclose the question as to whether the two terms may or may not always describe the same entity, or be subserved by many of the same brain processes. The terms Clark Kent and Superman have clearly different meanings, but it happens that they describe different aspects of the same entity. Anytime we sense something, it is logically possible that the object of the sensing could have existed without being sensed. The sensing that I have a left arm is not the same thing as having a left arm. I could have a left arm without sensing that I have one, or in rare instances even sense that I have one without actually having one. The same goes for the self; a “sense of self ” cannot be auto-

Chapter 5. Varieties of extended self and personality

matically equated with the self that is sensed in this sensing. As we shall see, this point holds true even for the most sophisticated and complex concepts of an “extended self.” In each case, it is simply a logical point that sensing something is not the same thing as having it or being it. Some people might want to collapse this distinction a priori on the grounds that the meaning of “self ” is inherently subjective. To be a self is to be a sensing being, and to sense anything is to be a self – at least in one sense of the word “self.” But even in this particular sense, it does not follow that to be a self is the same thing as to have a sense of myself. It is logically possible that I could be a sensing being, yet be engaged in sensing things other than myself. In that case, I might have a self, but not have a sense of self. Conversely, I might sense myself as being something other than what I actually am. Also, the meaning of “sense of self ” will vary, depending on which meaning of the term “self ” is in play. Among the various meanings of the term, at least several are probably applicable to something that is real, as opposed to socially or personally constructed, although they may describe different features that in most normal instances will all apply to the same organism. We can distinguish different aspects of the “self ” in the sense of primitive agency, discussed earlier – the simple ability to move rather than just be moved. A slightly more complex variation on this concept of capacity for self-movement is the idea that my will is the source of the movement. While asleep, for example, my body performs many minimally agentic actions in the sense that they are dynamical and robust rearrangements of many micro-constituents geared toward the organism’s balancing of various self-organizational purposes; yet I do not will my body to do any of those primitively agentic things. So what one might call the “ownership” of the primitive actions that one deliberately “wills” (Marcel 2003) may describe a “self ” that is very primitive, but not as primitive as the mere ability to move per se. I shall refer to this meaning of self as the “ownership” of actions. One of the problems it raises is the problem of mental causation. Because of this problem, many have decided (falsely, in my view) that the power of a willed choice to cause motion in one’s own body is entirely illusory (Wegner 2002). I shall return to this problem. I believe the “action ownership” aspect of the self is closely related to Panksepp’s idea of a “core self.” Panksepp makes essentially the argument that the core self means agency, in both the most primitive sense and in the sense of “owning” the willed actions. In his view, this very primitive agentic “core self ” is rooted in the PAG and upper brainstem, based on observations suggesting that this brain area initiates action in the interest of the organism’s motivations. This includes the notion of primitive agency, but adds a more subjective



 Curious Emotions

component with the feeling that a unified subjective entity, my core self, initiates the actions in question. Panksepp also links this point with his emphasis on the idea that the emotional systems are necessary for all forms of consciousness. He suggests in several places (1998, 2000, 2002) that the “core self ” may be necessary for a self in all other senses – including any capacity for subjective reflection or even subjectivity per se (these two phenomena are distinct, by the way, because subjectivity per se constitutes a “self ” in one sense even if not reflected upon). According to Panksepp, this would entail that any subjective consciousness, including the capacity to reflect on it, would be inseparably linked with a more basic capacity for agency. Therefore, it would necessarily have to be true that other forms of subjective consciousness are derivative from the core self ’s motivation to act and to generate sensorimotor action imagery. For example, if, as I have argued, there can be no perceptual consciousness as a function of the visual cortex without this consciousness somehow being necessarily a part of an action routine (or of a motivation to act or action imagery of some sort), then Panksepp’s claim that very primitive PAG-brainstem selfregulation is the root of the core self in the sense of owned primitive agency would be a very likely hypothesis. The self in some of the senses we have just distinguished seems to be more basic or primitive than the others, so the question arises as to whether they are dependent on the “core self ” as defined by Panksepp (similar to and certainly including “primitive agency”), and whether some result from overlays of planning, cognition, self-consciousness, or other mental activities added to primitive agency. If it is true that there can be no consciousness without at least unconscious action imagery, then even the subjective self – the “subject of experiences” – would as a consequence be dependent on the core self or the primitive agent. I have already argued that forms of subjective consciousness that are not commonly viewed as emotion – for example, perceptual consciousness – are necessarily dependent on emotion, motivation, and action-initiation (which may then be inhibited) in order to be experienced subjectively. If this were not the case, then we could have the ability to experience subjective consciousness in some forms (for example, subjective consciousness of perceptual objects) without a major contribution from the emotional brain areas. But evidence is mounting that consciousness is dependent on the core self ’s motivated actioninitiating capacity, and it seems that the emotional brain areas do play a major role in all forms of consciousness. One of the most coherent cases for the enactive view of this dependence of the subjective self on agency is made by Natika Newton in Foundations of

Chapter 5. Varieties of extended self and personality

Understanding (1996). The central purpose of her book is to show that the essential difference between the “knowledge” of computers (for example, Searle’s Chinese room) and conscious humans (for example, a person speaking Chinese) is that humans can act and plan actions, whereas computers only react. She then tries to show that all forms of consciousness, from perception to thought to our understanding of other people, are rooted in our ability to perform basic actions, and to grasp the action affordances of environmental situations and objects; our abstract thoughts are ways of incorporating environmental affordances into possible action schemas in more and more complicated ways. I also argued for Newton’s enactive thesis in Questioning Consciousness (1995) and earlier in this book on the grounds that it fits the ERP data we have discussed here. The subject’s consciousness of a suddenly presented novel stimulus is not merely a result of sensory processing, but also presupposes complicated subcortical emotional activation that is not dependent on the perceptual processing that goes on simultaneously. Thus, without the emotional activations, there would be no perceptual consciousness, although there might be blindsight or implicit knowledge. And of course the notion that the core self is rooted in primitive agency is explicitly built into the groundbreaking work by Varela et al. (1991), The Embodied Mind, which systematically advanced the “enactive” theory of consciousness. What, then, must be added to the core self or primitive agent in order to constitute the extended self that we think of ourselves as being in daily life? That question depends partly on what aspects of self-functioning are meant to be included in the extended self. As already noted, whichever meaning is at issue should not run together “self ” with “self-representation” as if they were the same thing; if the distinction is made between a self and a representation of the self, then an additional problem is highlighted: we can have subjective perceptual consciousness, and in a sense the subject of this consciousness could be a “self,” whether it then reflects on itself or not. So here again, if subjectivity is a “self,” then this subjectivity is necessarily dependent on emotion and agency. But if “self ” is meant to refer to a being that is capable of self-reflection, not all subjects will be selves in this latter sense. For example, given the anatomical and behavioral similarities of cats’ and dogs’ brains to our own, it is very likely that these animals have consciousness. But they cannot interpret the image of themselves in the mirror as themselves, and it is difficult to think of any behavioral indication that these animals can reflect on themselves. Certainly, it is entirely logically possible for something to be a subject of experience without being able to reflect on itself. This would simply mean that it can be conscious of the



 Curious Emotions

outer world and some aspects of its own body and interoceptive sensations, but not be able to reflect on itself as a whole. Still another “self ” that many current authors discuss is the sense of a “higher-order” or “social” self, which many equate with a “constructed” or “narrative” self. This notion of a narrative or constructed self is only a very recent one, and one that may already be going out of fashion among the postmodernist philosophers and literary critics who advanced it. But aside from this constructivist conception, there is also a traditional notion of self in the sense of “person” or “personality,”discussed by personalist philosophers such as Bergson, as well as Husserl and Aristotle, and of course by many psychologists under the rubric of “personality theory.” Husserl, for example, thinks of personality (the “psychophysical subject” discussed in his Lectures on Phenomenological Psychology) as a stable way of organizing subjective experiences. I.e., anyone could experience any given state of consciousness, so what makes one person a different “personality” from another is that he organizes the flow of states into a pattern that is unique to him. Sartre, as noted above, argued against such a possibility, but part of what Sartre ignores is that this difference in ways of directing the stream of experiences could be attributable (in agreement with Panksepp’s PAG/core self thesis) to the way emotions and motivations dictate the direction of attention at each subsequent moment in the stream. The higher-order pattern formed in the overall motivationally directed stream of consciousness could then be different for each different personality. So another meaning of “self ” is the idea of the unified “person,” which provides a continuity and unique character to the pattern in which different items in the stream of consciousness unfold. Many people could have pretty much the same experiences, yet they still would not have the same personality. If common sense inclines us to think of ourselves as equivalent with this “person” aspect of the self, we are also naturally prone to equate ourselves with what appears at first to be a completely different kind of self – the self in the sense of the “subject of experience,” mentioned above. It is quite possible, as various philosophers throughout history have thought (e.g., Aristotle, Descartes, Kant, and the phenomenologists), that every state of consciousness has a subjective dimension and an objective referent, and that the possibility of subjectivity per se is the self. I.e., wherever there is consciousness, there is a “self ” that has or executes this consciousness. So subjectivity itself would constitute a “self,” at least in one of its most important functions, regardless of what kind of narratives or constructions are lain over that “self.” Here again, to be a self and to have reflective knowledge of the self that one is are two different things that should be distinguished. So there are several notions of “self ”

Chapter 5. Varieties of extended self and personality 

that are not equivalent with anything that is constructed from socially learned categories, narratives, etc. If we sense ourselves both as subjects and as persons, then at least our common sense experience would seem to suggest that there should be a way to unify these two concepts: on the one hand, the subjective component of every experience that always stands ready to be reflected upon (this would be the standard Kantian definition of the “ego”), and on the other hand the “person” or “personality” in the sense of organizational unity across time. The latter would be a higher-order and partly social self, but not merely a constructed or narrative one, because the subject need not construct any narrative in order simply to direct attention in certain unique patterns that ultimately will form a unique structuring of the stream of consciousness. This unique unfolding pattern can pick out that individual “self ” from all the other possible ones, whether or not this unique self is able to “know itself ” accurately. To complicate matters further, another notion of extended self that goes beyond the “core self,” is the sense of planned actions emphasized by Newton (1996). “Self ” in this sense would require more than the primitive agency discussed in the previous chapter. It would require more sophisticated functions than the “core self,” and would manifest itself in social as well as physical relations, but still would not be a narrative or constructed self. If we cannot imagine ourselves as existing in the absence of planned action in an everyday sense – being able to plan a few moments in advance to reach out and grasp a coffee cup, for example – then one might suspect that this function too would find a place in a coherent notion of self, inclusive also of subjectivity and personality. And finally, we must not forget that all these senses of self differ from the idea of a “reflected” self, i.e., the self as it appears to us in reflection, which may or may not reveal the “true self ” that is doing the reflecting, as object relations theorists and self psychologists such as Kohut often stress. This reflected self might be a construction or narrative, but the “true self ” that is the subject doing the reflecting (and which might not be revealed in the results of the reflection) would not be a construction or narrative. This is the object relations theorists’ distinction between the “true self ” and the “narcissistic self.” Part of the problem, then, is that there are many senses of “self,” of which these last two we have just mentioned (“subjectivity” and “reflexivity”) are only two of the most prominent. Like Panksepp, I would like to argue that somehow the “core self ” that is rooted in the PAG-brainstem emotional systems and grounds the possibility of agency is more basic than these others. In order to establish this, and to clarify the ways in which the more extended “selves” are

 Curious Emotions

built up from this concept, we need to show how the other functions that are often labeled as “selves” in everyday usage and in the usage of various thinkers are dependent or somehow necessarily tied in with the “agentic” function of the core self. I want to argue in this chapter that it is possible to develop a synthesis of all these concepts: that all consciousness requires motivation and primitive agency in the sense of the “core self,” but also that there are other senses of the self that are not merely the narrative/constructed ones from which the object relations theorists distinguish the “true self ” – that the self in all these senses may be real and may in the final analysis be different aspects of more or less the same entity.

.

How emotion grounds the various senses of self

If the action-reaction distinction is formulated in terms of dynamical selforganization, and makes possible unified organismic action and correlative intentional experience, then it should not be a far-fetched hypothesis that emotionally motivated directionality in the stream of consciousness is what makes possible the organization of experience in the form of a self, person, or experiencing subject. I shall argue that several of the most important senses of “self,” “person,” and “subject” cohere if we understand them in terms of the way the organism’s emotional life unifies experience. The most primitive meaning of self is simply the notion of an organism that acts rather than merely passively reacting to stimuli. I have suggested that this primitive action initiator is related to Panksepp’s (1998) “core self ” – a primitive agency that may or may not be supplemented with a capacity for action planning, conscious subjectivity, or reflexivity. This basic starting point, that selves can act, serves as a necessary foundation for two other concepts of self, both of which are important for their own philosophical and psychological purposes – the “embodied self ”, which grounds specific subjective states in specific embodied actions and the “personal self ”, which weaves the various subjective states into a personality structure. The embodied self is an agent attuned to environmental affordances that can create subjective awareness; if this intentionality is reflected upon, it reveals a subjectivity in the form of a felt sense that is always available for reflection. The personal self – or personality – seems to be a higher-order phenomenon than any particular subjective state, and is constituted by the patterning of the various states in the stream of consciousness; it is what gives each individual a personality over and above the particular experiences, none of which are necessary for the unique identity of

Chapter 5. Varieties of extended self and personality 

the person. This pattern allows identification of an overall personality as well as a unity across the various states in the stream. On my argument, the functions of both these selves are united in the workings of the emotional-motivational self. At the end of this chapter, we shall explore the ways we can or cannot know about these selves – e.g., whether this attempted knowing leads to mere constructions, or whether it can found actual knowledge of what is real about ourselves. The first requirement of a self is that it be able to act. One of the first problems in any concept of self is therefore to make possible an understanding of how such a self ’s choices and motivational intentions give rise to certain results, such as the willed movements of the body, rather than having merely illusory causal power. If I decide to raise my arm, there seems to be a sense in which “my” decision causes the hand to go up. When a cancerous tumor grows in my brain, we want to say that this sense is lacking. “I” do not decide that the tumor is to grow, even though it is my own self-organizing organism that, in the final analysis, makes it grow. Just as in the case of the tumor, we know that numerous micro-mechanisms are necessary and sufficient precursors of raising my arm, and in both cases, we saw earlier that the micro-mechanisms do not thereby exhaust all the causal power over whether the arm goes up. The system self-organizes its own microconstituents by controlling and readjusting the background conditions under which any given micro-sequence can or cannot lead to the holistic outcomes needed to main self-organization. Yet the reason for this causal robustness of the process with respect to its micro-components does not show that there is freedom to make mental choices that can really determine behavior. The point we made earlier about self-organizing processes having causal power with respect to their micro-components is just as true for the growth of the cancerous tumor in my brain as it is for the execution of the choice to raise my arm. The feeling that the choice to raise the arm is what causes it to go up could then be only an illusion. If a unitary act – the choice to raise the hand – has real causal power, then as Juarrero (1999), Newton (1996) and other action theorists have pointed out, this is at the heart of what it means to act rather than merely passively reacting. On the other hand, if all the causal power is at the “sub-personal” level, just as it is for the growth of the cancerous tumor, then the movement of the arm, in the final analysis, is really just as passive on “my” part as the growth of the tumor, and is not truly active in the sense needed to ground the feeling of freedom included in the sense of agency. And if this feeling of agency is a mere epiphenomenon, with no real causal power, then so are

 Curious Emotions

the concepts of self just mentioned, since they depend as a minimal condition on the self ’s being able to act. Traditionally, philosophers have formulated the problem of mental causation in this way: the “causal closure of the physical realm” – the notion that every physical event, if it has a cause, must have a physical cause (Kim 1992, 1993) – seems empirically as true for physical and chemical events in the brain as it is elsewhere. Many empirical brain studies support this conclusion (see Ellis & Newton 1998a). Yet, when I raise my hand, it seems obvious that the decision to raise the hand makes the hand go up. If there is mental causation, this would mean that, if I make a mental choice, the choice can have causal power over the resulting movements of my body. If the choice is merely an ephiphenomenon or causal side-effect of a series of physical events, over which “I” as a unified being have no control, then the feeling that the choice influences the resulting action would seem to be only an illusion. Both the feeling that the choice has causal power and the behavioral outcome then really would be caused by a sequence of physical events over which “I” have no control, just as in the case of the growth of the tumor, even if it is true that there are selforganizing processes in my body that do control them. This would be a trivial sense of choice as far as “I” am concerned, in the same way that it would be trivial to say that, when a cancerous tumor grows in my brain, it is “I” who decides that it is to grow. Thus, if the feeling that a mental decision or intention has causal power is not to be a mere illusion, it is traditionally assumed that some combination of events other than the mental decision cannot be completely sufficient to produce the outcome independently of the decision. If A (a physical mechanism explainable at the micro-level) causes both B (a mental intention) and C (a physical or mental outcome), then A is sufficient for C independently of B, so B is precluded from having any real causal power over C. B is simply an epiphenomenon of A. Because B always accompanies A, it feels as if B is causally necessary for C, but if the micro-level causal account is correct then this feeling must be only an illusion. The same problem will arise whether the purposeful intention of the organism is conscious or unconscious. A man courting a woman may not be aware of making more eye contact than he normally would in order to achieve an objective, but this behavior may often be purposefully intentional nonetheless. But if the behavior is completely explainable by sub-personal physical processes in the brain, without reference to a self-organizing power on the part of the mental choice itself, then the causal power of the man’s intentions would

Chapter 5. Varieties of extended self and personality

seem to be only illusory, unless we can somehow reconcile it with the causal closure that apparently exists at the sub-personal level. In neurophysiological terms, to say that the man’s “self ” is driven to behave by hormone and neurotransmitter activity such as testosterone or vasopressin would place the “self ” in the cortex, because it is the cortex that receives the driving input from the subcortex and cerebellum. This could not be correct, because we have seen repeatedly that action is already initiated at the subcortical level, with or without further cortical refinements. So the acting self must include the subcortex, and cannot be a separate entity that merely receives causal inputs from the subcortex. In concrete terms, a man on a beach surrounded by naked women will not necessarily be “driven” by neurotransmitters to lust after them if he is already geared up toward a unified action to which they are irrelevant, such as saving a drowning person. How are we to account for this unified and unifying sense of agency that is not merely driven by sub-personal brain processes? There are really two interrelated questions here, and much of their mystery stems from their interrelatedness. The question as to how a mental state can cause a physical event is only modestly mysterious if we assume a simple identity between certain combinations of mental and physical events. It becomes much more mysterious when we ask how a unified act of will on the part of a unified decision-making agent can cause the various micro-level physical events which all together comprise the physical substrata for this very agency itself. That is, how can the “core self,” simply in the sense of a unified being (not necessarily with sophisticated action planning ability or moral agency), voluntarily initiate the movements of its own bodily components? This last question is mysterious for two reasons. Not only can a desire in the brain command the body to move, but also, and more paradoxically, a unified process called the “self ” or “person” can command the component processes in the brain to conform to the patterns of activity needed to execute the relevant thoughts, feeling states, perceptual imagery, and attentional activities. This notion that a whole being can somehow cause the movements of its own parts seems at odds with most scientific accounts of consciousness, in which the whole is built up from interacting components whose activities are caused in quite piecemeal fashion by various inputs, micro-mechanisms, and basic chemical reactions. We have already made some progress in developing resources to deal with this enigma, but more are required. We have seen that the micro-constituents do not exhaust the causal power of a self-organizing system, once we take account of the system’s ability to rearrange the background conditions under



 Curious Emotions

which one or another causal event at the micro-level may or may not be sufficient for its consequents. But again, this is as true for sub-personal as for personal ones. It is true for growing cancers as for conscious or unconscious decisions to execute bodily movements.

. Why not an illusory-choice model? It is tempting to just bite the bullet and opt for a solution which posits that the causal power of choice or purposeful intent is an illusion – as, for example, Wegner (2003) and Dennett (2003) have argued. We could then simply posit that the movements are caused by interacting sub-personal mechanisms, and leave the mental intention out of the causal picture. The feeling that we choose to act would then be explained as a feeling that results from those same subpersonal mechanisms – an epiphenomenon of the events that exert the real causal power (Dennett 2003; Jackendoff 1996; Searle 1984; Smart 1963, 1970; Wegner 2003). Against this purely “bottom-up” solution, it can be argued that the feeling that our consciousness often does play an active role in organizing the physical processes leading to action is not a mere illusion. Voluntary movement really is different from an unconditioned reflex or a habituated neural-firing sequence (Jeannerod 1997; Spence & Frith 1999). In playing a well-practiced piece on the piano, for example, most of the movements are not consciously initiated: as soon as I tell my hands which piece to play, they automatically execute the sequence of notes, as a series of micro-mechanisms that require little direction except from the micro-mechanisms themselves, orchestrated mostly by the cerebellum at an unconscious level (Schmahmann 1997). However, this is true only until I need to modify a learned sequence of motor commands. Then I must switch to the mode of voluntary, global directing of the micro-events (Jeannerod 1997; Spence & Frith 1999). I must purposely decide to hit the G before the A, and deliberately command the selected finger to move (Sudnow 1978/1999). There are both phenomenological and physiological differences between these two modes. It is not just an illusory difference. In the effortful movement, more widely distributed brain processes are quickly activated, and the whole pattern of the organization seems to be commanded, not just by the cerebellum, but in some sense by “me” as a whole. Brainstem arousal mechanisms activate neurotransmitters that permeate virtually all parts of the brain (Panksepp 1998; Watt 2000), and looping signals integrate the functions of the

Chapter 5. Varieties of extended self and personality 

cerebellum, thalamus, hypothalamus, frontal and prefrontal areas and the anterior cingulate (Schmahmann 1997), which then lead to imagery involving parietal, occipital and temporal lobes, and tentative action imagery involving the motor cortex and supplementary motor area (Damasio 1994, 1999; Jeannerod 1997). In short, virtually the whole brain must be integrated at the point when a new decision is made, by contrast to the linear command sequences that are sent from the cerebellum when it is functioning on “automatic pilot,” by means of conditioned responses and habitual motor programs (Schmahmann 1997). In the deliberately self-controlled mode, a pianist can decide to use the fourth finger or the fifth, or play loud or soft, depending on the effect intended or wanted. Widely distributed aspects of the pianist’s being participate in forming such motivations, as indicated by the analyses of the motivational brain systems just mentioned, and from this holistic formation flows the sequence of mechanisms needed to make the self-directed behavior happen. Widely distributed but holistically unified aspects of a golfer’s being must execute the decision as to when the swing will start; and if so, then there cannot be an isolated or highly localized mechanism that is responsible for the decision. In the case of intentional choices, we seem to be in a very different realm from a simple linear mechanism, such as amygdala activation by the sudden sight of a snake (as in LeDoux 1996). In that case, the standard explanation is quite simple, involving a very localized sequence of physical and chemical mechanisms. In the case of forming the mood that is to be conveyed in music, the needed brain activity seems to be extensively global, simultaneous, and dependent on the precise timing of billions of micro-mechanisms occurring all through the brain all at once (Haines et al. 1997). The overall pattern, which is the motivational state, must in some way organize the activities of the microconstituents. And this seems to be just the opposite of a situation in which the overall pattern is caused by those same micro-constituents. Some will say that there is no paradox here, since any conscious state, whether global or localized, is just caused by the interaction of previous physical mechanisms. While it is true that most scientific accounts of consciousness and mentality (perhaps excluding quantum brain theories) will regard every psychological state as having been caused by some combination of past events, this is not where the paradox lies. The paradox is how a holistic organizational pattern – once having been caused by whatever sequence of past events – can then have the power to organize its own component mechanisms so as to ensure that they will behave as dictated by the holistic organizational pattern. If an affective intention is to motivate behavior, then in order to maintain itself

 Curious Emotions

this affective state must use numerous simultaneous shunt mechanisms to ensure that just the right micro-events occur (Panksepp 1998), and that just the right timing is maintained among billions of neural firings (Anderson 2000), to allow the intended quale to continue being felt while at the same time all the micro-level events are being coordinated in such a way as to facilitate the flow of chosen actions from the affectively motivated choices of the unified self. The problem of mental causation can thus be formulated in terms of the need to choose between two theoretical alternatives, each of which entails its own difficulties. Either (1) a whole organizational pattern, occurring at the level of personal choices and intentions, has the power to control its own subpersonal physical constituents rather than only the other way around – which seems to fly in the face of causal closure at the sub-personal level (which in turn seems to be an empirical fact); or (2) the feeling that our choices determine our actions in a non-trivial sense is only an illusion. Let’s take a closer look at the second of these two alterantives. A classic finding that is often interpreted as supporting the illusory-choice solution is that, prior to a subject’s awareness of deciding to execute an action, the neurophysiological mechanisms that cause the action are already reliably measurable (Libet 1999). A measurable readiness potential is observable .5 second before a willed action, whereas the subject is aware of the choice only .1 second before the act. Libet assumes that this means that the actual choice occurs unconsciously .4 second before we consciously will it. This assumes that the readiness potential is the correlate of the initiation of the action, and this seems to be the standard assumption among those who cite Libet’s readiness potential as evidence that conscious choices are only causally irrelevant epiphenomena. If conscious intentions are merely epiphenomena, then the same could just as well be true for unconscious intentions. The apparent causal power of all purposeful intentions, even if unconscious, could just as well be a subjective illusion masking the real causal mechanisms, which are at the micro-level. The problem here is not limited merely to the realm of consciousness theory. Actually, Libet tries to save “free will” from this problem by postulating that the conscious choice, which occurs .4 second after the readiness potential, has a “veto power” over the act that has already been unconsciously caused, and that this veto power is not predetermined and is the source of free will. Although (conveniently) there currently is no way to measure the physiological correlates of the veto power, Libet posits that we know it must exist because of the selfreports of the subjective experience of subjects. He appeals to the sense that we can change our minds after having just begun to execute an act. Obviously,

Chapter 5. Varieties of extended self and personality 

a hitter in baseball must be able to check a swing much more quickly than the .5 second interval between the readiness potential and the actual swing. Libet concludes that an immaterial mind may still have a veto power over the physically initiated readiness potential. Libet goes on to say that, if this veto power itself were physiologically predetermined, then free will would be an illusion, and our conscious choice would have no causal power of its own, but would be a mere epiphenomenon. In Libet’s view, if we give up a metaphysical dualism of mind and body, then we also give up any causal power on the part of mental choices. He ignores the possibility, of course, that the choice itself could have its own physiological correlates at the point when it occurs, and thus could have the same causal powers that these physiological correlates have. He seems to assume here that, if physical event A realizes mental event B, and then B causes C, this means that B did not “really” cause C, since A “really” did. I.e., he assumes that if B itself is sufficiently explainable by its own physical realizers, then B can have no real causal power. This is in sharp contrast to most psychophysical identity theorists, who attribute the exact same causal powers to a mental event and its physical realizers, on the basis that the two are identical. In short, Libet’s assumption is that there must be a contra-causally free will in order for there to be a “will” per se. He therefore ends up endorsing a straightforwardly and admitted dualistic conclusion (Libet 1999). An alternative explanation of the Libet findings would be that the readiness potential correlates not with the decision to execute the action, but with a consideration of whether to execute that action, or an imagining of the action that is being considered, in order to then decide about it. According to brain imaging studies (Jeannerod 1994, 1997), imagining an act requires sending efferent action commands, but then inhibiting them. So what the readiness potential is actually measuring may be the imagining of the action, which already involves motor cortex and sensory motor area activation. It may not correlate with the decision to act or to the activation of the action at all. Still another alternative reading, which would be essentially equivalent to this one, would be that, at the point when we have made a decision to act, we have not yet completely committed ourselves to the action. We have initiated the action command, which stands ready to be frontally un-inhibited only if we do not in the meantime decide to “veto” it. This is in no way counterintuitive, and does not imply that the decision to act does not cause the action; the physiological antecedents of the tentative decision may precede the time at which the actual decision is made, by allowing the action to go forward uninhibited. On the other hand, when we have made the tentative choice, we have not really

 Curious Emotions

chosen yet, because we still have not decided whether to veto the action. But whether actions are decided at the point when they can no longer be vetoed, or prior to that point, the crucial question is whether their causal power is an illusion. Libet’s findings do not support this thesis. What they support is that (1) entertaining the notion of acting in a certain way, while leaving open the door to finally inhibiting the action, is different from (2) deciding once and for all to go through with the action – and there may of course be a time lag between these two points. The readiness potential, even on Libet’s own analysis, accompanies (1), not (2). Empirical experiments could be devised to decide between these two alternative explanations, and either of them would be consistent with a dynamical systems solution to the mental causation problem, which in my view allows that the person can have the power to control the interaction of micro-components, provided that the person’s ontological status can be understood in terms of patterns of physical self-organization which do not contradict ordinary causal laws. We already have seen in previous chapters that, if a dynamical system has the power to rearrange the background conditions needed for various microlevel causal sequences to obtain, then the system has a type of causal power over its own micro-constituents and subsystems that is not merely an illusory effect of the causal workings of the micro-constituents themselves. If process A is necessary and sufficient for process B, and if A could be realized by a number of different combinations of micro-constituents, x, y, and z, then neither x nor y nor z is necessary or sufficient for B. Moreover, if the overall system subsuming the flow from A to B is a complex enough self-organizing one, then it may have the power to actively seek out and replace the needed micro-constituents to subserve A, so that not only is A multiply realizable in terms of x, y, or z, but the system as a whole has the ability to ensure that A is likely to be realized, by seeking out y or z in cases where x is not available. The problem is not whether the relation that forms the higher-order, multiply-realizable pattern can be a “physical” relation. We saw earlier that this relation could easily be physical without thereby having its causal power reducible to the sum of the causal powers of its micro-level components. The question is only whether, in the case of mental intentions, the intention has real causal power over its own component processes in such a way as to make its body move. We have seen that there is no logical impossibility in this scenario. A whole self-organizing process can have causal power over its own micro-constituents. The real question, then, is whether this logical possibility can be fleshed out enough to provide for a notion of self capable of executing meaningful choice that is not merely dictated by its own component processes.

Chapter 5. Varieties of extended self and personality 

Now to flesh out this concept a little more specifically in terms of what is known so far about neurophysiology. Panksepp plausibly locates the most necessary aspects of this “core self ” in circuits dominated by the periaqueductal gray (PAG), which is immediately adjacent to the cerebellum, which of course is involved in coordinating automatic action routines. The PAG is involved in many kinds of feedback loops with the thalamus, hypothalamus and cerebellum as it monitors the body’s homeostatic problems and initiates actions. After receiving these various kinds of feedback about the body’s general homeostatic conditions, it also has the capacity to serve as the first step in transmitting an action command, since it projects efferent pathways to all the brain areas involved in coordinating and executing actions. It also controls the release of many of the neurotransmitters and circuitries associated with emotions. For example, “Probably the most important brain area for the actual integration of the overall anger response is in the PAG. . . ” (Panksepp 1998: 198). Indeed, a review of Panksepp’s seven basic emotion systems reveals that the PAG is the one brain area that is involved in virtually all of them (Watt 2000). Many of the actions initiated at the PAG level, with the help of the cerebellum and cerebellar-thalamic-hypothalamic-cortical loops, can be modified or inhibited after having been initiated, but the fact that the PAG is the starting point suggests to Panksepp that the PAG is the most indispensable core of the “self ” considered as the agent of action as opposed to mere reaction. Extending Panksepp’s thinking on this point, I suggested in an article coauthored with Schmahmann, Anderson and Newton that When brainstem-hypothalamic loops (1) sense a homeostatic imbalance for the organism, they (2) activate neurotransmitter pathways in the interest of general arousal, and begin sending efferent signals to alert the brain that action is needed. This process includes looping with the cerebellum, because the cerebellum can activate highly specific action sequences, based on past learning. The cerebellar-brainstem loops then send efferent action commands as far as the motor cortex. When (3) these action commands are inhibited by prefrontal signals, the result, as Jeannerod points out, is (4) action imagery. When the action imagery is fully informed by specific action commands in response to environmental stimuli, the resulting “object understanding” (Newton 1996) becomes an intentional image schema, which vaguely represents categories of possible objects, but only in terms of their action affordances. . . . (5) When the imagery resonates with the primary and secondary sensory areas (e.g. in the occipital or temporal lobes) either because the latter has been triggered by perceptual input or because the image schemata are so powerful that they activate the occipital or temporal areas on an efferent basis, the result is vivid perceptual (sensorimotor) imagery or actual perceptual consciousness. . . .

 Curious Emotions

The role of the cerebellum here is, instead of generating efferent commands to take some action in general, generating very specific action commands by initiating the thalamocerebellar and hypothalamocerebellar loops (Haines et al. 1997), and also hippocampal-cerebellar loops, so that the action that is commanded (and which then gets inhibited in the motor cortex to form consciously-accessible action imagery) is a highly specific action, rather than just an efferent command to act in a general way (i.e. “Don’t just stand there; do something!”). That this specificity arises early in the activation sequence has been seen experimentally. In eyelid conditioning the cerebellum responds within 20 ms of the delivery of the airpuff (Woodruff-Pak 1997); Coles et al. (1990) report finding a hippocampal ERP as early as 18 ms after presentation of a novel stimulus. Both of these ERPs occur long before the occipital 100N and 200N “mismatch negativity,” and probably reflect early activation of the hippocampal-cerebellar loops that Haines et al. (1997) report, along with cerebellar-PAG loops. (Schmahmann et al. 2001: 300–302)

We can think of the “core self ” – the agent in the most minimal sense – as a subsystem of the organism comprising those aspects that are needed to initiate action in the interest of the body’s self-organizational aims. But we can also speak of “higher order” selves than this. When higher emotional and intellectual processes come into play, they often can facilitate the body’s selforganizational aims more effectively than the “core self ” can by itself. Thus they too satisfy the notion of “self ” as unified agent of action. Indeed, they unify the organism more completely than does the core self. Damasio (1999) shows that certain neurological diseases and injuries can knock out certain intellectual or even attentional and perceptual functions, but without eliminating the person’s consciousness or sense that she is a person. When only cortical areas are eliminated, consciousness and the sense of self remain essentially intact. But as we consider neurological dysfunctions that affect regions lower and lower in the brain, more and more of the person’s ability to execute actions and make choices is lost. When the injury affects the upper brain stem areas, first the sense of self and the ability to function as an agent are gone, and then all consciousness is lost. However, this does not mean that other brain areas do not contribute to the identity of the person when they are functioning. In my view, an action is an action of the self, and not a merely mechanical reaction, in any instance where a large subsystem of the organism is acting to coordinate its own micro-constituents toward aims determined at the self-organizational level. This can occur minimally if only the subcortex and parts of the limbic system are intact, and these areas are certainly the ones that are most crucially necessary for even a minimal degree of agency.

Chapter 5. Varieties of extended self and personality 

In the top-down type of dynamical systems approach to purpose-directed action, the agent who makes the decision is a relation, requiring suitable physical substrata, and the motivated mental choice to raise one’s hand is another such relation, which I shall designate as R. This R is multiply realizable by a number of alternative sequences of physical mechanisms, or Ps, with each alternative sequence requiring its own set of background conditions. R is necessary and sufficient for the hand’s going up, whereas no particular set of Ps is necessary and sufficient for the hand’s going up except given certain background conditions which are substantially controlled by R. Note that we are assuming, as argued earlier, that in a suitably complex dynamical system R could have obtained without that particular set of Ps, by rearranging the background conditions in such a way that a different set of Ps could have subserved R. Thus, by the above definition of causation, R causes the hand to go up, yet we can also say that, given the background conditions that did in fact obtain, the Ps that subserved R also caused the hand to go up. But if the Ps that caused the hand to go up had been unavailable to do so, R could have rearranged the background conditions so that, under these altered background conditions, an alternative set of Ps could have caused the hand to go up. Not only can R meaningfully be designated as a cause of the hand’s going up, but in fact R is a cause of the hand’s going up under a broader set of possible background conditions than any given sequence of Ps. Moreover, R is both multiply realizable by alternative sets of Ps, and has some control over the realization of alternative sets of background conditions under which a given sequence of Ps will or will not be used as the particular sequence leading to the raising of the hand. Thus R has more control over the raising of the hand than any particular set of Ps subserving R can have, since it can be said that R is necessary and sufficient for the hand’s going up under many background conditions where that particular sequence of Ps would not have sufficed to make it go up (because the needed background conditions may not have been in place). The causal power of R therefore is not illusory, yet neither does it contradict the causal closure of micro-level sequences, because the micro-level sequence that actually does occur is necessary and sufficient to bring about the resulting action under the given background conditions. But these background conditions have already been influenced by a previous enactment of R, whose tendency is to manipulate background conditions in ways needed so that available microlevel substrata can produce the outcomes that will propagate R into the future. Nor does this causal power of R contradict the thesis that R itself may have been caused to be the way it is by some past set of micro-level events (for exam-

 Curious Emotions

ple, in ontogeny and phylogeny). The possibility that I may have been caused by a combination of environmental and hereditary factors to be the kind of person who would choose to study philosophy and psychology does not contradict the fact that I did choose to study them. If someone writes a will, we may ask for evidence that the person chose to write it to include certain specific contents, as opposed to being coerced by forces contrary to his or her own choice, and this question remains meaningful regardless of whether the personality of the person writing the will has been predetermined, resulting in a correlative causal predetermination of the person’s choice. In sum, we do make choices, and choices do cause actions, even though the actions can also be said to have been caused by substratum-level events, if certain background conditions for those causal relations have obtained. However, when a mental choice occurs, the choice is an aspect of a motivated self-organizational process that is a relation subserved by the relevant substratum events, and this self-organizing process itself is structured so as to have a strong tendency to go out and seek the needed substrata to keep its pattern of organization going. The relational process, then, has the power (within limits, of course) to replace its own components, and this power is essentially the basis for the causal power of mental choices over our bodily movements. Throughout this book, we have seen details of the way various motivational situations can be understood as the results of a self-organizing system. But now we must consider how the concept of primitive agency we have developed here (not necessarily complicated action-planning or moral agency, but simply the ability to act rather than merely passively react to external inputs, or for internal micro-events merely to passively react to each other) can help us to ground a concept of the self in several senses in which we want to say that human beings are or have “selves.” Two of the most important of these are what I am calling the “personality” and the “embodied self.”

. The embodied self and the personality In one sense, the self is different from each of its momentary conscious states in that the self is what gives unity and direction to the overall stream of states. This is why we recognize the difference between different individual personalities: The way we identify a person’s uniqueness has nothing to do with the fact that the person happens to experience this or that conscious state; most people either do experience or could experience most states, so experiencing these or those particular states is not what distinguishes their personalities from each

Chapter 5. Varieties of extended self and personality

other. Had I not experienced the pain of hitting my foot on a rock this morning, I would very likely still be the same person. If not in terms of a collection or bundle of states, then, how do we define personality in a sense that could unify these states into a self? One strategy is to identify the recognizable style with which the person chooses at each moment in the stream of consciousness what to direct attention toward at the next moment. The overall pattern of these attention-directions is what we recognize as the personality. This is also the sense in which the “psychophysical self ” is the ontological foundation for the “transcendental ego” in the thinking of Husserl (1931, 1970), and which Sartre (1937) denied – something that is not confined to one particular conscious state, but does have a subjective status, and at the same time serves to unify all the other states into a coherent pattern. This Husserlian transcendental ego should in no way be confused with the “transcendental self ” of Eastern philosophy, which is a mystical and metaphysical conception. For this reason, I shall refer to it simply as the “personal self ” or “personality” – that which provides the unity and direction across various conscious states, as opposed to the states themselves. A person could still be the same person without having experienced this or that particular state. What could serve both these purposes at once – on the one hand, unifying all the subjective states in the stream, while on the other hand having a subjective character in its own right? A likely candidate would seem to be the person’s emotional motivation to maintain and enhance consciousness per se. An emotional motivation can unify all the particular states, by motivating the direction of attention at each moment, toward homeostasis when that is what is needed, and toward greater complexity when that is what is needed. And at the same time, emotional motivations are subjectively experienceable, so that, contra Sartre, this emotional motivation can both serve the purpose of unifying all the conscious states in the stream and at the same time be a conscious state (Ellis 1986, 1996). The motivation referred to here can be thematized in terms of the organism’s motivation to maintain and enhance its pattern of self-organization. The ontological status of the personality, conceived of in this sense, might be viewed as a higher-order pattern composed of the particular conscious states in the stream. Given just one such state, we could not discern a personality or self in this sense; it takes a continuing progression of them to form such a pattern. So one way of defining the personality would be to say that it is a pattern formed by all the states in the stream, and which tends to maintain a certain contour or style in spite of variations in the particular states themselves.





Curious Emotions

But in contrast to this concept of the self as a pattern of states, we have seen that there is a simpler one that focuses on a more basic and elementary role that the self seems to perform within each state. Given any state of consciousness, there is a purpose-directed and aim-oriented dimension of the state, and this dimension also could be called a person or self (Newton 1996; Wider 1997). As we have already seen, evidence that motivational and affective processes are necessary for conscious cognitive processes (as opposed to nonconscious information processing) has been mounting, and it is now well known that afferent perceptual processes can be completely activated with no perceptual consciousness of the object occurring (Luria 1980; Posner & Rothbart 1992). Consciousness occurs only when the efferent system, beginning with the emotional brainstem/midbrain/limbic area, prompts the frontal lobe to start looking for (and thus generating vague image schemas of) environmental conditions that might be of interest to the organism, in light of its emotional-motivational purposes, and activates neurotransmitters to catalyze the corticothalamic loops needed to enhance important signals and direct either voluntary or involuntary attention to them (Ellis & Newton 1998b; Posner & Rothbart 1992, 1998). So it can be argued that, within each state of consciousness, there is an affective dimension at work, doing the directional work of a self. But this notion of a self within each state does not really contradict the idea of a self that unifies various states. We can conceive of the person as the selforganizing system whose motivational tendencies coordinate the pattern and direction of cognitive contents by motivating, at each moment, what is important for us to direct our attention toward in the next moment. Of course, this definition would extend “personhood” far beyond human beings; we could speak of cats, dogs, or even frogs as “persons” or “selves” in this sense. (The philosopher B. P. Bowne even goes so far as to define personhood in terms of value-directedness, and then conclude that even the ecosystem is “personal” in a primitive sense.) What makes something a self, in this sense, rather than just an inanimate mechanical system, is that it is purpose directed, that it acts on its environment rather than just reacting, and organizes itself in order to maintain its existence rather than merely having its organization be the result of external mechanical processes. Such a pattern actively seeks out, appropriates, and replaces the material substrata needed to maintain the overall pattern into the future. To be a self in this sense is simply to be an organism that lives and works concertedly as a unified being toward purpose-directed aims. Even insects would qualify as selves in this sense. A self is anything that strives; it wants, desires, is able to be “egoistic.”

Chapter 5. Varieties of extended self and personality

But this concept of self still seems different from the notion of a higherorder pattern formed by different states, because it seems to be present within one state. In principle at least, there might seem to be no need for a higherorder pattern of conscious states to make this kind of “self ” ontologically possible; it can be there within any conscious state, as the “self-ish” aspect of that state. For example, a desire to eat already includes an aspect of the desire which is sufficient for a concept of “self.” So it seems that, even if we confine ourselves to the “person” or “personality” dimension of self, this dimension can be conceived of in two different ways: We could think of it as a higher-order phenomenon, formed by relations between conscious states across time; or we could think of it as a lower-order phenomenon, the motivational component of any one state. While these two concepts of self initially appear to be different, we shall soon see that they refer to the same process when contextualized vis a vis the organism’s emotionalmotivational life.

The importance of the higher-order self. The notion of a higher-order self formed by the various states in the stream is important, because it makes possible distinguishing between different unique, individual selves or personalities. We know that Chopin is a completely different personality from Liszt, because, even though they may both have used most of the same chord progressions at various times, Chopin would not have used chord progressions X, Y and Z in the same pattern and in the same contexts as Liszt would. We distinguish individual personalities by seeing, not what specific states of consciousness they have, but by seeing the pattern formed by those states. Also, this higher-order concept of personality is important because it shows that emotion and motivation are among the essential determinants of all states of consciousness; emotion determines a continuous direction in which we will move from one state of consciousness to the next, and again to the next, across a good portion of time. And this directional continuity also indirectly determines the content of each subsequent state. For example, in the religious realm, a motivation to live eternally can motivate us to direct our attention away from the Problem of Evil, and toward the most convincing proofs we can find to convince ourselves of the supposedly cognitive and intellectual belief that God (in the sense of an all-powerful and all-good being) exists. Although the resulting belief appears to be based on intellectual and empirical premises (the cosmological argument, the argument from design, etc.), it also and more importantly results from the fact that the person has previously been motivated to direct attention toward certain things and away from others. Attention is likely



 Curious Emotions

to be directed away from the Problem of Evil and other arguments that might tend to undermine the belief system. The content of intellectual cognition is therefore driven by a more basic emotional level. This higher-order concept of self, as a pattern that obtains between states in their motivated sequences, also seems to make possible non-consummatoryreductive motivations in a way that the other concept of self – the self within one state – does not appear as easily to do. If the self is the pattern, then the self can be motivated to make an interesting pattern, and is not imprisoned within one particular state’s reductively driven desires. The pattern could then organize its components in a coherent direction that has meaning apart from the smaller purposes of each component; this requires both homeostatic motives and extropic ones. But this idea of a higher-order “self ” might seem to contradict the claim that the stream of consciousness is already directed by a lower-order self that exists in each state of consciousness. Another reason it conflicts with the notion of a lower-order directional self is that such a lower-order self existing in a specific state of consciousness might be easy to conceptualize and explain in terms of drive-reductive aims, whereas it is harder to make sense out of the higher-order pattern without granting the importance of non-reductive motives. Thus the two concepts of self – personality pattern versus single-state subjectivity – might even appear to reflect conflicting motivational processes. And, as we have already seen, one argument for the higher-order self is that it renders non-reductive motives possible in a more intuitive way.

The importance of the embodied self. Against the appeal of the higher-order self, the simpler concept of self also has important advantages, because it makes possible “self-reference” on the part of any particular state. When I say “my consciousness of wanting to catch the streetcar,” implicit in this consciousness is the “me” that wants to catch the streetcar, and there is no need for a higher order pattern of states in order for that to exist. In principle, a being that existed only for one second, and spent that one second chasing a streetcar, would have a self, because implicit in its chasing the streetcar would be a self (an “I”) that wants to catch the streetcar. The subject of the experience, simpliciter, is a self or ego. And this is what we seem to refer to when we “reflect” on an experience; the object of the reflection is the subject of that experience. The higher-order pattern used in the “trans-temporal” conceptualization of self mentioned above is an abstraction with respect to the immediate subject-of-experience component of each of the particular experiences, which are concrete. Since the self, if it is to exist, must be concrete and not a mere abstraction, we want to say that

Chapter 5. Varieties of extended self and personality

it is something to which it is possible to refer in any given experience, as the possibility of reflecting on that experience and seeing that it has a subject. Many philosophers are attracted to this simpler concept of self because it can serve as the basis of the subject that is capable of reflection on its own conscious states in any given moment. To say that the self is simply the subject of its own consciousness can be consistent with a dynamical systems view of agency, since agency in the primitive sense is also necessary for consciousness. This simpler concept of self – simply as the subjective side of any momentary conscious experience available for reflection, regardless of the way the stream of experiences is organized – is also attractive because we can explicate it in terms of concrete bodily experience in the following respect: In any given perception, there is a Gibsonian “affordance” attached to understanding the meaning of what is perceived. I identify the carpet I see by imagining what it would be like to crawl on it, run my hand over it, etc., although these imaginations are usually kept at a preconscious, habituated level. (Gibsonians, such as Newton 1996, would say that, earlier in infancy, the infant had to sense in some way what those affordances would be like in order to appreciate the identity of the carpet, and then the affordances became habituated and sedimented.) For example, in visual perception, a landscape presents a Gibsonian affordance of possible patterns of eye movements that could be easily executed in relation to that scene (Pribram 1991), and the eyes are continually engaged in active rather than passive movements and oscillations. As every artist knows, some scenes are “easier to look at” than others, and patterns on a canvas can conjure moods by facilitating and interfering with patterns of eye movement in the viewer. Newton (1996) supports this viewpoint by pointing out that the “understanding” that distinguishes humans from computers is based on the human’s ability to plan actions. Here again, humans act, whereas merely mechanical systems only react. To understand a planned action means to be able to form an imagistic/proprioceptive bodily sense of what it would be like to execute the action. To understand an object is to form a similar kind of sense of what it would be like to execute an action in relation to that object. Although we may believe that red is pasted to the surfaces of objects, in reality the red that we see arises only through our interactions with the object; as Merleau-Ponty (1942) emphasizes, if we (or our perceptual systems) did not act, we would not see. So we can explicate the simpler concept of the self by answering this question: When I reflect on my “self,” what am I reflecting on? The answer would be that I am reflecting on my embodied sense of what the affordances would feel like. I.e., when I “inhabit” my left hand as it is feeling what the right hand



 Curious Emotions

as an object feels like, my “being” or “self ” is “in” the left hand rather than the right; then I can shift to inhabiting the right hand, and sense what the left as an object feels like (Merleau-Ponty 1948/1968). Thus the difference between reflective and nonreflective experience can be understood in the following way: In nonreflective experience, we focus attention on the object of the experience rather than on the subjective affordances that make the experience possible; whereas reflective experience focuses attention on the bodily sensing of the affordances themselves. When I touch a tree trunk and feel the texture of the bark, I am focusing attention on the bark. When I allow my attention to inhabit the hand itself, feeling the bodily sensing of the texture, I am inhabiting my “self,” and thus reflecting on the self, the subjective side of experiencing. In phenomenological terms, we could say that to reflect on the self is to inhabit the noetic rather than the noematic side of any given experience (Husserl 1931). This concept has the advantage of making the self into an embodied being rather than an abstraction; the self is already an aspect of the body at any given moment, and can be reflected upon as such. In sum, both concepts of the personal self, the higher-order and the embodied, can offer persuasive justifications for themselves. Yet the two seem irreconcilably in conflict. Is there a way to combine the two?

Reconciling the two concepts. The self-organizational view entails that both these concepts of “self ” are aspects of one unified concept. Within each specific motivational tendency inherent in each conscious state, there is already a tendency to move toward greater and greater complexity, newness, variety, etc. – i.e., toward “extropy” rather than merely toward maximally energy-efficient homeostasis. If a momentary conscious state is motivated by a purposeful component, and if that purposeful component already includes a desire to maintain and enhance the overall organizational pattern of the organism as a whole, then the higher-order-pattern concept of personality is already included in each momentary state. Each momentary state is geared toward making a certain higher-order pattern possible, and the higher-order pattern at stake, in order to be “extropic” in addition to being homeostatic-drive reductive, must want to lead to higher order patterns that would include greater variety, complexity, etc. Its aims cannot be realized in one specific state; for there to exist a fundamental motivation toward “extropy,” a desire to maintain a certain energy level, to avoid complete rest, requires the higher-order pattern dimension. The concept of the embodied self suggests that, when we reflect on consciousness, what we are doing at the concrete level is to pay attention to our own intentional action in relation to the object of consciousness. Since under-

Chapter 5. Varieties of extended self and personality

standing an object always involves an implicit imagining of how it would be possible to act in relation to the object (the object’s “affordances” for us), we can reflect on the body’s orienting itself to the object in this way. On the other hand, when we have “unreflective” consciousness of an object, what we are doing is to intentionally act in relation to the object in this same sense, but in such a way that we do not pay attention to the acting that our bodies are doing. When we speak of the self as the embodied self, we mean that our self can inhabit any part of the body that we pay attention to. Whenever this embodied self is in a certain place in our bodies, that means that we are paying attention to how we act (as opposed to reacting) in that place. To reflect on the embodied self, then, is to reflect on this acting on the part of the body (as opposed, again, to merely reacting), to focus on oneself as acting in relation to the object of experience rather than focusing on the object itself, even though it becomes an object for us only through our acting on it, whether we focus on this acting or not. We constitute the appearance of the object by acting in relation to it; but only when we “inhabit” the acting of our bodies do we “reflect” on our embodied self. Yet the ability to act is itself a higher order process, an organization in terms of a complex, self-regulating system. Thus the self can be both an aspect of the body – an embodied self – and a higher-order self, i.e., a higher-order pattern than most specific states of consciousness are, which serves to organize and direct the specific states, and which could still have the same identity even without any particular one of those specific states. Moreover, the specific state could not exist except within the context of a higher-order pattern motivated toward maintaining and enhancing the pattern, regardless of whether certain momentary states have to be included or excluded in order to achieve that. For example, a person motivated toward a monastic existence can arrange, to a great extent, to exclude consciousness of sexual temptations from the stream of consciousness by carefully planning a way of life that avoids them as much as possible. The desire for the overall pattern is what counts in our personality makeup, not the desire that we would or would not have in the face of any particular stimulus. Presumably, the monk would be sexually interested in a sexual stimulus, just the same as anyone else, and this is not what makes him the person he is; what makes him who he is is the fact that he arranges the higher order pattern in the way that he does. Now, if we grant that the higher-order pattern is already implicit in the self-organizing pattern of the person’s being at any given moment, we can have an integrated concept of an embodied yet higher-order and trans-temporal self that unifies and gives direction to the various particular experiences within the



 Curious Emotions

stream. But so far, this concept is merely a concept. If the concept is not to be empty, we must also ask how it is that we can know about ourselves, in both the embodied and the higher-order sense.

. How can there be knowledge of the self? Suppose we grant for the reasons just outlined that each of us has a somewhat unitary “self ” or “personality” that includes the properties of subjectivity, agency, and higher-order personality. The problem remains that none of the arguments just presented seem to enable us to know anything about any particular, real-life self. Most importantly, they seem to provide no way of knowing about our own particular selves. In this sense, both concepts of the self are “mere abstractions.” If we cannot know ourselves, then there is still the possibility that, in each person’s case, the self in terms of which the person lives is merely a construction. Moreover, if we do construct selves for ourselves, then it may be these constructions that play the role of transcendent unification discussed earlier; the same constructions may also posit the values that motivate the direction of attention even at the level of the embodied self. Just as Eskimos immediately and directly perceive eleven distinct kinds of white, because they are motivated to do so, our constructions of self may motivate us to direct attention in the ways we do, and these values will also manifest themselves in our embodied, lived experience. So, even if we can reflect on this embodied lived experience, still, if the self is ultimately a mere construction, then either there is no “true” self to be known, by contradistinction to our “false” views of ourselves, or our views of ourselves are constructions that do not necessarily reflect any “truth” about ourselves. In either case, “self-knowledge” would become a dubious project. It is true that the subjective life has public manifestations; Beethoven’s numerous romantic infatuations, about which he was somewhat secretive, show up in his music. But the public manifestations may or may not tell us everything we could want to know about Beethoven’s self, and what they do tell depends on very subtle and difficult inferences that we might well get wrong. It is no coincidence that some students of Beethoven see him as an egomaniacal, wooden tyrant, while others see him as sensitive, caring, and lovelorn. Some of these constructions may resemble the real Beethoven, but they do not exhaust, let alone determine his personality.

Chapter 5. Varieties of extended self and personality 

Of course, the possibility remains that Beethoven himself may have “invented” the pattern of organization that is the personality “Beethoven.” But here we run into the problem of authenticity. We often invent personas for ourselves that include aspects quite different from the way we really are. We pretend to feel or think differently from the way we really feel and think – pretend these things even to ourselves. But pretending them does not make them so. There is still a real self, in both the embodied and the higher-order sense, even if we have a distorted view of it. So the question remains, how can we know about ourselves, other than knowing in an abstract philosophical way that there is some self, of some description or other? How can we know what our own personal selves are like? One of the prominent sources of misunderstanding of the emotional self is that it offers a temptation to a strong special case of Husserl’s “natural attitude.” Since our understanding of the material world is based largely on sensory imagery, there is a tendency to think of the self in terms of such objectified sensory imagery “of ” the self. The result is that we construe the self as if it were an external object, as if seen by someone else. This tendency leads not only to a misunderstanding of the self, but to a fundamental type of psychological condition rooted in such a misunderstanding – the narcissistic disturbance, which in its most advanced stages can rise to the level of “borderline personality disorder” (Lowen 1985). To confuse imagery of the self, in the everyday sense of “imagery,” with the self itself is the hallmark of narcissistic disturbance (Klein 1975; Kohut 1985). The narcissist thinks of herself as an object, as viewed from an external perspective, as the protagonist of a story as read by someone else, or by oneself in the act of viewing and judging oneself as if from an external perspective (Horney 1937, 1950; Miller 1981). This is not to deny that the experience of a “true self ” (I argued above that there is one) also crucially involves imagery. But this will be “imagery of the self ” in a different sense. The terms “imagery”and “self,” as well as the word “of ” linking them, have all remained notoriously ambiguous throughout the histories of philosophy and psychology. “Imagery” is often thought of as limited to visual imagery, or imagery in other sensory modalities such as audition. But there is also sensorimotor and proprioceptive imagery. One can imagine the way a leg cramp would feel without actually having one; or one can imagine oneself performing a bodily action such as grabbing one’s left foot, without actually doing so: these are proprioceptive and sensorimotor images, or what Jeannerod, Newton, and others have called “action images.” The enactive view implies that action images are crucial to understanding what the true self is.

 Curious Emotions

Sartre (1943/1966) distinguished between the “self-for-others,” as an object of experience, and the “self as it exists in itself ” – as a subjective entity or process, which may not be accurately transparent to itself. But it was also Sartre who first popularized the view that the second type of self, the “self in itself,” does not exist, and that there is only a narrative self, a constructed set of images that we tell ourselves are of ourselves, in which the protagonist is seen as an object, just as someone might see a person other than oneself. Hence Sartre’s (1943/1966) notorious claim that Peter knows just as well as Paul what Paul is feeling. I have rejected this argument in many places (for example, Ellis 1986, 1995, 1996b), and shall continue to hold here that there is a difference between the subjective self and the constructed self. For consciousness of the subjective self, it is action imagery and imagery of others that is central, not pictures of the self-for-others as in the attempt to view images of ourselves as we might appear from an external perspective. To offer a narrative account of myself is to present a series of objectified images that in principle are intelligible as well to others as to myself. Such a narrative thus captures primarily the self-for-others, and presents the temptation to construe myself as equivalent with this objectified series of images. As far as a subjective self is concerned, the narrative can capture it only to the extent that it reveals different subjective responses to the same objective events that others might also have undergone. Thus the events in the narrative do not touch on the nature of the subjective self per se, and less still do objective images of the self that is undergoing or creating the events. There can be images “of ” the self, as in looking at pictures of our bodies or stories about ourselves as seen from an objective or external point of view. These images of the self are images of the narrative or constructed self. But in another sense of “of,” we could say that even these images of the narrative self exist in the consciousness “of ” the true self. They are images “of ” the self in a different sense – they are images had by the subjective self in the attempt to experience itself. If the images are images of the self-for-others – the self as viewed from an external viewpoint, as an object – then what is viewed is merely the narrative self. But if the images are action images had by the subjective self in the context of action, and at the same time this action imagery helps to capture the true self in its intended actions, then the sensory images involved are not pictures of the self-for-others, but rather pictures of the objects in relation to which one acts, would hope to act, fear to act, etc., along with sensorimotor and proprioceptive action imagery as one evolves real or imaginary action schemas from the valuational emotions that motivate the action. As these action images become very complex and are related to complex environmental affordances,

Chapter 5. Varieties of extended self and personality

the question of course becomes increasingly pressing to what degree even these action images can diverge from an accurate reference to an authentic sense of the self. While “healthy narcissism,” including a narrative self, is a vital part of the self ’s functioning, too exaggerated a preoccupation with the narrative self, to the point where the narrative self is equated with the subjective self, or treated as more important than the subjective self, is a hallmark of narcissistic disturbance. But to argue in this way also requires showing that there is something in our actual experience that the label “subjective self ” points to. The imagery of the “self ” that is relevant here is not primarily objective, but consists of imagery of the self ’s actions in relation to the world, including the others toward whom one acts. For the narcissistically disturbed person, by contrast, objectified imagery of the self is so prominent in the self-concept that even the subjective self is relegated to a mere instrumental value, since the main purpose is now to view oneself as valuable, through one’s own eyes and the eyes of others, as an accomplished, amusing, or in some way other-pleasing persona. For this reason, the narcissist is caught up in the habit of measuring the value of himself and of others instrumentally – as depicted in objectified imagery. Rather than imagery of sensorimotor action schemas as they would relate the self ’s actions to the other, the imagery is caught up with the way things look, objectively. As played out in terms of narrative imagery, the objective events of my life are to a great extent replaceable; the constant for each person is the way he or she organizes and directs attention and the self-organizational actions that make experience possible. If Harriet Taylor had never met John Stuart Mill, would she have still been more or less the same person? The specific experiences of her life did not define her as who she was, since others could have had the same experiences. (After all, many other people did in fact meet John Stuart Mill.) As we have seen, there is an important sense in which the style in which the person organizes and patterns the flow of experiences is more important for defining the self than the what-content of those specific experiences. This patterning in the flow of experience is motivated actively, and thus can ground a concept of a subjective self that is not experienced in the mode of objectified imagery, but rather in action imagery. We have seen that the subjective self arises essentially from the selforganizing system’s capacity for action as opposed to mere reaction. It anticipates emotionally significant experiences by seeking out and appropriating opportunities to express schematically patterned action potentialities. A narrative concept of self will miss its mark, erring on the side of narcissistic images, if it seems to equate the self with the narrative, and does not capture the fact that



 Curious Emotions

if the same objective events were to occur in the lives of different individuals, there would be extremely different subjective responses to them and different uses of them for purposes of subjectively-motivated action. When Chopin was offered a series of concerts with the large audiences on which Liszt capitalized so successfully, Chopin sometimes took advantage of the opportunity, but he did so only reluctantly, as a sacrifice that must be made to secure the privilege of continuing to compose music and get it published. Objectively, his concert demeanor may not have appeared so different from Liszt’s. But subjective differences can express themselves across apparent objective similarities, and it is these differences more than the objective similarities – more than the objectivistic images in which people often are tempted to categorize themselves – that really identify the self as a subjective unity distinguished from others. Eugene Gendlin’s (1980, 1992a, 1992b) psychotherapeutic method well illustrates a way to avoid confusing the real psychophysical self with merely constructed objectifying imagery. Gendlin begins with the phenomenological observation that there is a hermeneutic circle when it comes to understanding what experience reveals to us, about either objective reality or about ourselves. The problem posed by the hermeneutic circle is that sedimented categories may distort the intentional meaning of any experience, to an undetermined extent. But in order to see the category structure as what it is, we must experience it in some way, and this experience too will be filtered through subjective categories. Thus our view of the categories itself may be inaccurate, so we can never know whether we have identified them successfully, nor whether we have successfully compensated for their distorting effect on our interpretation of our experiences. Affect plays an especially important role here because the categories reflect what is deemed important for organismic purposes, and “importance” is defined in terms of, and reflected by, our affective categories. Gendlin approaches this problem by noticing that, if a “bracketing” (or phenomenological reduction) of the categories that we use to characterize an experience serves to change the feeling or object in question, then the change must be epistemically positive – must move us a little closer to the truth about what we really think or feel. For example, if I ask myself whether my anger at my son for not taking out the garbage is really directed at my son, and if this asking serves to change the felt quality of the feeling toward my son, then this is an indication that I am on the right track in reinterpreting the feeling as not being about what my son did, or at least not primarily about that. This hermeneutic approach is also consistent with the enactive view that emotions are never passive reactions to objects, since active seeking and organization of experience is what makes the affective categories selective in the first

Chapter 5. Varieties of extended self and personality 

place. This requires a distinction between active and merely reactive processes, and Gendlin is firmly in Aristotle’s camp when it comes to this distinction: Affective categories are expressions of larger organismic purposes, not reactions to discrete stimuli. This is why the affective categories play such a powerful role in organizing and shaping the interpretation of all of our experience, even the experience of objective reality. But for this very reason, affective experience itself can reveal a great deal about the relationships between the self and its world. According to Gendlin, To begin philosophy by considering perception makes it seem that living things can contact reality only through perception. But plants are in contact with reality. They are interactions, quite without perception. Our own living bodies also are interactions with their environments, and that is not lost just because ours also have perception. . . . Our bodies. . . interact as bodies, not just through what comes with the five senses . . . . Merleau-Ponty. . . meant perception to include (latently and implicitly) also our bodily interactional being-in-the-world, all of our life in situations. . . . The body senses the whole situation, and it urges, it implicitly shapes our next action. It senses itself living-in its whole context – the situation. . . . From one ancient bone one can reconstruct not only the whole animal, but from its body also the kind of environment in which it lived. . . . The body even as a dead structure still contains all that implicit information about its environment. . . . My warmth or hostility will affect your ongoing bodily being whether you perceive it or not. You may find it there, if you sense how your body has the situation. (Gendlin 1992a: 4–13)

How do we know when we have adequately identified what a “bodily felt sense” of a situation is telling us about the situation? How do we know what a state of consciousness is “about” in relation to the “bodily sensed” environment? Gendlin offers as an example the “odd feeling of knowing you have forgotten something but not knowing what it is”: You are troubled by the felt sense of some unresolved situation, something left undone, something left behind. Notice that you don’t have factual data. Your body knows but you don’t. . . . You find a possibility. “Helen’s party! I forgot to tell Helen I can’t come to her party!” This idea doesn’t satisfy the feeling. It is perfectly true that you forgot to tell Helen you would miss her party, but your body knows it isn’t this that has been nagging you. . . . Your body knows you have forgotten something else, and it knows what that something is. That is how you can tell it isn’t Helen’s party. (Gendlin 1980: 38–39)

 Curious Emotions

The reason we know that a given answer will not work is that “I forgot to tell Helen I can’t come to her party,” though true, fails to elicit any “bodily shift” from the felt quality of the question as to what it is I am trying to remember. Ultimately, the fact that I have a body anchors me in reality, because the being of my body (in relation to other beings) is a reality. In effect, our interpretation of our own experience is laden with externally imposed presuppositions. When we bracket them, we use our imagination to envision what it would be like to experience a life situation minus some of those presuppositions. If our bodies react to this imagining with a perceptible “felt shift,” this feeling is a clue that we have hit on an important element of what the feeling was about. By continuing to separate presuppositions from direct experience in this way, Gendlin believes we can come closer and closer to an adequate understanding of what our feelings mean. Moreover, since the meaning of the feelings is action-oriented, and the action proclivities define the self, this means that we can come closer to a perception of who we are. That this method works the way it does is consistent with the enactive view of affect. If the self is an actor rather than merely a receiver of information, then we must understand ourselves by imagining how we could act in relation to the world, rather than in terms of receiving information about ourselves, as if we were observing ourselves from the outside. The “constructed” self may not be an adequate representation of what the self really is, but it does not follow that the self is not real. Constructed narrative selves tend to be misleading largely because we attempt to formulate the narrative in terms of objectified imagery, such as how we look or how our actions could be seen by others, rather than in terms of the way they feel from the inside. But, aside from narrative and constructed selves, there are also a number of different senses in which the term “self ” can refer to a real phenomenon. The narcissistic self is part of the working of this self-system, but should not be confused with the self per se. The narcissistic self is the self as seen by others and evaluated in terms of its instrumental values such as admirability, pleasingness and accomplishment. Our concepts of our own narcissistic selves tend to center around objectified imagery and narratives. But if we attempt to construct a real self based on such imagery, we fall into the trap of narcissistic disturbance. As we have seen, the real self must also include a subjective self, whose imagery centers not around objectified sensory imagery, but around complex sensorimotor and emotional action schemas as related to the action affordances of the world, and as we act in relation to others whose intrinsic value we can symbolize and experience in terms of the actions we can take or even imagine in

Chapter 5. Varieties of extended self and personality 

relation to them. If there is a streetcar to be caught, it is because the subjective self is a self-organizing system that really does want to catch it. In the next chapter, we can turn our attention to the way the motivationally driven selection for intentional objects helps toward an understanding of the intentional meanings of affective experiences, especially at the most subtle and sophisticated level that characterizes what we have called “higher human affects.” The way intentionality is explored in psychotherapy and the arts will provide important clues to the way the “aboutness” of emotions can be understood.

Chapter 6

Learning about emotions through the arts

In a study of the sociology of knowledge with respect to emotion studies, Bernard Baars (2000) suggests that scientific understanding in this domain has been severely limited by four self-imposed taboos of modern science. For methodological reasons, we shut ourselves off from the rich data available through (1) the study of consciousness; (2) the real-life emotional feelings of human adults in their natural habitat; (3) the subtle insights of the arts and humanities; and (4) the subjective data revealed in psychotherapy and psychoanalysis. During the past decade, the first two of these taboos – the ones against studying consciousness and emotion per se – have been suddenly and enthusiastically rejected by a new generation of cognitive theorists, neuroscientists, and philosophers of mind. Several new journals have specialized in the study of consciousness, and at least one (Consciousness & Emotion) specializes in emotion and its relationships to other conscious processes. In a book on the higher human affects, it seems appropriate (although a sharp break from tradition in emotion studies) to analyze what can be learned about emotion from all four of Baars’s tabooed disciplines, particularly the arts and humanities. This chapter will be guided especially by what the arts and humanities can teach us, as well as Baars’s fourth taboo, psychotherapy. In my view, these disciplines more than any others can clarify just where emotion fits into the overall conscious life, especially with regard to its intentionality. The process of experiencing art is like the process of psychotherapy, but in reverse. In psychotherapy, we begin with a murky feeling that seems to defy understanding, and work to discover imagery and words that can facilitate seeing the complex intentionality that is connected with the felt sense. By doing so, we allow the self-organizational motivational state of the organism to reorient itself so as to correct some of the imbalances that are impeding its freedom of movement, and move forward to feel still further emotions that need to be felt; we are no longer “stuck” in the one emotion. In this way, people are enabled to digest toxic emotions and their repetitive activation.

 Curious Emotions

In art, the process is just the reverse, but with a similar outcome: We are presented with readymade imagery by the artist, but the imagery is emotionally ambiguous and has been especially contrived so that it lends itself to use by the viewer to open up the unfolding of the progression of a variety of possible emotions so that we can explore them, understand them, and be guided into deeper and deeper levels of their meanings. By allowing emotional ambiguity with regard to what the images could mean, the artist gets us to give up Husserl’s “natural attitude,” in which we habitually try to establish a oneto-one correspondence between a static feeling and a simple environmental situation, and in its place we are encouraged to substitute the kind of exploration of the various interconnected meanings of the felt sense suggested by some of the examples we explored in the opening chapter of this book. To contrive such powerful imagery, the artist must understand how emotion and perception are related in a special way that is often missed by philosophers and scientists. There is much to be learned from what the artist knows and uses about this connection. What does it say about the organization of the psyche and the brain that – as every good artist knows – our emotional lives, including the priorities manifest in our patterns of curiosity, largely determine what we see, how we see it, and the patterns of searching and scanning used to look at a painting or a natural scene and interpret its literal and valuational meanings? Artists are experts at manipulating visual “affordances,” knowing with Helmholtz (1962) that the eyes must actively scan in order to see, and with Eleanor Gibson (1988) that the way a scene strikes us is determined by the patterns of activity it affords for the active eye and brain. Artists know that these movements are motivated by the motivational interests of the total living organism, and they use this knowledge in seducing us to see things in a new way. In knowing the difference between real art and mere “prettiness” or “entertainment,” they also know, with Panksepp (1998) and Rank (1924), that the emotions that structure perception go far beyond a simplistic stimulus-response hedonism. How can we put the artist’s working know-how into words and relate it to cognitive science in such a way that science learns from the artist, and not just the other way around? In Kantian terms, we could put the question this way: What are the neuroscientific and psychological conditions of possibility for the truth of what the artist knows and practices? What kind of description of the perceptual-emotional system would have to be true in order for what the artist knows how to do to work successfully? The three sections of this chapter will focus on three main conditions of possibility for the truth of the work-

Chapter 6. Learning about emotions through the arts 

ing knowledge of artists, whose conclusions can be preliminarily summarized as follows: 1. The total motivational purposes of our organisms lead us to see according to expectations and motivated interests that precede the presentation of a given stimulus. As Merleau-Ponty (1941) says, “We must look in order to see” (232). The emotions that direct attention are there before the moment of perception. To be precise, the ERP data discussed in the previous chapters (Aurell 1984, 1989; Srebro 1985; McHugh & Bahill 1985) show that the motivation to direct attention must begin about 1/3 second before the 300P parietal potential – i.e., 1/3 second before an occipitally-processed datum can be consciously attended to; if the datum does not appear as if it might speak to an organismic need or purpose, at least in some general way, then it does not enter into consciousness. General curiosity is one of the emotions with which nature has equipped us for this task (Panksepp 1998: Chapter 8), but it is superseded by stronger interests, such as the interest in a moving peripheral object, or an object whose retinal image is rapidly expanding, as discussed in the previous chapter. More specific emotional purposes related to the organism’s history and social context also determine the way the reticular activating system filters percepts (a good overview of this area can be found in Faucher & Tappolet 2000), and the way the anterior cingulate and frontal lobes direct selective attention to them (Posner & Rothbart 1992, 2000; Bachmann 2000), even prior to their being processed by primary perceptual areas such as in the occipital lobe (Aurell 1989). In each of these cases, we must have some interest in a stimulus if it is to enter into consciousness; attention is required for perception, at least if what we mean by “perception” includes a sense of being consciously aware of the object (Mack & Rock 1998). In a traditional way of looking at the brain, perception was supposed to feed information into the brain, which in turn led to thoughts about the information, which then led to emotional assessment of the information and finally action. Many investigators of neuropsychology and neurology (e.g., Damasio et al. 2000; Posner 1990; Posner & Rothbart 1998, 2000; Luria 1980; Panksepp 1998; Pribram 1991) are now beginning to realize that the artists were right all along, and that the situation is just the reverse: Our emotions gear us up for action, and then we search and scan the environment for relevant perceptual cues, which become conscious to the extent that they resonate with image schemas (Johnson 1987; Newton

 Curious Emotions

1996; Varela et al. 1993) that were already in process of being developed in response to frontal-limbic emotional purposes. Artists use their understanding of this fact – that in conscious experience the response must precede the stimulus – both to make pictures that are “easy to look at” (that present easy affordances for the dance of the active eye), and also to surprise and shock us, to create conflict, and to lure the eye into unaccustomed ways of looking that can allow meaningful selfmotivated emotional processes to unfold. 2. Art reveals motivations beyond entertainment, decoration, didactic illustration, or pleasure. In experiencing art, we want a form of symbolization that can intensify our emotional experience, rather than merely reduce it (as in hedonistic drive reduction). This requires that, as we have already seen on physiological grounds, there must be an exploratory, excitatory or extropic tendency that is independent of rather than derivative from any hedonistic drive-reductive tendency – whose purpose is not to maximize pleasure, achieve homeostasis, or (to use White’s classic definition of drive) eliminate a “physiological deficit external to the nervous system” (White 1959: 298). We want not only to reduce our drives, which brings pleasure in the straightforward sense meant by hedonists, but also to intensify the degree of consciousness we experience – to fully feel the value of that which we value, intensifying that feeling as a goal independent of whether the experience has a positive or negative hedonic valence (although we also prefer the positive valences, and set some upper limits to the intensification of unmet desires). Exploring this non-reductive aspect of motivation, which Rank (1924/1993) tried to conceptualize in terms of a “life force,” and which dynamical systems theorists (e.g., Kauffman 1993) thematize in terms of higher-energy attractors, is necessary if we are to understand why art speaks to us in a way that mere entertainment does not. This is why good art can disturb us, agitate us, or make us weep, and need not be “pretty” or “pleasant.” To know this, as sophisticated artists do know it in pragmatic terms, is to know that the intentionality of emotional experience is much more complex than most current theories acknowledge, that the object of perceptual or imaginative attention during the having of an emotion is often only a trigger, not what the emotion’s aims are really “about.” By concretely symbolizing an emotion, we explore far-reaching meanings that go beyond stimulus-response theories that would try to exhaust the meaning of the emotion by nailing it to the stimulus or type of stimulus that triggers the emotional “response.”

Chapter 6. Learning about emotions through the arts

3. Art can “inspire” us without giving us pleasure in net terms. We can sometimes find a uniquely important type of meaning by enduring or suffering through the misery, torment, and troubling disquietude of certain works, especially those falling under the category of “tragedy,” but also through other dark works such as horror stories, “film noir” mystery stories, “ironies,” and ugly, disorienting paintings and music. These painful artistic experiences are not merely an alternative means toward the end of pleasure, entertainment, or decoration, but offer their own special type of symbolization-matrix for the exploration of more “existential” emotions – emotions involving the affirmation of the values we can feel, rather than the attainment of the ends posited by those valuations. This effect shows further that the emotional life is not confined to means toward the end of any simple kind of pleasure or drive-reduction. Each of these points has important implications for the psychology and neurophysiology of perception and emotions and for the theory of consciousness. Section 1 of this chapter will pursue further the implications of the enactive view for the way perception is affected by emotion. Section 2 will then explore what art can teach us about the nature of the emotions. Section 3 will extend this analysis to the realm of existential issues, which involve the feeling of “inspiration” we discussed in Chapters 3 and 4.

.

An enactive dance form for the eye

In viewing a painting of a seacoast city featuring gabled houses with bay windows on hillsides, I find my eye pulled in by the painting. My attention is “attracted” to it, and I find the scene very “easy to look at.” Gibson (1988) would say that the scene affords a natural and gratifying pattern of eye movements. I can look at it for a long time, and the eyes do not become bored – since each eye movement naturally invites, welcomes, and harmonizes with the next – nor do they become too stressed by conflicting demands. Varela et al. (1993) would say that the organism “enacts” its own spontaneously self-organizing patterns, and the painting further enables certain self-enacted patterns of eye-movements, which are motivated by the organism’s emotional purposes (including curiosity). Paintings, rather than causing us to see and feel certain ways, only provide us with an opportunity to do so. More generally, the painting presents the eye with the opportunity to enact a dance that is felt as emotionally meaningful in



 Curious Emotions

a particular way that no other painting can accomplish. How is an artist able to make us feel this way? Yarbus (1967) and Pribram (1980, 1991) have shown on neurophysiological grounds that the eyes continually dance, with thousands of micromovements per second, and that without this active, self-generated movement, the eye could not see. These movements of the eye are not caused by the perceptual object, but by the organism’s own self-directed activity in the interest of questioning, scanning and testing the environment for clues that can be put together into a coherent interpretation of what is there. Contrary to popular opinion, even a single eye has depth perception (Pribram 1991: Chapter 4), because it enables the brain to compare the way an object looked from a certain angle at one moment with the way it looks from a slightly different angle a few milliseconds later, taking account of the efference copy, i.e., the registering of feedback from proprioception of the eye’s self-generated efferent activity (Jeannerod 1994) as it continues its tireless dance. The eye continually engages in elaborate self-generated activity designed to enable it to compute mathematically what the environment is like. These movements are motivated not only by general curiosity as to what is there – which is facilitative of our survival – but also by value priorities that make us more interested in certain specific kinds of objects than others – and this also facilitates survival. A newborn infant is already motivated to look for something that affords sucking, and when it finds the mother’s breast, it thenceforth understands the breast as the kind of object that presents this kind of affordance. There is now also good evidence that infants are preprogrammed to look for something that resembles a smiling human face (Meltzoff & Gopnik 1993; Barresi & Moore 1996). When it is found, their attention is pulled to it in the same way that a salient object on a stage can keep pulling our attention, even to the point of distracting us from the other events that are happening on the stage. A good theatre director knows how to use this fact to elicit the desired perceptual effects from the audience, just as a soccer player knows that a deceptive leg movement can lead the opponent to look for the ball where it is not, and thus not see it where it is. Seeing is conditioned by expectations, and to execute a Gestalt shift is to purposely change the pattern of our expectations so that it will be fulfilled by a different pattern of eye movements, or in other words to execute a different planned action schema (Newton 1996). In focusing on a Neckar cube, we know how to make it shift from one Gestalt to the opposite by focusing on just the appropriate spot on the drawing. The expectations dictate the pattern of the eye movements, which in turn determine whether the object is “seen as” this or that (Thomas 1989).

Chapter 6. Learning about emotions through the arts 

This point already has extremely important neurophysiological consequences for the way imagery functions in the brain. We saw in earlier chapters how well the Mack and Rock (1998) perceptual experiments confirm that we cannot consciously perceive something unless we first pay attention to it. When attention is fully occupied with one intentional object, another object will not register in awareness (although minimal information from it may be available to us on a blindsight basis, without perceptual consciousness); this remains true even if the unattended object is presented in the same area of the visual field as the visual task that is occupying our attention. (An exception to this rule, according to Mack and Rock, is that a large enough moving object can attract our attention away from a stationary one to which we are trying to pay attention. In motivational terms, the organism’s interest in moving objects obviously takes priority over just about any other environmental elements, both because of the possible threat to boundary integrity, and because of a moving object’s possible exploratory interest.) More generally, we must already be “looking for” general categories of objects before we will see instances of them. We have seen that perceptual experiments bear out this point. When the occipital lobe is activated by incoming visual data, there is no perceptual consciousness of the object until the parietal and frontal lobes are active (Bachmann 2000; Farah 1989; Luria 1980; Posner 1990; Posner & Rothbart 2000); yet the activation of the parietal and frontal lobes is not caused by the activity of the occipital lobe (Aurell 1984, 1989), but rather by action affordances sought by the emotional brain (Lethin 2002, 2004). This way of looking at the neurophysiological substrates of perception is already in conflict with certain more traditional ways of thinking of the causal ordering of neural pathways. As Goleman (1994) summarizes, “The conventional view in neuroscience had been that the eye, ear, and other sensory organs transmit signals to the thalamus, and from there to sensory processing areas of the neocortex, where the signals are put together into objects as we perceive them. The signals are sorted for meanings so that the brain recognizes what each object is and what its presence means. From the neocortex, the old theory held, the signals are sent to the limbic brain, and from there the appropriate response radiates out through the brain and the rest of the body” (Goleman 1994: 18). The findings we have cited by Aurell, Posner, Damasio, and Mack and Rock show that just the reverse is the case. The organism must first be geared up to look for data that are important for its purposes. This activates the frontal and limbic regions to begin “looking for” important categories of items, in

 Curious Emotions

effect forming vague sensorimotor imagery associated with the action affordances of those items, prior to presentation of the stimulus. If the presented stimulus then resonates with this self-generated activity, a more vivid sensory image of the object is formed in consciousness, and one that is felt as the image of a present perceptual object rather than as a mere mental image (Aurell 1984, 1989). For example, as we track the movement of a soccer ball, our expectation as to where we should look for the ball is at each moment motivated by categories of utility combined with the retention of the ball’s previous location. When the ball suddenly turns up where we are not looking for it, we do not see it, although we have a quite vivid mental image of the ball where it should have been. It then takes us at least a third of a second to find the ball again, by which time the opposing players may well have outmaneuvered us. The motivation to attend precedes perceptual processing – by a third of a second, in the case of a sudden novel stimulus. The object catches our attention because it presents a meaningful affordance for the overall motivated pattern of activity of the organism, which is already ongoing. In some instances, the emotional motivations not only pre-exist the presentation of the stimulus, but are also activated in a primitive, preperceptual way by the stimulus itself, through direct contact with the emotional brain prior to perceptual processing. As Gregory (1970) points out, “Whether the hand is burned by a match, a soldering-iron or boiling water makes little difference – it is rapidly withdrawn in any case. What matters is the burning heat, and this is directly monitored. The nature of the object may be established afterwards” (p. 12). Similarly, in visual perception, we are preprogrammed to treat certain kinds of items as having attentive priority – moving objects, objects whose Gestalt makes them seem novel, unusual or unexpected, and especially objects whose retinal image suddenly increases in size. This direction in the philosophy of mind is consistent with the notion that the most basic categories are purposeful, subject-related ones involving value, meaning, goal, effort, and action. These subject-related categories enjoy a certain priority over incoming sensory information. The categories of an organism’s consciousness are determined, not just by input from the environment, but also by the purposes of the organism. Judgments are then a function of the structure of subjects. For example, every purposeful being splits reality into obstacles versus helping factors, and this distinction grounds the categories of value and meaning. We are already in process of acting before we perceive, and in acting we generate “ideal possibilities.” The fact that I am an eating being implies the ideal possibility of something edible. The category system in-

Chapter 6. Learning about emotions through the arts

cludes a distinction between that which “affords sucking” and that which does not, prior to the presentation of any suckable stimulus. This implies, ultimately, that we perceive not only through our senses, but also through our emotions. Categories are based on structures of lived experience. This “enactive” approach to perception is also consistent with the fact that artists must effortfully train themselves to be able to “see what is there” (Edwards 1979), rather than having their perception almost completely dominated by a preconceived categorization of what is there. It is because our natural tendency is to see ideas rather than the specifics of their instantiations, that a child’s drawing of a hand might just as well have eight fingers as five, with no thumb, fingernails, or knuckles. As Edwards shows so clearly, seeing the preconceived category is the more natural tendency, which must be overcome if the artist is to learn to draw accurately, regardless of what culture the artist has grown up in. This point applies both to physical seeing, as discussed above, and to understanding, which is always threatened by the distortions of self-deception and prejudice. A typical technique that aspiring artists use to overcome this tendency is to draw the content of a photograph, first while viewing the photograph rightside up, and then while viewing it upside down. Invariably, the second drawing comes out more accurate, until the seasoned artist finally learns to ignore the distorting element in the rightside up image – namely, her own conceptual categorizations of the objects (Edwards 1979). There is some evidence that this process would involve using the corpus callosum in such a way as to prevent the dominant hemisphere, which organizes information in terms of abstract categories, from completely dominating the right hemisphere’s ability to organize it in terms of actual raw imagery, as it appears before we categorize it (Levy et al. 1972). In the same way, the artist must learn to allow the ventral stream of neural activation, which tends to look merely for instantiations of categories of utility, to be subordinated to the dorsal stream, which is more sensitive to the details of incoming data (Tucker 1981). Because all perception is motivated by categories of utility (Newton 1996), and because both the left and right frontal and parietal areas must be activated by lower, emotional brain areas (anterior cingulate, hypothalamus, hippocampus, and amygdala), we first tend to see abstract concepts, not the specific details of what is there. We see the sinister nature of a smile without noticing the details that tell us it is sinister (Wertz 1987); we notice the disorder in a room without noticing that a particular crooked picture frame is what makes it look disorderly (Merleau-Ponty 1942: 173). And when we have projected our own despised attributes onto “scapegoats,” we see the scapegoats as uglier and



 Curious Emotions

more despicable that they really are (Needleman 1968). So to think that we first see, and then develop abstract concepts out of our perceptions, is to place the cart before the horse. By showing the differences and the real relationship between concepts and percepts – the ways in which the former tend to prefigure and shape the latter – artists enable us to get beyond our preconceptual categories and to discover, if only to a limited extent, what it is like to see the world “as if for the first time.” When we see the way our categories contaminate what we see, we are also forced to consider what the object might look like if the categories had not prejudiced us to see it as we did. This produces an emotional as well as perceptual effect: Suppose every husband could, at every moment, see his wife as if for the first time? Suppose a political ideologue could see her favorite scapegoat in that way, and see the ways in which her own categories have distorted her perception of the scapegoat? Returning to the painting of the seaside city, with its gables and bay windows, suppose we grant that the affordance it presents is one that allows my eyes to dance in a way that I experience as emotionally meaningful, and that its ability to attract and sustain my attention is attributable to the emotional motivations to which those affordances speak. The question still seems very perplexing as to why this particular set of visual affordances is felt as such an emotionally meaningful one. To address this question, we should pay some attention to what makes artistic experience more than mere entertainment and pretty decoration. But this in turn will challenge some widespread assumptions about the nature of the emotions in general.

. Why does art move, and not just entertain? In light of the above analysis, to say that art “moves us” is slightly misleading. It might be more accurate to say that it presents us with a favorable opportunity to allow ourselves to move. Here the difference again revolves around action versus mere reaction. This would also be consistent with the fact that each viewer will get something different out of the same painting, and the same viewer will get different things at different times. The painting is not so much a bearer of a meaning, which has “one correct interpretation,” as it is a matrix of symbolization possibilities, presenting us with tools to use in intensifying, explicating and carrying forward our own emotional lives. Theodore Reik (1945), in his theory of romantic love, emphasizes that a person will not fall in love unless she is in a peculiar condition of “readiness”

Chapter 6. Learning about emotions through the arts 

to do so. Conversely, if a person is in this condition, Reik says, falling in love becomes much more likely for that person. This is why Shakespeare dwells so much on Romeo’s initial infatuation with Rosaline. What does this infatuation have to do with the story of Romeo and Juliet? It presents Romeo as being in the condition of readiness for love, so that when the opportunity for real love does present itself to him, we the spectators are also prepared to empathize more fully with the experience of falling in love, since this condition of readiness is a part of the experience (Branden 1988). Something analogous to this “condition of readiness” is required of the viewer of a work of art, if emotional meaning is to arise from the experience of the work. A good part of the resulting emotional event must be brought to the experience by the viewer. The viewer must know how to use that work as a vehicle to explore that emotional meaning in herself. This means, in part, that the viewer must want to have the kind of experience that the work is capable of yielding for that viewer at that time. Otherwise, the viewer will turn her attention to a different work, one that offers the emotional affordances that can be meaningful for her at that time. In the experience of an emotional response to a work of art, we can see that the relationship between the work and the feeling it evokes is not a direct causal one. Just as with the dance of the eyes, the dance of the emotional life is already ongoing, and the work must speak to that. The organism is already trying to maintain a dynamical, self-organizing pattern of activity, at a certain level of energy or intensity, and is already in process of dealing with thousands of emotional issues that pertain to that goal – some very mundane and pedestrian, others important at an existential level (Gendlin 1980). What does a good artistic image “afford” the emotional life of the viewer? We have seen that to say that it causes pleasure would fail to make the distinction, recognized by every serious artist, between real art on the one hand, and mere entertainment or decoration on the other (this distinction is pursued more elaborately in Ellis 1981). If the work of art does not have as its only purpose to cause pleasure, and if, as we have seen, it would be an oversimplification to see the artwork as a stimulus that causes a response, then we must say that the work on the contrary offers an affordance for a project of emotional exploration that is already underway. What is this project? We can infer from the above that the relation of a work to the feeling it evokes must be similar to the relation between what Gendlin (1980, 1998), in his “focusing” method of psychotherapy, calls a “felt sense” and its “handle.” The “handle” is an image or symbolization that can be used to “pull up” a felt sense, in the way that typing the code name of a computer program can “pull

 Curious Emotions

up” the program. I.e., entertaining a particular image may serve to enable me to intensify a feeling in myself that was only vaguely and dimly present before. For example, the way I feel when under impossibly stressful life demands might be intensified by thinking of the image of the Dutch boy holding his finger in a dike, and then realizing that still another hole has sprung in the dike, which he can barely reach with another finger. A person with alexithymia is someone who does not know how to find this kind of imagistic or symbolic “handle” (Gendlin’s 1980 term) for the bodily feelings of an emotional state, and therefore can be characterized as “lacking knowledge” of his or her own emotional states. This lack of knowledge is not merely a matter of lacking a label for communication purposes, but has important psychiatric consequences. The person is unable to understand, deal with, or work through his or her emotions because of the inability to use a “handle” to work with them, to get them to change and move forward. The person cannot pull the emotion “into focus,” in terms of a detailed awareness of what its “felt sense” subjectively “is like.” What does finding a good “handle” with which to intensify and “pull into focus” a felt sense do for the focuser? In the example of the Dutch boy’s symbolizing the way I feel about overwhelming life pressures, if I ask myself what it is about the whole situation I am facing that makes me feel that way (perhaps even an inevitable existential situation), I may then feel that what really bothers me is the statue-like immobility of the boy in the image, that he is not free to move, and is not available for any creative activity. At this point, the issue is no longer whether the Dutch boy will be able to hold back the sea, but rather whether being immobilized in that statue-like pose is an acceptable way of being. At this point, something may dramatically shift inside the focuser, who now sees the entire situation in a completely new way. The “stuck” place in the emotional progression has become unstuck, and the focuser can move forward in a direction that is naturally motivated and will be felt as meaningful. Because I am no longer “stuck,” I feel more fully alive, and am freed to feel things I would not have felt before. My emotional being has been brought back to life, and I am now available for it in a way that I was not before. If the mental image in my own mind of the Dutch boy and the dike can accomplish this purpose, then certainly a carefully crafted work of art can accomplish it. Suppose I want to remember a general mood that I tended to be in during the Fall of 1970. I know how to “pull up” this emotional meaning in myself. I can do it by listening to Tchaikovsky’s Fifth Symphony, which because of the way it is structured, especially in its harmonic dimension, allows those kinds of sad feelings to be explored. Whatever I am motivated to feel, I will feel it

Chapter 6. Learning about emotions through the arts 

more intensely if I can give some kind of symbolic expression to it, which allows a more concretely embodied form of activity that makes the feeling more extreme. Because of the music, the body becomes fully engaged in executing the emotion, and it is intensified and more sharply focused. If, on the other hand, I want to pull up the sense of a murky feeling that I do not understand, I do not have readymade imagery that is guaranteed to work. This is where the expertise of the artist must come into play. The artist provides complexes of imagistic material that afford use for a wide variety of focusing purposes. When we look at Degas’s The Absinthe Drinkers, a study of a somewhat disengaged couple sitting at a table in a pub, we have imagistic data with which to work in intensifying and carrying forward the intricate complexes of emotional meanings that come to life as we explore the painting. There is no stereotyped emotional meaning that the painting “depicts” or even is meant to “evoke,” but rather raw material for an emotional life with certain common threads that we all share because we are all human, we all value relationships, and we have feelings about them. It thus comes as no surprise that other artists, for example Picasso and Lautrec, have taken up the same subject – a couple who seem to be ignoring each other – and explored its meaning in still further directions, giving us still further tools with which to explicate the emotional meanings involved with this imagery. It is beyond our scope here to discuss differences between good and great art, but that could be done within this same framework. It might be argued that great art lends itself to addressing really important emotional issues that are inevitably there for all of us – we might call them “existential” issues – such as the brutality of the fact that we are finite, or the yearning of alienation that we all must experience because we cannot exist except interpersonally, and need certain kinds of authentic relationships in order to be who we are. Even in the most abstract art, issues of conflict, finitude, attainable and unattainable bliss, shock, positive valuation, and despair arise. A Kandinsky and a Baselitz afford very different moods just as surely as do a major key and a minor key in music. I have discussed these issues in some detail elsewhere (Ellis 1981, 1986, 1996b). This effect depends on our directly experiencing rather than simply intellectually acknowledging the felt senses involved. The artwork must invite us to really make contact with the emotional meanings to which it speaks, rather than just thinking about them. There is a difference between being aware that something has a certain meaning and directly experiencing this meaning. We may have it from reliable sources that the city of Toledo is beautiful, but knowing this is a completely different experience from actually perceiving the beauty of the city directly by looking at it or at a painting of it. Similarly, to know that

 Curious Emotions

a being or experience is unique and fleeting, thus irreplaceable and unduplicated, is different from directly experiencing its uniqueness and fleetingness. Art makes possible the direct experience of intrinsic values associated with the concrete existence of ontologically embattled, irreplaceable conscious beings, and then magnifies the intensity of this direct experience of intrinsic value by shocking, disturbing, disquieting. It must pull disequilibrium to at least enough of an extent to motivate us to see in a new way, rather than being content to reinforce our complacent prejudices, as a pretty picture of a cute little kitten might do. Such an analysis makes sense only against a background of motivational theory in which consummatory happiness and pleasure are not the only ultimate driving forces, even in an indirect way; drive-reduction (in White’s sense) is not the only end of human being. Rather, the end is to exist as the form of being that one is motivated to be, and to engage in the kinds of symbolizing functions and actions necessary for this purpose. There is something basic about the structure of intensely conscious beings (such as human beings) that allows us to feel the pull of such a powerfully transforming value experience as is offered by the challenging and disquieting imageries of the kind of art that is more than pretty decoration. Suppose we begin with the assumption that the first desire of a self-organizing conscious being is to be in the form of a self-organizing conscious being – to be conscious at a certain level of intensity. This desire is not fulfilled merely by taking steps to reduce drives, or even to feel pleasure. It seeks not the reduction of emotional consciousness, but a heightening, a carrying forward of one conscious state into the next in such a way that the overall progression remains interesting, alive, and meaningful. This entails what we might call “value-expressive” as opposed to “drive reductive” activity. I.e., we act not just in order to attain the end which our value feeling posits as desirable, but also in order more fully to feel the value feeling itself and to affirm that which we value. As Freud (1959) finally recognized, if energy-efficient drive reduction were the only ultimate motive, then life would degenerate into a monotonous homeostasis. Living beings want not only to reduce their emotional consciousness through drive reduction, but also to maintain and enhance their patterns of life and consciousness. This existential point can also be thematized neurophysiologically by thinking in terms of self-organizing dynamical systems that tend to maintain their pattern at a fairly high degree of energy and complexity – that tend to be drawn into higher-energy attractors as well as lower energy ones (Freeman 1975; Globus 1992; Mac Cormack & Stamenov 1996).

Chapter 6. Learning about emotions through the arts

Philosophy and science are only in the infancy of understanding how it is that the self-maintaining processes of living organisms are able to appropriate, replace, and even reproduce the physical substratum elements needed to continue and elaborate the patterns of these life processes or, as MerleauPonty (1942) calls them, “psychophysical forms.” But the data of the life world are there: If given a choice whether to be a very happy dog or a somewhat less happy human being, most of us on our better days would choose the less happy human being – implying that there is something worthwhile other than capacity for happiness, and which humans have more of than dogs have. Art does not serve merely as a means to increase our happiness, or our success in attaining the things we value in life, but rather as a vehicle through which to express our commitment to those values.

. Love and other non-consummatory motivations Much of the higher emotional life derives from motives that orient themselves not only around achieving the ends that we feel are valuable, but also around intensifying our experience of the value of the ends – the “inspirational” effect that was so basic in the system of emotional categories sketched out in Chapters 3 and 4. Many forms of love, including the love of a mother, of an erotic partner, or sometimes of a comrade in arms, are of the “value-intensifying” rather than the “end-attaining” variety (Ellis 1996b). I.e., they are motivated to intensify our experience of the value of a posited goal or object, rather than merely to take steps to realize the goal. Each of these kinds of love also resists reduction to purely “consummatory” motivations in the sense we discussed in Chapters 3 and 4. A mother will endure considerable suffering for the sake of children, as will (quite notoriously) a person in erotic love. Literature and the arts are quick to pick up on this non-consummatory dimension of love. It is an interesting phenomenon from the artist’s viewpoint, because the very non-consummatory nature of the subject affords exploration of the emotional life in which the consummatory motives are not assumed to be paramount, and neither is the consummatory-pleasure inducing or entertainment dimension of the aesthetic experience presumed to be the overriding artistic consideration. The classic instance of an artwork that seems to eschew the dominant importance of the consummatory dimension is the entire genre, not only in literature and drama, but also in music and visual art, that falls under the category of “tragedy.” Philosophers since the time of Aristotle have tried to explain



 Curious Emotions

what is sometimes called the “tragic paradox” – the fact that in watching a tragedy we positively want to experience the gut-wrenching feelings of grief and pain that we hope will be so intense as to move us to tears. A similar phenomenon also extends beyond drama to real life: We want to look at the picture of a deceased loved one so that we can grieve, to revisit the park where we spent time with a lost love, so that we can weep, to get the pianist to play the sad song again so that we can again reflect on how much sadness inevitably pervades life. Some might wish merely to write off such tendencies as symptoms of “clinical depression,” but labeling a phenomenon does not ensure that we have understood it. This tragic paradox, the desire for a painful catharsis in tragedy or in life, would not seem so contradictory if it were not for the usual assumption that humans are motivated always and only to maximize happiness and minimize suffering, or at least that we are willing to endure suffering only if it leads to some greater happiness in the long run. It is not so paradoxical that we can sometimes “enjoy” painful experiences in an equivocal sense of “enjoy” where “enjoy” is defined to mean simply that we sometimes choose to undergo painful experiences, as when a guilty person wants to punish herself. The real paradox arises from the conflict between this wanting to have painful experiences and the usual, everyday assumption that human motivation is fundamentally hedonistic. Aristotle tried to resolve the paradox without rejecting the hedonistic assumption by saying essentially that we are willing to pay the price of suffering if through it we can achieve a greater gain by learning something, since the purpose of drama is learning. But Aristotle’s account leaves us somewhat in the dark as to what kind of learning it is that could not have been packaged in a pleasantly entertaining way, rather than uniquely requiring us to suffer. Why is it that, after watching a videotape of Dustin Hoffman’s brilliant portrayal of Willy Loman, I watch it again the next night so that I can weep some more? What do I learn with the second viewing that could not be learned with the first? Others have tried to resolve the paradox by saying, with Clive Bell, that the vicarious suffering in tragedy is not a real suffering, and thus is not so bad. (See also Edith Stein 1989; Max Scheler 1970; and Husserl 1913 on this point.) But this resolution by itself would still leave unanswered the question as to why vicarious suffering serves a purpose that vicarious pleasure could not serve just as well or better. Hume suggests that unpleasantries are required to lend interest to a story, which would become boring if only happy events occurred; tension is required

Chapter 6. Learning about emotions through the arts 

for release in a good entertainment. But the problem is that a good comedy can be every bit as entertaining and interesting as a good tragedy; and as far as tension and suspense are concerned, a well-made mystery or action movie may well contain just as much of these qualities as a good tragedy, yet given a choice we sometimes prefer to watch the tragedy. What does tragedy offer that neither non-tragic forms of entertainment nor entertainingly packaged learning experiences (such as good philosophy lectures) can offer? We can make progress in making sense of this problem by combining some phenomenological observations made by Unamuno, in The Tragic Sense of Life, and by Levinas in Totality and Infinity. Levinas emphasizes that love, the appreciation of the value of another being through an intense experience of empathy, is often pulled by a vision of the other’s vulnerability and finitude. “Eros aims at the other in his frailty [faiblesse]” (256). Similarly, Unamuno stresses that the compassion dimension of love is largely determinative of its force, because we appreciate the value of something more fully when we focus on its inevitable finitude, thus contrasting its uniqueness and irreplaceability with the tragic possibility of its non-being. Through such tragic experiences, we fully appreciate the value of being, as instantiated in a love object or literary protagonist for whom we feel the intense combination of admiration and compassion; such an experience is different from mere pleasure or entertainment. On this interpretation, the unique value of the tragic experience as such makes sense if we reject the assumption that the only motive of human beings is to maximize happiness and pleasure. As we have seen, such an assumption ignores another dimension of personality, the dimension that affirms the value of its object independently of whether the intensification of this value leads to achieving pleasure-maximizing results. Most of the people in the history of humanity, and a sizable number of Third World people today, would probably agree that the amount of travail and hardship involved in getting through life is much too great to be vindicated in a hedonic calculus by the limited amount of pleasant experience that we might entertain any reasonable hope of contriving to add to the other side of the ledger, even if we are very lucky. This is one reason most people in most cultures have believed in some form of religion – because it offers an experience of the value of being, not by removing or hedonically counterbalancing life’s suffering and woe, but by making us feel that life’s value is positive in spite of any possible amount of suffering and woe (or at least in spite of very substantial amounts). Nor can it be a coincidence on this score that tragedy evolved from a religious ritual. Most of us in Western urban-industrial economic systems are lucky, comparatively speaking, yet it is a commonplace that the pleasures we are able to

 Curious Emotions

grab, by themselves, are “not enough” to give life meaning – to give it enough value to vindicate its sorrows and hardships. Indeed, this must certainly be one of the messages of tragedy: Try though we may, it is hubris to think that we can ultimately win any battle in which the score is kept through a hedonic calculus alone. To be sure, tragedy in the end should leave us inspired and uplifted, not downcast; but we are uplifted not because we now are convinced that we can win the struggle for a positive hedonic balance, but because we no longer feel that the value of human existence is contingent on the status of any hedonic calculus. This is not to deny, of course, that we value the pleasure of achieving the ends of love and of consummatory well-being; but the intensification of the experience of their value is also important independently of whether it leads to the attainment of those ends. If we explore the phenomenology of this experience of the non-consummatory intrinsic value of being – the valuing of life supported in all the prototype emotions that motivated complex and energetic life activity not geared toward consummatory satiation – we can then see why tragedy, and more generally works of art that involve a tragic dimension, are particularly suited to make possible such an experience. I have already mentioned that some authors minimize the suffering in tragedy because it is only a vicarious suffering. The fact remains, however, that the worse this vicarious suffering becomes, and the more the skillful tragedian can turn the knife in us once all hope is lost, the better the experience works – provided, that is, that the suffering is contrasted against an equally persuasive appreciation for the intrinsic value and importance of the sufferer as an instance of a type of being whose meaning is monumental enough to inspire gut-wrenching sorrow for the person’s ultimate defeat or destruction (as J. W. Krutch 1956 argues persuasively). By contrast to this celebratory aspect of the tragic experience, it is often noted that without an extreme depth of empathy and appreciation for the intrinsic value or “nobility” of the protagonist (which is lacking in merely depressing litanies of the misery and ultimate worthlessness of the human lot), the tragic effect fails to be achieved; thus we feel no catharsis, and we leave the theater frustrated and unsatisfied. At worst, the spectacle fails even to depress us, but simply loses our interest altogether. The role of empathy in tragedy can teach us a great deal about the nonconsummatory aspects of real-life human love. If tragedy is to have an inspirational effect – to intensify our feeling of the positive value of being, rather than merely its futility, we must experience this value as instantiated in the tragic protagonist through empathy. The tragic effect requires that we vicariously appreciate not the instrumental or extrinsic value of a being (as reflected in

Chapter 6. Learning about emotions through the arts 

capacity for great deeds, power, or social status), but rather the person’s intrinsic value. And to do this requires setting up a dynamic in which our empathic admiration, not for the person’s accomplishments or grandiosity, but rather for the person’s intrinsic value qua embattled yet existentially meaningful conscious being, must be intensified by our compassion, while at the same time our compassion is intensified even more by the feeling of empathic admiration for the person qua ontologically endangered yet intrinsically valuable. If we do admire a protagonist in this way, the feeling of admiration is sharply intensified by our compassion for the tragic flaw which will result in her downfall. These same elements of admiration (for the person’s intrinsic value), compassion, and empathy can be noticed in real life feelings of love. There is often an intensification of the feeling of love in the instant when we have reason to focus on the finitude of the love object – when we empathize with the embattledness of her existential project in the face of the forces that not only endanger her physical existence, but also threaten the prospect of authentic being of the person she has the potential to be. Moments when we sense that the other’s being is ultimately faced with conditions of endangerment (including social or existential endangerment) are moments when the concrete feeling of love is pulled very sharply. The intensification of admiration through compassion for the vulnerability and finitude of a tragic protagonist (or, in real life, a love object), transforms mere admiration into an instance of the kind of experience that can be designated as “awe-inspiring.” Admiration for a great hero such as Hector becomes an overwhelming awe when the threat of his heroic death becomes real, and more still because his cause is already doomed to lose; the sharpest possible intensification of this awe is achieved (assuming that we do initially admire and empathize with him) in the instant when he is run through by the blade of Achilles, who then drags his mutilated body through the street in front of his aggrieved widow. But our admiration for Achilles in turn is also intensified to the point of awe because we know that the Achilles heel makes him also vulnerable and finite. In general, our admiration for all that is worthwhile in the struggle of conscious beings against the inevitable ontological conditions that embattle them is magnified by our compassion for the inevitably tragic condition of finitude. Such an analysis makes sense only against a background of motivational theory in which pleasant satiation is not the ultimate driving force; drivereduction is not the end of human being. Rather, the end is to exist as the form of being that one is motivated to be, and to be in this pattern independently of the desire for pleasure and happiness (which of course is also there). In dy-

 Curious Emotions

namical systems terms, the system needs to maintain its structures of activity at a certain level of complexity just as much as it needs to reduce drives so as to achieve homeostasis. To ignore the emotions that are not satiation-driven is thus to ignore half of the emotional life. Tragedy, for example, does not serve primarily as a means to increase our happiness, or our success in attaining the things we value in life, but rather as a vehicle through which to express our commitment to those values. Not only in tragedy, but in all artistic activity, we sek to intensify value feelings rather than only to attain what they value, and we want the emotions to have an opportunity to unfold by finding vehicles of expression – Gendlin’s “handles.” By expressing values, as Gendlin (1998) and Langer (1957) point out, we give our feelings about them concretely embodied symbolization, and in this way we feel their value more fully; we thus also feel more intensely the value of being per se. We began this chapter with the observation that perception always offers an enactive affordance for the activity of the eyes (and brain); thus some scenes are “easy to look at,” and evoke different kinds of moods depending on what kind of enactive affordances they present for the eyes. Visual art sets up the possibility of a “dance form for the eyes.” Art can do this only because perception is active rather than passive, and begins with efferent activity which then motivates afferent processing. The limbic system “categories” that motivate looking-for are different from perceptual categories; they are broken down in terms of emotional affordances. Art plays with our looking-for, using it to make us engage in different afforded actions that relate to different limbic (emotional) categories. Children not only fail to “draw what they see” (drawing instead what they conceptualize that they ought to be seeing), but in fact they fail to see what their eyes see; they have to be especially trained not only to draw what their eyes see, but also to see it. Normally, we fail to even know what the actual perceptual appearance of the sinister smile was – only that it was sinister – unless we happen to be trained as artists, or are confronted with an effective painting of a sinister smile. The actions that art affords are not just reductively motivated, but are also motivated by self-actualization motives that can be understood in terms of selforganizing tendencies of the whole organism viewed as a dynamical system. We “use” art for the purpose of symbolizing our emotions, rather than just reacting to it emotionally. The extropic motives can be phenomenologically experienced in terms of the “value expressive” as opposed to the “drive reductive” tendencies. I.e., some of the actions in which we engage are meant to affirm the value of what we value, rather than only to attain whatever it is we value.

Chapter 6. Learning about emotions through the arts 

Understanding the meaning of art calls for a more complex theory of the meaning of emotional experience than has so far been attained. The object that is in conscious attention during the feeling of an emotion is not the whole story of the intentional object of the emotion, and often not even an important part of its object – i.e., it is not the object in relation to which our actions could serve the purpose of the emotion. A frustrated person who is insulted by another, thus triggering anger, is not caused to be either angry or frustrated by the insulting person, any more than a grieving person’s sadness is caused by a funeral dirge. In both cases, the objective triggering event presents certain affordances, which may or may not speak to the enaction needs of an already-ongoing emotional progression. The insult would not even have been so angering had it not been for the underlying holistic progression of frustration that contextualized the stimulus. The emotion is not triggered by the stimulus per se, but rather by the stimulus in a certain context. Emotions are aroused not by simple stimuli, but by their meaning for us in a total context determined by ongoing and dynamical organismic purposes. Emotions call for explication; they arise from the total life process, which is a dynamical system – not as a simple causal result of a simple stimulus. Art affords not only a meaningful self-directed dance of the eyes or ears, but also a meaningful dance of this self-motivated process of emotional exploration.

Chapter 7

Dynamical systems and emotional agency A closer look

In some sense, living organisms involve “agency” at a primitive level: they act to appropriate and regulate their own components, rather than being merely the random summation of the behavior of the components, in passive reaction to other components or to external inputs. If motivation entails more than just movement, but also involves complex self-movement and self-regulation of some sort, then the presence of motivation in an organism presupposes agency in a minimal sense. If a theory of emotion is to be grounded in a concept of dynamical action as distinguished from passive reaction, one of our foundational concerns must be to reconcile micro-level physical causation with the motivational determination of action. Sometimes this kind of determination is referred to as the phenomenon of “free will,” but that term can engender some philosophical confusion around the issue of determinism: the fact that organisms act out of a unitary determination of will rather than merely react to external inputs does not entail that there are physically uncaused events, or “contracausal freedom” in the philosophical sense. But the problem of agency in its most primitive sense is one of the major stumbling blocks for a self-organizational theory of emotion, especially if we hope to avoid a straightforward epiphenomenalism in which the apparent causal power of purposeful intentions would be only an illusion, as Wegner (2003) argues. This chapter will develop and defend a particular kind of dynamical systems account of the way motivational causation is rooted in the self-organizational purposes of the organism as a whole, in order to show how the whole can exert real causal power over the rearrangements of its own constituents. Any coffee drinker facing a public performance knows how a few strong cups can make the difference between a masterful performance and a lackluster or even depressing one. At the micro-level, we could explain this effect by saying that caffeine facilitates release of acetylcholine and norepinephrine (by interfering with inhibitory neuromodulatory systems), which together with the

 Curious Emotions

normal increase in cortical arousal associated with the perceived possible danger of failure cause the performer to become energized; but they do so only because the overall plan (the intention to drink the coffee) sets up the background conditions for these micro-level causal relations – in this case, internal causal relations that result from drinking the coffee. Without the coffee, the situation might even be depressing, which in turn could lower the performer’s energy level still further. The presence of the caffeine can determine the difference between two different patterns in which the organism can maintain its overall homeostatic balance: a downward spiral into lethargy and apathy, or an upward spiral into inspiration and higher energy. By choosing to drink the coffee, the performer does not change the physical principles that govern the micro-level causal relations within his own body. What he does is to rearrange the background conditions that are presupposed by these micro-relations. No causal relation is sufficient except under certain general background conditions. Both the micro-mechanical explanation, in terms of chemical reactions in the brain, and the intentional, self-organizational explanation, in which the organism’s overall pattern affects the details of the micro-mechanisms (in this case, by means of a conscious plan), must be regarded as valid, but in different respects. The trick (and the purpose of this chapter) is to integrate these two types of explanations into a coherent account of the micro-level and holistic explanations. This compatibility is needed if different disciplines in the study of emotion (neurophysiology, psychology, etc.), which study different levels of organization, are to be commensurable. We do not yet know, and may not soon know, all the micro-mechanisms that subserve conscious plans, but what we do know implies that there is some such combination of micro-mechanisms; and Physics 101 tells us that at the aggregate micro-level the causal antecedents of a system should be sufficient to determine its outcomes (even if at the sub-aggregate level of subatomic particles there should be indeterminacy). But at the same time, we know that our conscious choices – for example, the decision to drink the coffee prior to the performance – can rearrange the micro-constituents themselves by altering some of the background conditions presupposed by their causal interrelations. While in the example just mentioned it is a conscious decision that leads to the rearrangement of the background conditions for the needed causal mircorelations, the same thing happens on an unconscious basis all the time – for example, when the organism regulates its body weight by balancing hunger and availability of food with the amount of sleep and the desire to expend energy in various ways and at various intervals, or even when it regulates blood pressure, heart rate, or release of adrenaline in order to maintain its overall

Chapter 7. Dynamical systems and emotional agency

self-organizational balance. In all these cases, micro-level explanations can be given for each discrete event, but the fact that these events occur in the patterns they do is not just happenstance; it is somehow determined at a higher level of organization, by a tendency of the whole organism to maintain its definitive patterns and thus avoid disintegration. The bio-medical model of psychiatric treatment, as discussed extensively by Zachar (2000), attempts to change the functional pattern of the whole by manipulating a specific event in a micro-level causal sequence. Since the desired outcome is a holistic balance, sometimes this approach works, and other times it results merely in an adaptation of the organism, rearranging its other parts to compensate for the micro-level change, so that the organism as a whole can remain in its previous pattern. By contrast, when a depressed person decides to jog every day, she is rearranging the system holistically, in hopes that micro-level chemical reactions within the body will change. What is really needed, to make a theoretical place for genuine causal power on the part of purposeful intentions, is a type of dynamical system that is “topdown” in the philosophical sense; i.e., the higher level of organization must have certain powers not reducible to the lower, but without violating the causal closure of sequences of events at the lower level of organization. The higher level of organization sets up the background conditions on which the lower level events depend for their causal sufficiency; given those background conditions, every lower level event has a lower level cause that is sufficient for it, yet would not be sufficient for it had not the right background conditions been in place both within and outside of the organism’s boundaries. Ironically, this will later turn out to be tantamount to a “bottom-up” process in the neurophysiological sense: It is primarily the subcortical emotional areas that transmit the body’s overall self-organizational imbalances, and thus its aims, through the primitive mammalian ventral base of the cortex, to the higher parts of the cortex. No dynamical systems account of cognitive or cortical functioning can be adequate unless it includes a major and unique role for the emotional brain. It is the emotions that make the brain’s many functions serve the purposes of the body as a whole, not merely of the brain as if it could be a complete dynamical system in its own right. The brain is not a dynamical system in whose service the rest of the body is merely an appendage, but just the reverse: the body evolved first, and then the brain was added as an appendage to it. Selforganization must be understood at the level of the whole body, not just of the brain as if it were a complete dynamical system.



 Curious Emotions

An understanding of this dynamical process is crucial for setting the parameters for a theory of emotion. Emotions aim to act in the service of selforganizational purposes, are motivated by those purposes, and intentionally refer to environmental affordances related to them. Their physiological underpinnings are globally distributed and include rich systems of homeostatic feedback loops, massive overcausation, shunt mechanisms within dynamichomeostasis cycles, autocatalysis (where a terminal biproduct of a reaction catalyzes its earlier steps), and multiple realizability. Thus the aims, objects, and causes of emotions hinge on the process through which an organism can cause its own components to be appropriated and used to maintain a selforganizational structure across multiple replacements and purposeful readjustments of the components that subserve the pattern of the system’s activity. In some sense that needs to be carefully elucidated, the whole therefore has a certain kind of causal power over its own parts. To understand such a capability requires difficult theoretical work, because it must be reconciled with what is known about physical causation at the micro-level, where higher-order properties and processes usually seem to be epiphenomena – causally irrelevant surface properties – floating over the top of the causal powers of the micro-constituents. The micro-level causal sequences always seem to be sufficient to account for whatever happens, and thus are “causally closed” (Kim 1992, 1993, 1998). How then can the system appropriate, use, and reorganize its own components at will, as seems to occur when an emotional purpose, which is physiologically holistic, can cause actions that make use of the body’s own parts, both morphologically and at the level of chemical and electrochemical micro-events in the body’s nervous system? To work out the needed causal dynamics for a dynamical system with genuine causal powers that are not merely epiphenomena of its microconstituents – so that the causal relevance of motivations and emotions is not merely a subjective illusion masking the real causal powers at the microlevel – is a philosophical problem. The specifically psychological and physiological discussions that come later will depend crucially on this basic theoretical groundwork. We must carefully consider what is needed to reconcile the notion of self-causation in the sense of self-organization with the requirement for causal closure of the micro-level components of self-organizing systems. In this way, we can preserve the unity of the study of emotion and motivation at the levels of various disciplines all the way from inorganic and organic chemistry, which can make valuable contributions to understanding the workings of the emotional brain, to phenomenology, psychotherapy, and the humanities, which also have important methods and insights to contribute. In the end,

Chapter 7. Dynamical systems and emotional agency 

there must be no contradictions between the basic principles of any of these various disciplines.

.

The causal power of dynamical systems

Dynamical systems theories sometimes attempt to avoid strict psychophysical reduction, in the sense of reduction to micro-level causal mechanisms, but at the same time without violating the causal closure of physical events (Juarrero 1999; Prigogine 1996). Since dynamical systems seek out, appropriate, and replace physical substrata needed to continue their structural pattern, it is often proposed that the system can be relatively autonomous with respect to its components, yet the components can still constitute closed causal chains (Juarrero 1999; Kauffman 1993; Monod 1971). But how can systems have causal power over their substrates, if each component is sufficiently caused by other components? If a micro-level cause is sufficient, then by definition it should exhaust the explanatory domain, and render any other antecedents (for example, at the level of higher-order processes) unnecessary. The key to this question is to examine the way causal sequences presuppose background conditions for the causal relation. Suppose every causal relation requires background conditions, without which it is insufficient, as proposed by Mackie (1974). For example, flipping a switch causes a light to come on only if certain background conditions are in place – good bulb, wiring, etc. We can then say that a dynamical system is structured with a tendency to change background conditions for causal relations anytime needed substrates for the pattern’s own maintenance are missing; under the changed background conditions, alternative causal relations become sufficient to maintain the pattern. The system controls the background conditions under which one or another causal relation can subserve the system’s overall pattern, while the components remain causally closed under their given background conditions (Ellis 1986, 1992, 2000c, 2001a). In this case, a biological organism, with its multiple shunt mechanisms, homeostatic feedback loops, and maintenance of causally robust patterns of activity across replacements of their micro-constituents, is an excellent candidate for a system that can exercise real causal power over its own components. For example, when synapses are artificially flooded with serotonin, as with many recreational drugs, the brain automatically compensates by reducing the number and size of serotonin receptors, in an attempt to move the organism back toward the preferred holistic balance (Valenstein 1998: Chapter 5). In this

 Curious Emotions

case, the self-organizational process as a whole has altered a larger structure, the configuration of its receptors, to change the background conditions presupposed by the specific causal micro-mechanisms that lead to the euphoric response – the chemical reactions produced by the serotonin. By doing so, it maintains balance within larger homeostatic feedback loops, thus rendering the specific micro-mechanism (the specific serotonin reaction) less causally efficacious than it otherwise might have been. The organism thus can manipulate its own micro-components in the interest of producing a constant outcome across variations in the micro-level processes. It is true that the process takes this action by working through other micro-mechanisms whose effect is to reduce the size and number of the serotonin receptors. Micro-level causal explanations of these processes could also be cited; but the background conditions needed for those micro-mechanisms to work have also already been determined by the body’s tendency toward balance at a more holistic level. Ultimately, the body attempts wherever possible to readjust the background conditions for any micro-mechanism if this is what is needed to maintain the interest of balance at the larger level of organization. Unfortunately, certain recreational drugs such as “ecstacy,” which inputs directly to the micro-level mechanisms by flooding the synapses with serotonin and dopamine, can overwhelm the body’s attempts to compensate, trapping it in an irreversible change even after the excess serotonin and dopamine are withdrawn. The result often can be permanent or semi-permanent depression (Valenstein 1998: Chapters 4 and 5). What makes such drugs “artificial” is that they act directly at the level of micro-mechanisms, from the “bottomup” direction, and in this way often can act so strongly as to wash out the body’s capacity for “top-down” self-regulation, rather than allowing the selforganizational process level to regulate the level of the micro-mechanisms, as it normally would. I cite this example only to illustrate the practical applicability of a selforganizational hypothesis. The example does not prove the hypothesis, since it could possibly be explained in alternative ways. But what I intend to do here is to show how the organism as a whole process can have genuine causal power not only to regulate, but in fact to appropriate and manipulate the micromechanisms of which it is composed, vis a vis structures allowing for overcausation, autocatalytic loops, numerous shunt mechanisms, buffer mechanisms, and multiple realizers at the micro level, all of which tend in many instances to enable the system to control the background conditions presupposed by any discrete causal sequence at the micro level. This will mean that the organism’s motivations can arise endogenously at higher organizational levels, and can di-

Chapter 7. Dynamical systems and emotional agency 

rect genuine action as opposed to passive reaction. This is what makes agency – the ability of the organism to direct the use of its own parts, including and especially at the micro-level, rather than only being directed by them – different from a situation where the appearance of agency would be a mere illusion, since each part would really be moved only by micro-mechanisms below the organizational level at which motivations occur (as argued by Wegner 2003 and Dennett 2003). In addition to self-organizational structures complex enough to appropriate their own micro-constituents, all that is needed for genuine agency in its most primitive sense is that the organism not only regulate itself, but also act on its environment. Thus, in the sense of primitive agency intended here, animals are more “active” than plants, and animals that plan their locomotion and food-acquisition are more “active” than amoebae. Plants, by contrast, are primarily “reactive,” although they exhibit a minor degree of activity in moving toward light and water. So motivation and emotion, whether conscious or unconscious, can affect behavior only in organisms capable of acting on their environment in order to preserve self-organizational structure. Such an “agentic” structure can be linked to lower-level, merely mechanical causal sequences by means of a theory of complex dynamical systems. Our discussion of this problem is complicated by the fact that there are currently different notions of what kind of causal relations are involved in a dynamical system. Some dynamical systems approaches (for example, Kelso 1995) have no quarrel with the reduction of self-organizing systems to conglomerates of one-way, bottom-up causal sequences, so that the pattern of the self-organizing system itself can be viewed as merely an epiphenomenon of the way the micro-level constituents behave, which just happens to work out in such a way as to self-organize the system as a whole. Newton’s characterization of the property of self-organization in complex dynamical systems would be consistent with this reductionist viewpoint: The self-organizing properties of complex systems . . . incorporate a natural tendency toward order, which arises spontaneously among molecules when in sufficiently complex groups, in a way that can be explained entirely by physical mechanisms and involves no mystery. It does, however, allow the emergent order to be conceptually distinguished from the substratum in ways that appeal to some theorists. Mental states are not reducible to the individual states of the substratum, but they are physical states nonetheless, and obey physical laws. (Newton 2000: 91)

 Curious Emotions

But Newton also goes further and endorses a more radical thesis in selforganizational theory (for example, in agreement with Anderson & Mandell 1996; Ellis 2000a; Juarrero 1999; Kaufman 1993; Monod 1971; Weiss 1968) which holds that when a dynamical system is complex enough, not only is it multiply realizable with respect to its physical substrata, but it can also play an active causal role in seeking out, appropriating, replacing, and reproducing the substrata that are needed to maintain the organizational pattern of the system. There seems to be controversy as to whether the self-organizing system has real causal power independently of the causal powers of its separate micro-level constituents, and I shall return to this question. Dynamical systems theorists flesh out the notion of self-organization in terms of open thermodynamic systems, of which biological organisms are examples. Open thermodynamic systems continuously exchange constituent components and energy with their environment, yet maintain homeostatic constancies across these exchanges; these constancies preserve continuity of structural organization into the future. A behavior pattern into which the system has a strong tendency to settle, and in which it can find a good degree of holistic balance, is called an “attractor” or “basin of attraction.” The organism learns and remembers new perceptual patterns by creating new basins of attraction structurally related to the learned stimulus pattern (Alexander & Globus 1996; Freeman 1987, 1988). In Freeman’s work on olfactory learning in rabbits, as summarized by Alexander and Globus, “There are different basins of attraction, in the form of limit cycles, for different odors the rabbit can recognize. . . . Upon presentation of a novel odor, the olfactory bulb is pushed into chaotic activity [allowing] formation of a new limit cycle attractor to suit the novel stimulus” (Alexander & Globus 1996: 42). When the image or intentional state is later reproduced in the animal’s mental life, what the organism is doing, as Newton (2001) puts it, is to motivatedly re-enact a complex pattern of neurophysiological behavior corresponding to the basin of attraction. A stimulus can disturb the holistic balance of a system – which then is restored by settling into a different basin of attraction. For present purposes, we can think of a dynamical system as an open thermodynamic system that exchanges energy and materials with its environment while maintaining continuities of structure at a level of organization higher than the level of the substratum components that are continually being appropriated and replaced by the system, and is causally robust with regard to the various inputs into the system. For example, Kauffman (1993) argues that selforganization occurs spontaneously given a sufficient diversity and number of entities capable of a sufficiently large number of potential chemical reactions

Chapter 7. Dynamical systems and emotional agency 

and an autocatalytic structure. A self-organizing system that can catalyze the reactions that maintain its own existence is a “collectively auto-catalytic system.” This requires that a system not be too rigidly organized, but that it have different basins of attraction that it can shift into and out of depending on the need for self-maintenance. Part of the appeal of this notion for purposes of the problem of motivational causation is that it may offer resources with which to clarify the relationship between the causal power of the organizational pattern of a system on the one hand, and on the other hand the specific causal powers of specific components of the system (due essentially to the way in which it is multiply realizable). Monod (1971) and Kauffman (1993) specifically claim that the concept of a dynamical system with a structural tendency to maintain itself as a whole does not contradict the normal causal laws that constrain the discrete interactions of the components of the system, but rather supplements them with a different kind of analysis – a structural analysis – that is also needed for a complete explanation. If so, then such an approach to the phenomenon of self-organization could help to address the problem of motivational causation by attributing a top-down causal power to motivational intentional states, which could be identified with structural or relational properties of the system; these structural properties would have the power to maintain themselves across multiply realizable replacements of their own physical substrata. This claim is made only for very complex dynamical systems – complex enough for the pattern itself to seek out and replace the components needed to maintain the pattern. We can thus think of a dynamical system as one whose organization creates a strong tendency to maintain itself across various alternative causal mechanisms at the level of the components whose higher-order structural relations instantiate the system. In a complex dynamical system, not only is the structural pattern multiply realizable with respect to alternative sets of substrata, but it also plays an active role in bringing it about that one or another of the combinations of substrata needed to maintain the overall pattern will obtain. Kauffman and Monod define this dynamical relation in terms of both selfmaintaining and self-organizing systems. A self-organizing system not only has a strong tendency to maintain its pattern in the way just defined, but also has a strong tendency to come into being in the first place. All self-organizing systems are self-maintaining, but the converse does not necessarily hold. For purposes of applying the theory to the mind-body problem (for example, Edelman 1989, 1992; Freeman 1987; Staminov & MacCormack 1996; Thelen & Smith 1994; Thelen et al. 2001) we can assume that the dynamical system (the biologi-

 Curious Emotions

cal organism) already exists, and need worry only about how the system is self-maintaining across multiply realizable substrata, some available subset of which the system actively seeks out, appropriates, replaces, and reproduces. In speaking of “biological organisms,” there is no assumption that artificial systems could not manifest such structures, thus in principle that such systems could not have conscious minds. The determinative factor is not whether the system is composed of certain specific elements such as silicon or carbon, but rather the structural dynamics of the system. To be a biological organism, in this sense, is to be structured in a certain self-maintaining and self-organizing pattern, not merely to be composed of certain elements. Some of the theorists for whom complex dynamical systems can have causal power over their constituents explicitly contrast dynamical systems against simpler connectionist systems by virtue of emphasis on the top-down causal role of the higher-order pattern that not only maintains itself across replacements of its components, but seeks out and organizes usable components. Alexander and Globus are particularly straightforward about the implications of their view of dynamical systems for the causal analysis of events. “[In] connectionist models. . . transformations are carried out by individual neurons. . . by receiving weighted activations from other neurons” (Alexander & Globus 1996: 32). By contrast, in agreement with Freeman (1975), Globus (1992) and Nicolis (1986), Alexander and Globus characterize edge-of-chaos dynamical systems in this way: “[When] interconnectivity within a particular scale of organization reaches a critical limit, that scale of organization becomes a module in a larger scale of organization. . . [leading to] cascade effects whereby changes at one scale of organization can modify other scales in an exploding chain of reactions up and down the multi-scale structure” (Alexander & Globus 1996: 38). The higher level of organization is claimed by such theorists to “constrain” what can possibly occur at the lower level. Juarrero (1999) makes the same claim, arguing that complex systems can have causal powers not completely reducible to those of their micro-level components. How do these different “scalar” levels interrelate causally? A favorite analogy (Harth 1983; Alexander & Globus 1996) is the relationship between laminar and turbulent flow in fluids. “In turbulent mode the macro-scale turbulence is an expression of microscopic variations in the structure of the flow. In the stable flow, this between-scale communication is reversed. . . . ‘Laminar flow’ refers to fluid moving in an ordered fashion. . . ” (Alexander & Globus 1996: 42). In laminar flow, the overall structural pattern “constrains” the causal relations at the molecular level. The way global wave forms in the brain constrain the discrete interactions of their components is supposed to be similar:

Chapter 7. Dynamical systems and emotional agency 

“Freeman (1988) calls the highly stimulated state wave mode, indicating the dominance of the global wave-form. . . . The wave-form now causally constrains the activity of the individual neurons. The wave mode is analogous to the laminar phase in the flow metaphor. The global wave-form constrains the interactions of individual neurons” (Alexander & Globus 1996: 49–50). Juarrero (1999: 114ff.) uses the same analogy, reaching the same conclusion. But a problematic question is immediately suggested by this example: Both turbulent and laminar patterns describe causal interactions between water molecules, with each molecule’s behavior explainable in terms of the behavior of others. In principle, each molecule’s behavior can be completely explained in this way, without any reference to the overall pattern of flow; it thus seems that, if there is causal closure at the molecular level, the overall pattern can add nothing to the causal explanation of any molecule’s behavior that has not already been explained at the molecular level. If so, the causal power of the dynamical system, if not literally identified with the sum of the causal powers of its micro-level constituents, at least reduces to just another epiphenomenon of the sum of the behaviors of the constituents. So if such a model is applied to the mind-body relation – especially to the relation between purposeful intentions and their neurophysiological substrata – it entails just another form of epiphenomenalism, or alternatively a reduction by identification of the causal power of the dynamical system to the causal powers of its micro-level constituents. The mental decision to raise my hand still has no causal power beyond the causal powers of its separate physical components, which ultimately are single neurons, molecules, and electrons. Thus, if causal closure is not rejected, it would seem that the intentional decision to raise my hand can have causal power only if reduced to the causal powers of its separate physical substrata, which already exhaust the causal power of the system. But reducing the causal power of the purposeful decision to that of its individual physical substrata likewise does not help us understand in what real sense a unified agent or intentional state can have the power to rearrange its own constituents. It is important to note here that rejecting causal closure per se would also be an unhelpful move. To reject causal closure would not just mean rejecting the notion that every physical event has a physical cause; it would also mean denying that even those physical events that do have causes have physical causes. A mere epiphenomenon of the micro-level events is not the result that the top-down variety of dynamical systems advocates want, nor can it help anti-epiphenomenalists avoid the conclusion that the causal power of an intentional action choice is only illusory, since the agent’s intention itself is only

 Curious Emotions

a byproduct of micro-level sequences that exhaust the real causal power. What is needed is that the structural pattern of the system should make a causal difference to the maintenance of the system, independently of the existence of any particular components at the substratum level, as long as suitable components can be found when sought by the system. In philosophy of mind especially, the top-down type of dynamical systems theorist wants to be able to explain phenomena that presumably are not exhaustively explained at the substratum level; for example, such theorists want to deny that mental phenomena are exhaustively explained by connectionist systems in which transformations are carried out by individual neurons. What would dynamical systems theory have to do to make possible some sort of causal power for the organizational structure of a system that cannot be exhausted by its substrata, yet also does not violate physical causal closure?

. How can top-down systems avoid violating causal closure? Robert Francescotti (1998) points out that, when it comes to multiplyrealizable phenomena as are typical in brain functioning, there is seldom a discrete micro-level event that is necessary and sufficient for a given outcome. Instead, what is necessary and sufficient for the outcome can be a disjunction of several alternative mechanisms, any one of which can realize the outcome. And the causal antecedent for the outcome also is frequently realizable by several alternative micro-mechanisms. So, instead of a simple linear causal sequence, A if and only if B

what we have is a disjunction of antecedent mechanisms leading to a disjunction of consequents, (A1 or A2 or A3 ) if and only if (B1 or B2 or B3 )

In the case of the euphoric “high” from recreational drug use, the effect may be achieved either by flooding the brain with serotonin (A1 ) or by flooding it with dopamine (A2 ) or by inhibiting the reuptake of serotonin at the receptors (A3 ), or by a number of other micro-mechanisms. Similarly, the euphoric outcome could be facilitated by increased dopamine, or by stimulating any of several different hypothalamic centers facilitated by a variety of opiates. Of course, only one of the alternative antecedents and only one of the alternative consequents is likely to occur in any real situation, but we should not make the mistake of assuming that the one is necessary or sufficient for the

Chapter 7. Dynamical systems and emotional agency 

other. Instead, we should think of the alternative antecedent mechanisms, A1 , A2 , and A3 , as several different ways of realizing a certain complex physical relation, R1 . Similarly, the alternative consequents, B1 , B2 , B3 , are alternative ways to realize an outcome relation, R2 . We can then understand the causal relation between the input side and the outcome side of the equation in terms of multiply realizable physical relations: R1 if and only if R2

where R1 could be realized by A1, A2, or A3, and R2 could be realized by B1 , B2 , or B3 . Multiply realizable phenomena alone are not enough to create the more complex structures of self-organizing systems, where an R could have real causal power over the sequences of micro-level events that actually obtain. But to understand the structure of this type of system, we need further resources. Suppose we also assume, as commonly done in theory of causation (for example, Ellis 1986, 1991, 1995, 2000c; Mackie 1974), that no causal antecedent can be necessary or sufficient for its consequent except given certain background conditions which are presupposed by the causal relation. For example, serotonin will stimulate the hypothalamic pleasure center only if many other chemical balances and structural patterns are in place in both the hypothalamus and the serotonin receptors. We can then formulate the problem of top-down causation for dynamical systems in the following way: How can the system have an overall relational pattern (R) such that R has any effect on the causal relations between the physical substrata, P1 , P2 , etc., where P1 is sufficient to cause P2 under the given background conditions? A promising strategy is to consider the possibility that R is structured so as to have a strong tendency for various components of the system to rearrange themselves such that, if the existing background conditions are not conducive to A1 causing B1 under those conditions, then the system becomes rearranged so that A2 will occur, and will cause B2 , which in turn can subserve R2 just as well as B1 could, because of R2 ’s multiply realizable nature. If the causal power of the system as a whole is explained in terms of its being structured so as to have a tendency to rearrange the given background conditions for any discrete causal relation within the system, then it becomes comprehensible how the system as a whole, by virtue of its structure, can actively replace one discrete causal process with another, according to what is needed to maintain the structure of the system, without violating the causal closure of the discrete causal relations between the components. To say that the system rearranges the background conditions for a specific causal relation is to deny neither the causal sufficiency

 Curious Emotions

of that relation itself (under appropriate background conditions) nor the previous causal determination of the genesis of the self-maintaining system itself. Nowhere would physical causal closure be violated. Someone might worry that the logical possibility of such a system does not show that there are any systems of this kind, let alone that motivating intentional states can be conceived of as aspects of such systems in relation to their physical substrata. But whether there are any such systems is an empirical question. Kauffman, Monod and others cite numerous examples of biological organisms that seem to behave according to a principle of self-organization in the sense that the pattern shows a strong tendency to appropriate needed substrata for the maintenance of the pattern. When a given mechanism for maintaining a 98-degree body temperature fails, the organism finds some alternative way to achieve it. When the victim of a mild stroke has lost the neural substrata for certain conscious functions, new cells are appropriated to subserve the relational behavior of those that were destroyed. Now these examples do not prove that the relevant process in each example is exerting real causal power to appropriate and replace constituents, rather than being caused by the discrete actions of the constituents, as top-down dynamical systems theory would have it. All they show is that there are patterns that do act systematically in such a way that, in the final analysis, the continual replacement of parts needed for the pattern’s continuity does tend to be achieved in a very robust way, and with a great degree of flexibility with regard to initial conditions. And a dynamical systems model that coheres with all relevant causal facts would be a plausible way of accounting for this robust continuity of the pattern across such a broad range of substratum replacements. A more positive reason for believing that some systems do fit the dynamical systems causal analysis is that a dynamical systems hypothesis would be consistent with the observed fact of mental causation in a way that no other kind of causal analysis seems coherently able to accommodate. By saying that mental causation is an observed fact, I mean just this: In a situation where a subject intends to raise her hand and then does so, it is usually true to say that if she had not intended to raise the hand then it would not have gone up. Unless we are to assume that the mental intention is a non-physical epiphenomenon, then it must be equivalent with some combination of physical processes which, like all physical processes, have physical consequences. So the notion that mental intentions have causal power cannot be regarded as merely illusory, on pain of lapsing into a non-physical dualism. If dynamical systems theory can account for the phenomenon of mental causation in a coherent way, whereas competing accounts entail that the causal power of a mental intention over the

Chapter 7. Dynamical systems and emotional agency 

resulting bodily movements is purely illusory, then dynamical systems theory would have a decided advantage over the competing theories. A pressing problem is whether a theory that assigns a top-down causal role to a process over its own substratum elements can even be a coherent causal account in the first place. Making such a possibility into a coherent theory requires some conceptual tools if such a scenario is to be clearly distinguished from one in which the discrete behavior of the constituents simply causes the pattern of the process, as in most connectionist systems. To establish that such a theory could cohere with a reasonable causal scenario is the task to which we can now turn. To make the idea of self-maintaining systems as determining background conditions of causal relations a little more precise, we can say that a selfmaintaining system is a system whose organizational pattern, R, is such that, given any physical substrata for R (A1 causing B1 or A2 causing B2 or. . . etc.), (1) if A1 causes B1 , A1 all by itself is neither necessary nor sufficient for B1 given R, although there are certain possible background conditions, BC, such that, given BC, A1 is or would be necessary and sufficient for B1 , depending on whether BC obtains or not;

and (2) given this power of the system to rearrange background conditions for causal relations within it, the causal relations that actually occur in the system have a strong tendency to combine in such a way as to guarantee that pattern R will continue to obtain in the future (where the system is defined by the realization of the relational pattern R).

Turning our attention to condition (1), we can see that in any self-maintaining system R, it is not the case that, if A causes B, then A occurs if and only if B occurs, given R. What is true is that (I) Given R, (A1 or A2 or A3 or. . . ) if and only if (B1 or B2 or B3 or. . . )

In ordinary language, R has the ability to manipulate the background conditions needed for any micro-level causal relation such as A1 if and only if B1 , A2 if and only if B2 , etc., so that at least one of these causal sequences will occur. And at the same time, (II) Given R, (A1 or A2 or A3 or. . . ) will obtain in enough instances to ensure the continuation of organizational pattern R in the future.

 Curious Emotions

It is true that there is some possible set of background conditions, BC1 , such that Given BC1 , A1 if and only if B1

and there is some possible BC2 such that Given BC2 , A2 if and only if B2 .

And Given BC3 , A3 if and only if B3 .

But in a self-maintaining system, R → [(BC1 or BC2 or BC3 ) & (BC1 if and only if A1 ) & (BC2 if and only if A2 ) & etc.]

in enough instances to ensure the continuation of R. That is, R ensures (in enough instances to keep the system going) that if A1 is available to cause B1 , the needed background conditions will be arranged to facilitate this sequence (i.e., BC1 will obtain); but if A1 is not available, whereas A2 is, then the background conditions (i.e., BC2 ) will be arranged so as to facilitate the sequence A2 → B2 , which in turn can subserve the relation R just as well as the sequence A1 → B1 could have done. Thus, in a self-maintaining system, R ensures that both (I) and (II) above obtain in enough instances to ensure the continuation of R. In most physical situations, which are not self-maintaining, this is not the case. This is a typical organizational pattern in biological systems, for example the Krebs energy cycle. In every instance where A causes B, there are some given background conditions under which A occurs if and only if B occurs, but these background conditions do not constitute a self-maintaining system. The fact that, given certain background conditions, A occurs if and only if B occurs, does not make this set of background conditions equivalent with a self-maintaining system, although these background conditions may occur within a self-maintaining system. In sum, a self-maintaining system is organized in such a way that if the background conditions do not obtain under which A1 occurs if and only if B1 occurs, then there is a very strong tendency that the background conditions under which A2 occurs if and only if B2 occurs will obtain, or the background conditions under which A3 occurs if and only if B3 occurs will obtain, or etc., as a result of the organizational structure of the system.

Chapter 7. Dynamical systems and emotional agency 

This notion of a “strong tendency to combine in such a way as to guarantee that pattern R will continue to obtain” needs to be fleshed out a little. There is an important question as to how a self-organizing system can exercise such a tendency. But there are really two different questions packed in here. One is whether there are any such systems, and the other is, if there should be such systems, how it is possible for them to act in this way. The first question is an empirical one, and can best be answered by means of concrete examples, as already mentioned. When the cells of an embryo are transplanted from one brain area to another, sometimes even in a completely different species, at certain stages of development, they take on the functional properties of the new brain area in spite of the alien origin and function of the original cells (Kandel & Schwartz 1981). In effect, the functional organization of the system appropriates the alien cells for its purposes. In the same way, when stroke victims attempt to use a paralyzed limb during concerted amounts of time over a period of several weeks, different brain cells and synapses are appropriated to serve the function of the old destroyed cells and synapses. If this effort is not exerted in this way, the new cells are not appropriated to serve the lost functions. These examples illustrate that there are instances where functional properties of a larger system appropriate micro-components as needed – within certain limits, of course – rather than merely resulting from the interaction of the micro-components. The second part of the question is how self-organizing systems accomplish this purpose. The main principle seems to be that the pattern of the system, multiply realizable by different possible components of this pattern, causes flexibility in the arrangement of the background conditions needed for any given causal micro-sequence to obtain. In effect, the design of the system includes conditions that allow a robust array of antecedents for massive overcausation of a certain outcome. This can be seen, for example, in the shunt mechanisms built into the Krebs energy cycle. Because the system is patterned to contain these mechanisms for overcausation, there is no one element of the system that is either necessary or sufficient for the final outcome – the conversion of energy into ATP for storage. Instead, the initial structure of possible shunt mechanisms is necessary and sufficient to ensure the outcome, provided that minimal alternative chemical substances that are usually readily available can be obtained. A biochemical structure including auto-catalysis further increases the probability that this kind of mechanism will obtain. If none of the alternative substances are indeed available, of course, then the outcome is not produced, the system fails (dies), and ceases to be a self-organizing one.

 Curious Emotions

While not all self-organizing systems involve mentality, motivation, emotion, or purposiveness, what I am suggesting is that some of them do. An organism in which the whole actively seeks out replacement components and conditions needed to replicate its pattern into the future has motivation in a primitive sense. When it also uses a system of monitoring how well it is doing in maintaining the various homeostatic patterns needed to serve the selforganizational aim, of the kind discussed in our introduction, then it can be said to have at least non-conscious emotions. And when it is also able to form imagery (action imagery as well as perceptual) related to the intentional meaning of the emotions – holistic and thus murky though those meanings must be – then it can have conscious emotions (Damasio 1999; Ellis 2000a, b, d). What enables a process to have causal power independently of the particular micro-constituents that subserve it is essentially that the process could have been subserved by other micro-constituents if the actual micro-constituents had failed. So if Process A is causally necessary and sufficient for subsequent process B, there is no specific set of micro-constituents of A that are necessary and sufficient for B. Moreover, in a complex self-organizing system, the whole is structured so that, in may types of cases, if one set of micro-constituents for A do not occur, then an alternative set will be found. This point has extremely important implications for the question of intertheoretic reduction of the various levels of scientific explanation applicable to the human brain – psychology, biology, organic chemistry, inorganic chemistry, and ultimately subatomic physics. If the discrete steps in chemical reactions exhaust the causal power in the system, we end up with a motivational system that is merely driven by complicated networks of inorganic reactions rather than also driving them to serve its own self-organizational aims. This in turn would lead to a very different view of the causal mechanisms as well as the intentional meanings of emotions, especially with regard to their aims. Motivations and emotions are primarily processes that seek out alternative sets of micro-constituents. In the simplest case, hunger is the organism’s attempt to find replacements for chemical substrata for the Krebs energy cycle. In order to do so, the organism must form a vague, holistic understanding of how the system is “off,” and seek out environmental affordances geared toward appropriating the needed components for the continuation of the organism’s own processes.

Chapter 7. Dynamical systems and emotional agency 

. The emotional brain as an enactive system The emotional brain areas – the PAG, midbrain, brain stem, medulla, hippocampus, amygdala, and cingulate, along with the Newman and Baars (1997)/ Watt (2000) “extended reticular thalamic activating system” or ERTAS, which includes the hypothalamus and thalamus (see also Faucher 2002; Panksepp 1998) and parts of the frontal cortex – are the areas most involved in ensuring that brain functioning uses one overall self-organizational causal process as opposed to many separate ones. One of the functions of these emotional brain areas is to keep track of how the whole body is doing in maintaining the best balance of all its self-organizational patterns and transmit this information to the cortex; it uses the cortex to process and execute the needed bodily actions in view of cortically analysed sensory information. If causal processes in the brain consisted only of linear sequences, then it should be possible to explain the outcomes of the process, such as intentional conscious states, by tracking sequences of micro-level neural events. While a century of sophisticated physiology, biochemistry and perceptual psychology have explained a great deal by means of such mechanisms, they have not explained why some processes in the brain are capable of consciousness and intentionality while others are not. The micro-level analysis of vision, for example, can track sequences of neural signals from initial retinal impact through the effects they cause in the thalamus, and the further signals relayed from there to the occipital lobe. Different layers of sensory cortex in occipital and temporal lobes perform computations on the incoming perceptual signals (Hubel & Wiesel 1959; Richardson 1991; Weiskrantz 1986). These transformations are obviously unconscious, because, as we have already discussed, when there is virtually complete occipital activation in response to a completely unexpected stimulus (indicating that the transformations are virtually complete), the subject still lacks perceptual consciousness unless there also occurs a parietal 300P electrical potential (Aftanas 2001; Aurell 1983, 1984, 1989; Damasio et al. 2000; Federmeier & Kutas 2002; McHugh & Bahill 1985; Srebro 1985). Occipital and temporal lobes can do everything they normally do in processing the perceptual data, including the 100P occipital potential and the 200N “mismatch negativity,” without the subject having consciousness of the stimulus. (In the terminology of event-related potentials, as explained earlier, the numbers refer to milliseconds after presentation of the stimulus. Thus extensive processing occurs during the first 250 ms of processing, with or without the consciousness accompanying the 300P.) These unconscious occipital transformations fit a computational paradigm: cells in consecutive layers of sensory cortices analyze different features of per-

 Curious Emotions

ceived objects – lines, angles, shapes, colors, etc. These sequences of transformations are unimaginable on a conscious basis; we cannot imagine consciousness of color without shape or vice versa, yet our sensory cortices “compute” these properties separately and then recombine them. But at the point where we wish to use such micro-level analyses to explain why the consciousness of the object accompanies the 300P, we encounter the anomaly mentioned in the “predictions” section of our introduction: Given the close proximity of the occipital and parietal lobes, the question arises as to why it would take a nerve impulse, which normally travels about 100 miles per hour, approximately a fourth of a second to travel only a few millimeters. If the imaginative activity of the parietal lobe (signalled by the 300P) were really caused by the nerve impulse which supposedly would travel to it from the occipital lobe, the impulse should be delivered virtually instantaneously. It follows that whatever is happening during this fourth of a second that is also needed in order for consciousness of the object to occur cannot be caused merely by the passive receiving of the nerve impulse to the parietal lobe from the occipital lobe, which in turn receives it from the incoming stimulus. Thus it appears that the response is not caused by the stimulus. The reason for this anomaly, as we have seen, is that a more global process must be underway when these linear processes occur. Earlier than the completion of occipital (sensory cortex) processing, the emotional brain, with the help of the cerebellum and its extensive looping signals with the thalamus and hypothalamus (Haines et al. 1997; Schmahmann et al. 2001), has already initiated tentative action commands relevant to general action affordances of the object for the organism. The action commands are inhibited when they reach the motor and premotor areas with the help of inhibitory frontal activity which is partly controlled subcortically. The parietal area, with its body-schema representation of the bodily actions correlative to the perceived object’s motivationally salient action affordances, stands poised to be activated in resonance with the perceived object when the occipital perceptual analysis has finally advanced far enough. When the inhibited action commands are modified to fit the stimulus, so that the electrical wave patterns of these diverse brain areas can come into resonance – only then can the parietal P300 occur, corresponding with perceptual consciousness rather than only non-conscious processing of the information. Why are such diverse brain activities needed to cause consciousness of a signal whose analysis into a perceptual representation at first seems to occur in a relatively discrete series of linear neural pathways involving only the “visual area” of the brain? On the self-organizational view, the reason is that conscious-

Chapter 7. Dynamical systems and emotional agency 

ness and intentionality can arise only from a dynamical causal situation. The emotional brain areas must play an especially unique role in such a dynamical causal process, because they are the areas that monitor and adjust for the intricately interrelated homeostatic balances in all the body’s systems, and aim to maintain the best possible balance among them all, all at once. For example, the motor cortex cannot process or inhibit action commands from the cerebellum (which according to Haines et al. 1997, result from cerebellar-PAG-hippocampal-thalamic loops) unless well supplied with dopamine and norepinephrine, whose main nuclei are in the upper pons and midbrain (Faw 2000; Panksepp 1998). The amydgala-hippocampal loops, activated within 18 ms after presentation of a novel stimulus, cannot direct attention in the ways necessary for consciousness of a visual object (Mack & Rock 1998) unless serotonin (5HT) is in the right balance; the 5HT nuclei are mostly in the raphe nuclei of the brainstem (Faw 2000) – again, areas embedded deep in the emotional brain. All this regulation of neurotransmitters by the emotional brain results from its function in monitoring the total complex interconnection of all the body’s various homeostatic systems in a holistic way, dictated by the selforganizational structure’s one overriding aim: to maintain the ongoing pattern of activity in its definite structure, or as close to that as possible. It is the emotional brain system, more obviously than any other, that forces the microprocesses of the rest of the brain to subordinate themselves to the purposes of the body as a whole. Again, this in no way subtracts from the causal sufficiency of the micro-sequences, given the relevant background conditions. But the background conditions are continually being readjusted in accordance with the whole self-organizational system’s demand to preserve the continuity of its processes across robust replacements of the micro-elements relevant to maintaining that ultimate outcome, with concomitant interrelated continuous readjustments of its own subsystems.

. Objections and responses Motivational causation seems to manifest all the tendencies of complex dynamical systems – including features that traditionally are neglected or rejected by many motivational theories: Among the more controversial of these features are relative autonomy of emotional systems vis a vis their micro-constituents, and a tendency toward high-energy by preference over more energy-efficient homeostatic tendencies. The higher-energy basins, in my view, ground uncon-

 Curious Emotions

ditioned, endogenous drives toward exploration, nurturance, play, curiosity, social bonding, and other activities that are not consummatory in the traditional sense. In light of these controversial implications, we often find (myself included) that, in spite of the theoretical plausibility of a top-down dynamical systems account, a nagging intuition sometimes may continue to urge against it. The remainder of this chapter will therefore address a few of these concerns about the causal structure of self-organizational systems, especially as related to the function of the emotional brain. Consider our example of the performer who drinks six cups of coffee before going onstage. Under the given background conditions, the higher energy level and feeling of inspiration are sufficiently explained by the chemical effect of the caffeine in causing the release of increased ACh and NE, and similar micro-level explanations could be given for his decision to drink the coffee in the first place. Thus it might be argued that, in order really to make the case for the compatibility of ordinary causal micro-mechanisms with the causal power of a self-maintaining system over its own constituents, such a top-down type of dynamical systems theory must show that it is the overall relational pattern of the system (R) that actually drives or causes the rearrangement. But the objector might hold that it is the causal relations and interactions at the microlevel that really drive the rearrangement, and that the overall relational pattern of the system is just a higher-level, coarse-grained description of what is going on at the micro-level. What is really doing the causal work, in the case of mental causation, is “down there in the neurons,” and the relation R is simply token-identical with a particular combination of them. The response is that, first of all, the physical relata for a relation (call them P1 , P2 , etc.) could have occurred without being in that relation. So the relation R is distinguishable from the Ps in that it could have been the R it is, and thus could have created the background conditions needed for subsequent effects, without the particular Ps that, under the given circumstances, happen to subserve it – as long as other suitable Ps could have been found. This means that whether or not R obtains makes a difference in determining subsequent causal relations, while R’s making this difference is not dependent on the action of the specific Ps that, under the given circumstances, happen to subserve it. Other Ps not only could have, but in fact would very likely have subserved the same R if those Ps had not, and it is the R that is necessary, not those specific Ps. For example, the performer’s desire for a higher energy level could have been fulfilled by eating six candy bars rather than by drinking the six cups of coffee. In that case, the chemistry of the micro-mechanisms would have been substantially different, but the aim would have been achieved just as well.

Chapter 7. Dynamical systems and emotional agency

The complex set of relations needed for the desired energy level and feeling of inspiration would then have been subserved by alternative micro-mechanisms. Secondly, it is true that the R ultimately is determined by some Ps (at the micro level), but the R is determined by previous Ps, not by the Ps that serve as the substratum for the current R (since the R plays a role in selecting which Ps will subserve it at each subsequent moment). Therefore, the R is not just a courser-grained description of what is happening in its micro level substrata; the R has the causal power to appropriate different micro-level substrata as needed at each subsequent moment, by rearranging the background conditions to facilitate available causal sequences capable of subserving the ongoing pattern at that moment. Thus the R’s causal power is not token-token identical with the causal power of its micro-level substrata. It is still true that the R could have and very likely would have had that causal power without those particular Ps, although of course the R’s causal power is accountable for by the causal mechanisms that caused that R to be organized in the way it is in the first place; and those causal mechanisms are ultimately (at some point) traceable to previous micro-level events. For example, the performer’s decision to alter his energy level by nutritional means – either drinking coffee or eating candy bars – was in turn determined by some previous R or other, and this previous R may be explainable in terms of still earlier Ps that predetermined it; but the important point for our purposes is that the R now occurring as an outcome of this process – the state of increased energy and enthusiasm – is not identical with those earlier Ps. In case this sounds like a trivial point (that all it means is that at each instant in time the operative R at that time is caused by the Ps in the immediately previous instant), notice that it means much more than that: it means that not all of the causal work is being done “down there in the neurons.” It is also being done by the way in which the neurons are organized, which is a higher-level relational property of the system irrespective of which micro-level constituents are serving as the relata for the relation. I have claimed that the current R, caused by previous Ps, is doing indirect causal work now, in the sense that it is rearranging background conditions needed for the Ps to be causally sufficient for each other. I have taken this to imply that the current R is not identical with or caused by the specific current Ps, having been caused by previous Ps. But there still might be a problem with this answer: Why could one not still hold that the previous Ps caused the current Ps, and that it is the current Ps, not the R, that are doing the causal work? The problem with this viewpoint is that, if the R had not arranged the background conditions in the right way, then some of the earlier Ps would not have



 Curious Emotions

been necessary and sufficient for the subsequent ones. The important point for our purposes is that the R makes a difference as to whether a given P is able to do the causal work assigned to it or not. Dynamical systems theory does not need to claim that the R itself is not in turn caused by previous Ps. For example, when someone floods the synapses of the brain’s pleasure mechanisms with serotonin by taking “ecstacy,” the effect of the serotonin molecule on the receptor at the micro-level is sufficient to produce euphoria, given certain background conditions that usually obtain in the brain. But the brain is also organized in such a way as to compensate for the excess serotonin by changing those background conditions so that a given amount of serotonin is no longer sufficient for this purpose. Addiction – which includes the brain’s self-reorganization to compensate for excessive micro-level manipulations (so that more input is required for the same effect) – is a reflection of the ability of a relational pattern to change the background conditions needed for certain direct micro-level events to be sufficient for their micro-level outcomes. I.e., it is a reflection of the causal difference made by a larger relation, R, with respect to some of its own micro-constituents, by rearranging background conditions. Unfortunately, organisms are not infinitely malleable in this regard: If they were, no micro-level manipulation could ultimately harm them, and they would be immortal. We have seen that R cannot be literally identical with its currently subserving Ps, since a relation is not the same as the relata that are in that relation, and there are conditional statements that are true of the R but not of the specific relata that subserve it; most importantly, the R is necessary for subsequent effects for which those Ps are not necessary, since if those Ps had not occurred in relation R, others would have, because of the way R was structured. Thus there is a conditional statement that is true for R but not for the Ps that subserve the R, namely that the R is necessary for certain subsequent occurrences for which the Ps would not have been necessary, if other suitable Ps had been the ones to subserve R. I am emphasizing that the overall relation R does have real causal power for which the particular subserving micro-constituents would not be literally necessary and sufficient. If the R were literally identical with the Ps that are in relation R, then the R and its subserving Ps would both do all the same “causal work”; but we have seen that the R is not literally identical with its Ps. Moreover, the R cannot be literally identical with the previous Ps that caused it, because things that do not exist at the same time cannot be literally identical; and also, if X causes Y, then X and Y cannot be literally identical with each other. So R is literally identical neither with its currently subserving Ps nor with any previous

Chapter 7. Dynamical systems and emotional agency 

Ps. In sum, no R’s causal powers are reducible to the sum of the causal powers of any set of lower order Ps, for three essential reasons: (1) a relation, by definition, is not the same thing as the relata that are in that relation; (2) the R does not exist at the same time as the Ps that caused it; and (3) as we have seen, there are conditional statements that are true of R that are not true of the Ps and vice versa. For example, R could have (and very likely would have) existed without the currently subserving Ps, if other appropriate Ps could have been found. If no excess serotonin were available to flood the synapses, a similar result could be produced by inhibiting the reuptake of the serotonin, or by flooding the synapses with dopamine instead. R exerts its causal power, then, by being a necessary and sufficient condition for an outcome for which no set of actually occurring Ps is necessary; nor could those Ps have been sufficient for that outcome without that R. That is, if the Ps were arranged in some different R, then there would be a different outcome, and the R in question not only could but would have led to that outcome without those Ps if they had not been available (because R is structured so that other Ps would have ended up serving as its substrata in that case). This point illustrates the importance of the distinction Weinberg (1995) has made between what he calls “petty reductionism” and “grand reductionism” in science. Petty reductionism is defined as the view that “things behave the way they do because of the properties of their constituents.” By contrast, “Grand reductionism. . . [is] the view that all of nature is the way it is because of simple universal laws [which would] reduce the world of physical phenomena to a finite set of fundamental equations” (p. 39). According to these definitions, Weinberg says that “petty reductionism in physics has probably run its course” (p. 39). Dynamical systems theory, in order to make sense of the problem of motivational causation, does not need to reject Weinberg’s grand reductionism – only what he calls petty reductionism.

Conclusion

The main purpose of this book has been to sketch out an enactive account of the emotions that makes room for the “higher” and even the “existential” human emotions, and is not derivative by its very nature from a short list of consummatory-reduction needs. In fact, I have argued that higher emotions are just as likely to be unconditioned and hardwired as are the consummatoryreductive ones. The reason is that conscious beings must be self-organizational dynamical systems, geared toward maintaining certain levels of energy and complexity as well as toward homeostasis and boundary protection. This theory enables us to make a meaningful distinction between self-motivated action and mere reaction. A self-organizing system is one that actively appropriates and replaces the substratum components needed to keep the pattern going, rather than being the passive causal outcome of the interactions of the components. Of course, this is just what living organisms do (Monod 1971; Kauffman 1993). If a dynamical systems account resolves the mental causation problem, it can also resolve the most intractable aspect of the mind-body problem, the “hard problem” (Chalmers 1995). Chalmers argues that if we can show that certain physico-chemical antecedents cause the raising of my hand, and that they operate according to the same physical and chemical principles as in nonconscious parts of nature, then giving a complete physical explanation of all such brain events would still leave out of account anything that would explain why there is consciousness. Dynamical systems theory can answer Chalmers’ objections to physicalism by showing why only certain types of physical systems – complex dynamical ones that include emotional motivations (which of course can sometimes be unconscious) – can have consciousness. And this view seems consistent with a concept of the self as simultaneously available to reflection within any given state, and providing causal power and directionality within the state. To be an actor rather than a mere reactor means to be a system that readjusts its own parts in order to maintain and enhance the continuity of the functioning of the whole, i.e., to be a higher order pattern. Parts of systems can

 Curious Emotions

re-act, but if the system as a whole is a self-maintaining Gestalt (what MerleauPonty 1942, calls a “psychophysical form”), then it can act rather than just react. Of course, to “act” just means to behave in a way that is determined by the tendency of the whole to readjust its parts rather than to be pushed in partes extra parts fashion. Thus consider even a fairly complex mechanical system, i.e., one in which everything that happens to any given part can be described exhaustively in terms of some specific other part’s effect on it, without reference to a tendency of the whole to maintain its overall pattern: Such a mechanical system still cannot “act.” The tendency toward self-organization is the form of inertia that counteracts the inertia that describes objects’ tendency to seek a lower energy level or conserve energy; we might call it an inertia of “action” – a tendency for patterns to maintain themselves by appropriating and replacing the needed substratum elements to facilitate the continuation of the pattern. How can there be such an “inertia of action”? Do all physical things not conform to laws describable in terms of conservation of energy? The inertia of action stems from the fact that patterns in nature show a greater or lesser tendency to maintain themselves over changes in their parts, and such systems control the background conditions under which this or that causal sequence can take place. When we have this kind of inertia on the part of a complex pattern, we say that we have “purposiveness” in nature, even where there is no consciousness involved – e.g., when the organism regulates its heartbeat and blood pressure. There is no violation of the principles of chemistry or physics in such systems. At the level of the substratum for the process, each event has sufficient causal antecedents within the substratum level. What makes self-maintaining and self-organizing systems (of which “living” organisms are examples) different from merely mechanical ones is that the self-maintaining system is organized in such a way that it controls some of the needed background conditions for certain mechanical-causal relations within the system, as when cells are transplanted from the embryo of one species and are appropriated by a completely different brain area of a completely different species, or as in stroke recovery. Systems that “act” in this sense can show purposiveness to the extent that they can maintain their organizational continuity over disruptions of their parts. The enactive approach suggests that not just affective states, but in fact all conscious states are driven by emotion and motivation, because our interest in looking for potentially valenced environmental conditions is a precondition for attention and perception as well as thought. I sketched a theory of the way this priority of the efferent over the afferent takes place in the neurophysiology and phenomenology of conscious processes. The notion that we must first

Conclusion 

“respond” to a stimulus, in order to direct our attention toward it, before we can even see the stimulus is paradoxical only if we assume that the parietal lobe can be activated only as a result of prior occipital activity, which in turn results from prior optic stimulation originating from the environment. But I have reviewed evidence that this is not the case. Instead, what happens is that the parietal lobe is activated by frontal, limbic, and subcortical processes as a result of emotional-motivational activity triggered by thalamic arousal by the stimulus (which in turn arouses the amygdala much more quickly than perceptual processing can occur) only if the stimulus is generally felt as possibly emotionally important for the organism’s purposes (LeDoux 1996; Luria 1980; Posner 1990; Posner & Rothbart 2000; Damasio 1994). The needs of the organism as a whole must first motivate the process of “looking for” the kinds of environmental stimuli that might be important for the organism’s purposes, with the “kinds” categorized prior to perceptual processing in terms of rough and ready potential action affordances. At this point (and prior to the completion of occipital processing), the frontal lobe becomes active in a number of ways, including the inhibition of these first rough-and-ready action commands, resulting in preconscious action imagery (Jeannerod 1997). Jeannerod’s work shows that subjects form vivid mental imagery of the actions they intend to perform only when the action commands are inhibited by frontal activity. As the preconscious, sensorimotor sensing of these action affordances develops more and more precisely, with the help of the inhibitory role of the frontal lobe and also the increasingly refined thalamocortical loops, the parietal lobe then begins to entertain vague sensorimotor images, or pragmatic concepts, of the kinds of emotionally important objects that might be present in the environment. If and when this frontallimbic-parietal activity, once having been developed, finds itself resonating with patterns of activity in the occipital lobe (which reflects sensory stimulation) – only then does perceptual consciousness of a visual image occur (Ellis 1995). When the parietal 300P occurs, the parietal lobe is not activated in response to the occipital lobe’s activity at all. Instead, the organism must purposely activate the frontal and parietal lobes to “look for” emotionally important categories of objects which the thalamus has already alerted the organism might be relevant, and this “looking for” activity has already begun the forming of visual or conceptual imagery (including proprioceptive and sensorimotor imagery associated with possible action affordances) prior to any occipital activity’s having any effect on our perceptual consciousness (since at this point the impulse has not yet “traveled” from the occipital to the parietal lobe). Rather than the frontal-parietal system’s being a response to an occipital stimulus,

 Curious Emotions

the frontal-parietal activation must already have taken place before perceptual consciousness is possible, and the frontal-parietal pattern is what determines whether any given perceptual input will even register in consciousness, i.e., will be attended to. As always, we see that the organism must act on its environment in order to be conscious of it; consciousness cannot result from a mere passive reaction to incoming input. Rather than a stimulus’ causing a response, it is the response which must occur first, and then act on the incoming afferent signals to produce a stimulus. Correlatively, the intentional referent of an emotional feeling should not be automatically equated with the psychophysical cause of the feeling. The cause consists primarily of the physiological events occurring in the nervous system, which I am arguing function on a self-organizational basis and have self-organizational aims. The intentional object can be a real or imaginary environmental event that the organism can or cannot use for the purposes defined by those self-organizational causal processes. Every emotion thus in reality has an extensive and potentially infinite disjunction of possible objects. In many instances, the most immediate of these objects may afford use only as a convenient symbolizing vehicle for the intensification, exploration, and unfolding of an emotion. The organism selects its emotional intentional objects depending on its already ongoing purposes. Both cognitivism and sensationalism in the theory of emotions tend to overemphasize the affective receiving of interoceptive/proprioceptive information from the body’s viscera, and to de-emphasize the efferent action circuits through which emotions are expressed, and ultimately through which – in beings complex enough to be capable of consciousness – they are felt. Cognitivism views conscious emotion as evaluating sensation-like inputs (i.e., interoceptions and proprioceptions) that have already been received. And sensationalism, the view that emotions consist primarily of bodily feelings like sensations, to which we then externally attach causal stories, entails the same problem. Both cognitivism and sensationalism view emotions as resulting from proprioceptive/interoceptive inputs. What sensation and interoception have in common is that they are afferent, and we can therefore receive them in a way that in a sense is passive. Theorists influenced by the information-processing paradigm, or by the behaviorist stimulus-response paradigm, want to view emotion as the causal outcome of a stream of processed inputs. But the crucial problem with such views is that emotion, involving action and thus action imagery, is not passive, and does not merely passively result from the receiving of any input, either sensory or interoceptive.

Conclusion 

The notion that the consciousness of emotion involves sensorimotor imagery, and not just sensory and interoceptive imagery, is a key point because emotions are guides to action. This fact more than anything else is what makes emotional feelings different from other forms of consciousness. Action imagery, studied by Jeannerod (1997), Newton (1982, 1996, 2001), and other recent workers, is an efferent process delivered to the cortex by the subcortical and limbic emotional brain areas. It is different from either sensation or interoception in this regard. Sensation and interoception are passive – we receive them from the environment, or from our own periphery, as input by means of afferent pathways. Action imagery is the only kind of imagery that by its very nature is efferent, and therefore directly connected with emotions. The central aim of affective processes in complicated animals is not merely to consume needed materials and maintain homeostasis, but rather to act in such a way as to maintain the appropriate level of complexity and energy (extropy) while at the same time attending to homeostatic and boundaryprotection needs. With such complexity characterizing the aims of emotions, there will seldom be only one possible response that can achieve the organism’s ultimate objective. Thus it will tend to be the exception rather than the rule that we observe a one-to-one correspondence between stimulus and response. And on those occasions where such a one-to-one correspondence does appear to obtain, it may be only because the organism has not yet found alternative ways to achieve its objective, and thus, at that particular point in time, has habitually chosen one particular response. Many therapists have noted that inflexible one-to-one correspondence between specific stimuli and specific responses is one of the most universal earmarks of neurosis (for example, see Kohut 1985; Miller 1981; Muller 1991; Zachar 2000). One of the serious drawbacks of trying to construe supposedly “higher” emotions as derivative from “basic” ones is that it can mislead us into assuming that, just because some particular response pattern has become habituated (for example, physically fighting when angry), this means that every element in that response pattern is instinctually hardwired, and that when an alternative response pattern is found to achieve the same ends (for example, using a relatively nonviolent set of tools, such as a legal system, to control the adversary’s behavior), the alternative response is somehow less “natural,” and depends on a continuing “repression” of the natural response. Indeed, certain martial arts such as judo illustrate that controlling an adversary need not involve inflicting much damage. In many contexts, a preoccupation with harming the opponent, as opposed to controlling him or destroying his own psychological composure, may more often impair than help in combat. But it is not necessary to sup-

 Curious Emotions

press a natural instinct in order to take this perspective. Anger itself aims to remove the frustration or threat to one’s freedom, prerogatives, territory, or con-specifics. This can be achieved by controlling or getting rid of the adversary in whatever way possible; harming him is only occasionally required. To be sure, an elevated level of general arousal is beneficial in certain combat situations, and anger can facilitate very rapid arousal: this is its primary purpose in terms of natural selection. But on the “derivativist” account of the human emotional life against which I have argued, even when nonviolent means are used to control the adversary, one still primarily wants to fight and harm or kill the adversary, and this natural instinct must be “tamed” and “repressed” by civilization. One of the many questionable assumptions here seems to be that lower animals and primitive humans inflict violent harm on (or even kill) members of their own species more frequently than do “civilized” beings, who have learned to repress this natural instinct. Of course, such an assumption is plainly at odds with statistical facts about violent crime and warfare (for example, see Pincus 2001). I have argued that the essential aim of anger is merely to remove an obstacle to the organism’s already ongoing patterns of activity; the anger per se does not care what means are used to do so. I have emphasized that we should be careful not to fall uncritically into what Husserl (1913) calls the “natural attitude.” We normally direct our attention toward objects in the world. We observe that conjoined physical events cause effects in each other at a simple mechanical level. When we then notice a conjunction between a simple physical event and an emotional effect in ourselves, it is natural to apply the categories of everyday physical observation to ourselves, assuming that the simple environmental event that we have observed must be what has caused our emotional response. More often than not, however, this is simplistic and misleading, and most mature adults learn better. When someone insults me, the feeling that is present does not just inform me that something must be done about this particular insulting behavior. It also and more importantly tells me that I must come to terms with the fact that people in general can and do frequently insult me, that I am generally vulnerable in this respect, partly for contingent and partly for existential reasons. Contained in the felt sense of the moment are also a number of second order feelings about the fact that life is such that people can insult each other, the fact that people are vulnerable, the fact that I am finite and limited, and many other troubling issues. If there are contingent reasons why people often insult me, such as my own personality foibles, second order feelings about those are already contained in the initial response as well. If I am too meek or too arrogant, worries about these related problems will further intensify the

Conclusion 

instantaneous reaction to the insult. Moreover, the presence of such secondorder but already contained sentiments can be detected through subsequent reflection on the felt sense that was instantaneously triggered. This murkiness in the intentionality of emotional responses suggests that, as Ben Ze’ev (2002) has emphasized, we must guard against oversimplifying accounts that do not respect the subtlety of emotional meanings. I have argued that to respect this subtlety entails crediting emotions with enactive and not simply reactive meanings, thus opening the way to view the “higher” emotions as occupying a central place in the most basic motivational structure of human beings, and to understand these more complicated emotions as basic and important aspects of the life of an active and complex dynamical organism.

References

Aftanas, L., Varlamov, A., Pavlov, S., Makhnev, & Reva, N. (2001). Event-related synchronization and desynchronization during affective processing: emergence of valence-related time-dependent hemispheric asymmetries in theta and upper alpha band. International Journal of Neuroscience, 110(3–4), 197–219. Alexander, D. & Globus, G. (1996). Edge of chaos dynamics in recursively organized neural systems. In E. MacCormac & M. Stamenov (Eds.), Fractals of brain, fractals of mind (pp. 31–74). Amsterdam: John Benjamins. Anderson, C. (2000). From molecules to mindfulness: How vertically convergent fractal time fluctuations unify cognition and emotion. Consciousness & Emotion, 1, 193–226. Anderson, C. & Mandell, A. (1996). Fractal time and the foundations of consciousness: Vertical convergence of 1/f phenomena from ion channels to behavioral states. In E. MacCormac & M. Stamenov (Eds.), Fractals of brain, fractals of mind (pp. 75–126). Amsterdam: John Benjamins. Aristotle (1942). Poetics. New York: Oxford University Press. Aurell, C. G. (1984). Perception: A model comprising two modes of consciousness. Addendum II: Emotion incorporated. Perceptual and Motor Skills, 58, 180–182. Aurell, C. G. (1989). Man’s triune conscious mind. Perceptual and Motor Skills, 68, 747–754. Bachmann, T. (2000). The microgenetic approach to the conscious mind. Amsterdam: John Benjamins. Baker, J. & Allen, G. (1968). A course in biology. London: Addison-Wesley. Barresi, J. & Moore, C. (1996). Intentional relations and social understanding. Behavioral and Brain Sciences, 91, 107–154. Bell, C. (1928). Art. London: Chatto and Windus Ltd. Ben Ze’ev, A. (2000). The subtlety of emotions. Boston: MIT Press. Ben Ze’ev, A. (2002). Emotions are not feelings. Consciousness & Emotion, 3(1), 81–89. Bernstein, M., Stiehl, S., & Bickle, J. (2000). The effect of motivation on the stream of consciousness: Generalizing from a neurocomputational model of cingulo-frontal circuits controlling saccadic eye movements. In Ellis & Newton (2000a). Bertalanffy, L. von (1933/1962). Modern theories of development: An introduction to theoretical biology. New York: Harper. Bickhard, M. (2000). Motivation and emotion: an interactive process model. In R. Ellis & N. Newton (Eds.), The caldron of consciousness: Motivation, affect and self-organization. Amsterdam: John Benjamins. Bickle, J. (1992). Multiple realizability and pschophysical reduction. Behavior and Philosophy, 20, 75–88.

 Curious Emotions

Bickle, J. (1998). Psychoneural reduction: The new wave. Cambridge, MA: MIT Press/ Bradford Books. Bickle, J. (2002). Philosophy of neuroscience: A ruthlessly reductive account. Dordrecht: Kluwer Academic Publishers. Bickle, J., Bernstein, M., Heatley, M., Worley, C., & Stiehl, S. (1999). A functional hypothesis for LGN-V1-TRN connectivities suggested by computer simulation. Journal of Computational Neuroscience, 6, 251–261. Bowne, B. P. (1908/1936). Personalism. Norwood, MA: Plimpton Press. Boyce, G. (1998). Review of R. D. Ellis, Questioning consciousness: The interplay of imagery, cognition and emotion in the human brain (John Benjamins), at Darwin and Darwinism website, www.human-nature.com/darwin/index.html Branden, N. (1988). The psychology of romantic love. Toronto: Bantam. Bunge, M. (1979). Ontology II: A world of systems. Dordrecht: D. Reidel. Campbell, N. (1987). Biology. Menlo Park, CA: Benjamin/Cummings. Chalmers, D. (1995). Facing up to the hard problem of consciousness. Journal of Consciousness Studies, 2, 200–219. Clark, A. (1997). Being There: Putting brain, body and world together again. Cambridge: MIT. Coles, M., Gratton, G., & Fabiani, M. (1990). Event-related brain potentials. In Principles of Psychophysiology (pp. 413–453). Cambridge: Cambridge University Press. Courtemanche, R., Pellerin, J. P., & Lamarre, Y. (2002). Local field potential oscillations in primate cerebellar cortex: modulation during active and passive expectancy. Journal of Neurophysiology, 88(2), 771–782. Dalton, T. (2000). The developmental roots of consciousness and emotional experience. Consciousness & Emotion, 1, 57–92. Damasio, A. (1994). Descarte’s error. New York: Putnam. Damasio, A. (1999). The feeling of what happens. New York: Harcourt Brace. Damasio, A. (2003). Looking for Spinoza. New York: Harcourt. Damasio, A. R., Grabowski, T. J., Bechara, A., Damasio, H., Ponto, L. L., & Parvizi, J. (2000). Subcortical and cortical brain activity during the feeling of self-generated emotions. Nature Neuroscience, 3, 1049–1056. Dashiell, J. F. (1925). A quantitative demonstration of animal drive. Journal of Comparative Psychology, 5, 205–208. Davidson, D. (1970). Mental events. In L. Foster & J. W. Swanson (Eds.), Experience and theory (pp. 79–102). Amherst: University of Massachusetts Press. Dennett, D. (1991). Consciousness explained. Boston: Little, Brown and Co. Dennett, D. (2003). Freedom evolves. New York: Viking Press. De Preester, H. (2002). Intentionality and the inside/outside distinction in sensitive systems. Consciousness & Emotion, 3, 65–80. Donoghue, J. P. (2002). Connecting cortex to machines: recent advances in brain interface. Nature Neuroscience Supplement, 5 (November 2002), 1085–1088. Edelman, G. (1989). The remembered present. New York: Basic Books. Edelman, G. (1992). Bright air, brilliant fire: On the matter of mind. New York: Basic Books. Edwards, B. (1979). Drawing on the right side of the brain. Boston: Houghton Mifflin. Ekman, P. (1999). Basic emotions. In T. Dalgleish & M. Power (Eds.), Handbook of cognition and emotion (pp. 45–60). Cichester, UK: John Wiley & Sons.

References 

Ellis, R. D. (1981). The decoration-illustration-art trichotomy. Auslegung, 8, 105–121. Ellis, R. D. (1986). An ontology of consciousness. Dordrecht: Kluwer/Martinus Nijhoff. Ellis, R. D. (1991). A critique of concepts of non-sufficient causation. Philosophical Inquiry, 13, 22–42. Ellis, R. D. (1995). Questioning consciousness: The interplay of imagery, cognition and emotion in the human brain. Amsterdam: John Benjamins. Ellis, R. D. (1996a). Ray Jackendoff ’s phenomenology of language as a refutation of the appendage theory of consciousness. Pragmatics & Cognition, 4, 125–137. Ellis, R. D. (1996b). Eros in a narcissistic culture. Dordrecht: Kluwer. Ellis, R. D. (1999a). Why isn’t consciousness empirically observable? Emotion, self-organization, and nonreductive physicalism. Journal of Mind and Behavior, 20, 391–402. Ellis, R. D. (1999b). Dynamical systems as an approach to consciousness: Emotion, selforganization, and the mind-body problem. New Ideas in Psychology, 17, 237–250. Ellis, R. D. (2000a). Consciousness, self-organization, and the process-substratum relation: Rethinking nonreductive physicalism. Philosophical Psychology, 13, 173–190. Ellis, R. D. (2000b). Efferent brain processes and the enactive approach to consciousness. Journal of Consciousness Studies, 7, 40–50. Ellis, R. D. (2000c). Three elements of causation: Biconditionality, asymmetry, and experimental manipulability. Philosophia, 29, 1–21. Ellis, R. D. (2000d). Integrating the physiological and phenomenological dimensions of affect and motivation. In R. Ellis & N. Newton (Eds.), The caldron of consciousness: Motivation, affect, and self-organization (pp. 91–10). Amsterdam: John Benjamins. Ellis, R. D. (2001a). Implications of inattentional blindness for “enactive” theories of consciousness. Brain and Mind, 2, 297–322. Ellis, R. D. (2001b). A theoretical model of the role of the cerebellum in cognition, attention and consciousness. Consciousness & Emotion, 2, 300–309. Ellis, R. D. & Newton, N. (1998a). Consciousness and the brain: An annotated bibliography. http://earthlink.net/∼dravita (Internet site.) Ellis, R. D. & Newton, N. (1998b). Three paradoxes of phenomenal consciousness: Bridging the explanatory gap. Journal of Consciousness Studies, 5, 419–442. Ellis, R. D. & Newton, N. (Eds.). (2000a). The caldron of consciousness: Affect, motivation, and self-organization. Amsterdam: John Benjamins. Ellis, R. D. & Newton, N. (2000b). The interdependence of consciousness and emotion. Consciousness & Emotion, 1, 1–11. Farah, M. (1989). The neural basis of mental imagery. Trends in Neuroscience, 12, 395–399. Faw, B. (2000). Consciousness, motivation, and emotion: Biopsychological reflections. In Ellis & Newton, The caldron of consciousness (pp. 55–90). Amsterdam: John Benjamins. Faw, B. (1997). Outlining a brain model of mental imaging abilities. Neuroscience and Biobehavioral Reviews, 3, 282–288. Federmeier, K. D. & Kutas, M. (2002). Picture the difference: electrophysiological investigations of picture processing in the two cerebral hemispheres. Neuropsychologia, 40(7), 730–747. Francescotti, R. (1998). What multiple realizability does not show. Journal of Mind and Behavior, 18, 13–28. Freeman, W. (1975). Mass action in the nervous system. New York: Academic Press.

 Curious Emotions

Freeman, W. (1987). Simulation of chaotic EEG patterns with a dynamic model of the olfactory system. Biological Cybernetics, 56, 139–150. Freeman, W. (1988). Why neural networks don’t yet fly: Inquiry into the neurodynamics of biological intelligence. Proceedings of the IEEE International Conference on Neural Networks, Volume 2 (pp. 1–7). San Diego: IEEE (Institute of Electrical and Electronics Engineers). Freud, S. (1925/1959). Beyond the pleasure principle. New York: Bantam. Gallagher, S. & Marcel, A. (1999). The self in contextualized action. Journal of Consciousness Studies, 6(4), 4–30. Gallagher, S. & Jeannerod, M. (2002). From action to interaction. Journal of Consciousness Studies, 9(1), 4–30. Gallant, J. L., Conner, C. E., & van Essen, D. C. (1998). Neural activity in areas V1, V2, and V4 during free viewing of natural scenes compared to controlled viewing. NeuroReports, 9. Gazzaniga, M. (Ed.). (2000). The new cognitive neurosciences, Cambridge, MA: MIT Press. Gendlin, E. (1962/1998). Experiencing and the creation of meaning. Toronto: CollierMacmillan. Gendlin, E. (1981/1982). Focusing. Toronto: Bantam. Gendlin, E. (1992a). The Primacy of the Body, Not the Primacy of Perception. Man and World. Gendlin, E. (1992b). Thinking Beyond Patterns: Body, Language, and Situations. In B. den Ouden & M. Moen (Eds.), The presence of feeling in thought. New York: Peter Lang. Gibson, E. J. (1988). Exploratory behavior in the development of perceiving, acting, and the acquiring of knowledge. Annual Review of Psychology, 39, 1–41. Gillett, G. R. & Mcmillan, J. (2001). Consciousness and intentionality. Amsterdam: John Benjamins. Globus, G. (1992). Towards a noncomputational cognitive neuroscience. Journal of Cognitive Neuroscience, 4, 299–310. Goleman, D. (1994). Emotional Intelligence. New York, Bantam. Gottwald, B., Mihajlovic, Z., Wilde, B., & Mehdorn, H. M. (2003). Does the cerebellum contribute to specific aspects of attention? Neuropsychologia, 41(11), 1452–1460. Gregory, R. L. (1970). The Intelligent Eye. New York: McGraw-Hill. Griffiths, P. E. (1997). What emotions really are: The problem of psychological categories. Chicago: University of Chicago Press. Haines, D., E., Dietrich, E., Mihailoff, G. A., & McDonald, E. F. (1997). Cerebellar-hypothalamic axis: Basic circuits and clinical observations. In J. Schmahmann (Ed.), The cerebellum and cognition (pp. 84–110). New York: Academic Press. Harlow, H. F. (1950). Learning motivated by a manipulation drive. Journal of Experimental Psychology, 40, 228–234. Harth, E. (1983). Order and chaos in neural systems: an approach to the dynamics of higher brain functions. IEEE Transactions on Systems, Man and Cybernetics, 13, 782–789. Hatsopoulos, N. G., Harrison, M. T., & Donoghue, J. P. (2001). Representations based on neuronal interactions in motor cortex. Progress in Brain Research, 130, 233–244. Helmholtz, H. (1962). Helmholtz’s Treatise on physiological optics. J. P. C. Southall (Trans.). New York: Dover.

References 

Hess, G. & Donoghue, J. P. (1999). Facilitation of long-term potentiation in layer II/III horizontal connections of rat motor cortex following layer I simulation: route of effect and cholinergic contributions. Experimental Brain Research, 127(3), 279–290. Horney, K. (1937). The neurotic personality of our time. New York: W. W. Norton. Horney, K. (1950/1991). Neurosis and human growth. New York: W. W. Norton. Hubel, D. H. & Wiesel, T. N. (1959). Receptive fields of single neurons in the cat’s striate cortex. Journal of Physiology, 148, 574–591. Hull, C. L. (1952). A Behavior System. New Haven: Yale University Press. Hume, D. (1960). Of tragedy. In Richard Levin (Ed.), Tragedy. New York: Harcourt, Brace and World. Humphrey, N. (2000). How to solve the mind-body problem. Journal of Consciousness Studies, 7, 5–20. Husserl, E. (1913/1970). Logical investigations. New York: Humanities Press. Husserl, E. (1913/1931). Ideas. W. R. Boyce Gibson (Trans.). London: Collier 1931; from “Ideen zu einer Reinen Phänomenologie und Phänomenologischen Philosophie,” 1913. Husserl, E. (1900/1962). Phänomenologische Psychologie. Den Haag: Martinus Nijhoff. Hutto, D. (2000). Beyond Physicalism. Amsterdam/Philadelphia: John Benjamins. Ito, M. (1993). Movement and thought: Identical control mechanisms by the cerebellum. Trends in the Neurosciences, 16(11), 448–450. Jackendoff, R. (1996). How language helps us think. Pragmatics & Cognition, 4, 1–34. Jackson, F. (1986). What Mary didn’t know. Journal of Philosophy, 83, 291–295. Jeannerod, M. (1994). The representing brain: Neural correlates of motor intention and imagery. Behavioral and Brain Sciences, 17, 187–244. Jeannerod, M. (1997). The cognitive neuroscience of action. Oxford: Blackwell. Joseph, R. (1982). The neuropsychology of development: Hemispheric laterality, limbic language and the origin of thought. Journal of Clinical Psychology, 38, 4–33. Juarrero, A. (1999). Dynamics in action: Intentional behavior as a complex system. Cambridge, MA: MIT/Bradford. Kagan, J. & Berkun, M. (1954). The reward value of running activity. Journal of Comparative Physiological Psychology, 47, 108–110. Kandel, E. & Schwartz, J. (1981). Principles of neural science. New York: Elsevier-North Holland. Kauffman, S. (1993). The origins of order. Oxford: Oxford University Press. Kelso, J. A. (1995). Dynamic patterns: The self-organization of brain and behavior. Cambridge, MA: MIT/Bradford. Kim, J. (1992). Multiple realization and the metaphysics of reduction. Philosophy and Phenomenological Research, 52, 1–26. Kim, J. (1993). The nonreductivist’s troubles with mental causation. In J. Heil & A. Mele (Eds.), Mental causation (pp. 189–210). Oxford: Oxford University Press. Kim, J. (1998). Mind in a physical world: An essay on the mind-body problem and mental causation. Cambridge, MA: MIT Press. Klein, M. (1975). Envy and gratitude. Hogarth Press, English Institute of Psycho-Analysis. Kohut, H. (1985). Self psychology and the humanities. New York: W. W. Norton. Krutch, J. W. (1956). The tragic fallacy. In The modern temper: A study and a confession. New York: Harcourt, Brace.

 Curious Emotions

Kuhn, T. (1962). The structure of scientific revolutions. Chicago: The University of Chicago Press. Langer, S. (1957). Problems of art. New York: Scribners. LeDoux, J. (1996). The emotional brain. New York: Simon & Schuster. Lethin, A. (2002). How do we embody intentionality? Journal of Consciousness Studies, 9(8), 36–44. Lethin, A. (2004). Exposing the covert agent. In R. Ellis & N. Newton (Eds.), Consciousness & emotion: Agency, conscious choise, and selective perception (pp. 157–180). Amsterdam: John Benjamins. Levinas, E. (1969). Totality and infinity. The Hague: Martinus Nijhoff. Levy, J., Trevarthen, C., & Sperry, R. W. (1972). Perception of bilateral chimeric figures following hemispheric deconnexion. Brain, 95, 61–78. Libet, B. (1999). Do we have free will? Journal of Consciousness Studies, 6, 47–58. Lowen, A. (1985). Narcissism: Denial of the true self. New York: Macmillan. Luria, A. R. (1980). Higher cortical functions in man (2nd ed.). New York: Basic Books. Mac Cormack, E. & Stamenov, M. (Eds.). (1996). Fractals of brain, fractals of mind. Amsterdam: John Benjamins. Mack, A. & Rock, I. (1998). Inattentional blindness. Cambridge: MIT/Bradford. Mackie, J. L. (1974). The cement of the universe. Oxford: Oxford University Press. Marcel, A. (2003). The sense of agency: Awareness and ownership of action. In J. Roessler & N. Eilan (Eds.), Agency and self-awareness (pp. 48–93). Oxford: Clarendon. Maslow, A. (1974). Motivation and personality. New York: Harper & Row. May, R. (1969). Love and will. New York: W. W. Norton. McHugh, D. E. & Bahill, A. T. (1985). Learning to track predictable target waveforms without a time delay. Investigative Ophthalmology and Visual Science, 26, 932–937. Meltzoff, A. & Gopnik, A. (1993). The role of imitation in understanding persons and developing theories of mind. In S. Baron-Cohen, H. Tager-Flusberg, & D. Cohen (Eds.), Understanding other minds: Perspectives from autism (pp. 335–366). Oxford University Press. Merleau-Ponty, M. (1941/1962). Phenomenology of Perception. Colin Smith (Trans.). New York: Humanities Press. Merleau-Ponty, M. (1942/1963). The structure of behavior. Boston: Beacon. Merleau-Ponty, M. (1948/1968). The visible and the invisible. Evanston: Northwestern University Press. Miller, A. (1951). Tragedy and the common man. Theatre Arts, March, 1951. Miller, A. (1981). The drama of the gifted child. New York: Basic Books. Monod, J. (1971). Chance and necessity. New York: Random House. Montgomery, K. G. (1955). The role of the exploratory drive in learning. Journal of Comparative Physiological Psychology, 47, 60–64. Muller, R. (1991). The marginal self. Atlantic Highlands, NJ: Humanities Press International. Nagel, T. (1974). Physicalism. Philosophical Review, 74, 339–356. Natsoulas, T. (1993). What is wrong with appendage theory of consciousness. Philosophical Psychology, 6, 137–154. Natsoulas, T. (2000). On the intrinsic nature of states of consciousness: Further considerations in the light of James’s conception. Consciousness & Emotion, 1, 139–166.

References 

Needleman, J. (1968). Being in the world: Selected papers of Ludwig Binswanger. New York: Harper & Row. Newman, J. & Baars, B. J. (1993). A neural attentional model for access to consciousness: A Global Workspace perspective. Concepts in Neuroscience, 4(2), 255–290. Newton, N. (1982). Experience and imagery. Southern Journal of Philosophy, 20, 475–487. Newton, N. (1989a). Error in action and belief. Philosophia, 19, 363–401. Newton, N. (1989b). On viewing pain as a secondary quality. Nous, 23, 569–598. Newton, N. (1993). The sensorimotor theory of cognition. Pragmatics & Cognition, 1, 267– 305. Newton, N. (1996). Foundations of understanding. Amsterdam: John Benjamins. Newton, N. (2000). Conscious emotion in a dynamic system: How I can know how I feel. In R. Ellis & N. Newton (Eds.), The caldron of consciousness: Motivation, affect, and self-organization (pp. 91–108). Amsterdam: John Benjamins. Newton, N. (2001). Emergence and the uniqueness of consciousness. Journal of Consciousness Studies, 8, 47–59. Nicolis, J. S. (1986). Dynamics of hierarchical systems. New York: Springer Verlag. Nikolaev, A. R., Ivanitsky, G. A., Ivanitsky, A. M., Posner, M. I., & Abdullaev, Y. G. (2001). Short-term correlation between frontal and Wernicke’s areas during word association: an event-related potential analysis in human subjects. Neuroscience Letters, 298, 107– 110. Nissen, H. W. (1930). A study of exploratory behavior in the white rat. Journal of Genetic Psychology, 37, 361–376. Nunn, C. (2001). Review of R. D. Ellis & N. Newton (Eds.), The Caldron of Consciousness (John Benjamins). Journal of Consciousness Studies, 8, 76–78. Ornstein, R. & Thompson, R. (1984). The amazing brain. Boston: Houghton Mifflin. Panksepp, J. (1998). Affective neuroscience. New York: Oxford University Press. Panksepp, J. (2000). The neuro-evolutionary cusp between emotions and cognitions: Implications for understanding consciousness and the emergence of a unified mind science. Consciousness & Emotion, 1, 17–56. Panksepp, J., Siviy, S. M., & Normansell, L. A. (1984). The psychobiology of play: theoretical and methodological perspectives. Neuroscience and Biobehavioral Reviews, 8, 465–492. Pavlov, I. P. (1929). Lecons sur l’activite du cortex cerebral. Paris: Legrand. Pincus, J. H. (2001). Base instincts: What makes killers kill? New York: W. W. Norton. Plutchik, R. (1980). A general psychoevolutionary theory of emotion. In R. Plutchik & H. Kellerman (Eds.), Emotion: Theory, research, and experience, Vol. I (pp. 3–33). New York: Academic Press. Posner, M. I. (1990). Hierarchical distributed networks in the neuropsychology of selective attention. In A. Caramazza (Ed.), Cognitive neuropsychology and neurolinguistics: Advances in models of cognitive function and impairment (pp. 187–210). New York: Plenum. Posner, M. I. & Rothbart, M. K. (1992). Attentional mechanisms and conscious experience. In A. D. Milner & M. D. Rugg (Eds.), The neuropsychology of consciousness (pp. 187– 210). London: Academic Press. Posner, M. I. & Rothbart, M. K. (1998). Attention, self regulation and consciousness. Philosophical Transactions of the Royal Society of London B, 353, 1915–1927.

 Curious Emotions

Posner, M. I. & Rothbart, M. K. (2000). Developing mechanisms of self regulation. Development and Psychopathology, 12, 427–441. Pribram, K. (1980). Mind, brain, and consciousness: The organization of competence and conduct. In Julian Davidson & Richard Davidson (Eds.), The Psychobiology of Consciousness (pp. 47–64). New York: Plenum Press. Pribram, K. (1991). Brain and perception: Holonomy and structure in figural processing. Hillsdale, NJ: Lawrence Erlbaum. Prigogine, I. (1996). The end of certainty: Time, chaos, and the new laws of nature. New York: Free Press. Pylyshyn, Z. (1973). What the mind’s eye tells the mind’s brain. Psychological Bulletin, 80, 1–23. Rank, O. (1924/1993). The trauma of birth. New York: Dover. Reik, T. (1945). Psychology of sex relations. New York: Rinehart. Richardson, J. (1991). Imagery and the brain. In Cesare Cornoldi & Mark McDaniel (Eds.), Imagery and cognition (pp. 1–46). New York: Springer-Verlag. Rolls, E. (1999). The brain and emotion. Oxford: Oxford University Press. Runeson, S. (1974). Constant velocity – not perceived as such. Psychological Research, 37, 3–23. Salmela, M. (2002). The problem of affectivity in cognitive theories of emotion. Consciousness & Emotion, 3 (forthcoming). Sartre, J. P. (1966). Being and nothingness. New York: Washington Square Press. Sartre, J. P. (1969). The transcendence of the ego. New York: Noonday. Scheler, M. (1970). The nature of sympathy. London: Archon. Schmahmann, J. (Ed.). (1997). The cerebellum and cognition. New York: Academic Press. Schmahmann, J., Anderson, C., Newton N., & Ellis, R. (2001). The function of the cerebellum in cognition, affect and consciousness: Empirical support for the embodied mind. Consciousness & Emotion, 2, 273–309. Searle, J. (1984). Minds, brains, and science. Cambridge, MA: Harvard University Press. Searle, J. (2000). Consciousness, free action, and the brain. Journal of Consciousness Studies, 7, 3–22. Serruya, M., Hatsopoulos, N., Fellows, M., Paninski, L., & Donoghue, J. (2003). Robustness of neuroprosthetic decoding algorithms. Biological Cybernetics, 88(3) (March 2003), 219–228. Shevrin, H. (2001). Event-related markers of unconscious processes. International Journal of Psychophysiology, 42, 209–218. Smart, J. J. C. (1963). Materialism. Journal of Philosophy, 6(0), 651–662. Smart, J. J. C. (1970). Sensations and brain processes. Philosophical Review, 68. Reprinted in C. V. Borst (Ed.), The mind-brain identity theory (pp. 141–56). London: Macmillan. Spence, K. W. (1956). Behavior theory and conditioning. New Haven: Yale University Press. Spence, S. A. & Frith, C. D. (1999). Towards a functional anatomy of volition. Journal of Consciousness Studies, 6, 11–29. Spitz, R. A. & Wolf, K. M. (1946). Anaclitic depression: An inquiry into the genesis of psychiatric conditions in early childhood. P. A. Study of the Child, II. New York: International University Press.

References

Srebro, R. (1985). Localization of visually evoked cortical activity in humans. Journal of Physiology, 360, 233–246. Stein, E. (1989). The problem of empathy. Washington: ICS Publications. Sudnow, D. (1978/1999). Ways of the hand. Cambridge, MA: MIT/Bradford. Thelen, E., Schoner, G., & Scheier, C. (2001). The dynamics of embodiment: A field theory of infant perseverative reaching. Behavioral and Brain Sciences, 24, 1–86. Thelen, E. & Smith, L. (1994). A dynamic systems approach to the development of cognition and action. Cambridge, MA: MIT Bradford. Thomas, N. (1989). Experience and theory as determinants of attitudes toward mental representation. American Journal of Psychology, 102, 395–412. Treisman, A. M. (1964). Selective attention in man. British Medical Bulletin, 20, 12–16. Tucker, D. (1981). Lateral brain function, emotion and conceptualization. Psychological Bulletin, 89, 19–43. Unamuno, M. de (1972). The tragic sense of life Princeton: Princeton University Press. Varela, F., Thompson, E., & Rosch, E. (1991/1993). The embodied mind. Cambridge: The MIT Press. Watt, D. (2000). The centrencephalon and thalamocortical integration: Neglected contributions of periaqueductal gray. Consciousness & Emotion, 1, 91–114. Watt, D. (1998). Affect and the ‘hard problem’ neurodevelopmental and corticolimbic network issues. In Consciousness Research Abstracts: Toward a Science of Consciousness, Tucson 1998 (pp. 91–92). Wegner, D. (2003). The illusion of conscious will. Cambridge, MA: MIT. Weinberg, S. (1995). Reductionism redux. New York Review, Oct. 5, pp. 39–42. Weiskrantz, L. (1986). Blindsight: A case study and implications. Oxford: Oxford University Press. Weiss, P. A. (1968). The living system: Determinism stratified. In A. Koestler & J. R. Smythies (Eds.), Beyond reductionism – New perspectives in the life sciences. The Alpbach symposium (pp. 3–55). London: Hutchinson. Wertz, F. J. (1987). Cognitive psychology and the understanding of perception. Journal of Phenomenological Psychology, 18, 103–142. White, R. (1959). Motivation reconsidered. Psychological Review, 65, 297–333. Wider, K. (1997). The bodily nature of consciousness. Ithaca: Cornell University Press. Wohlschläger, A. & Bekkering, H. (2002). The role of objects in imitation. In M. Stamenov & V. Gallese (Eds.), Mirror Neurons and the Evolution of Brain and Language (pp. 101– 114). Amsterdam/Philadelphia: John Benjamins. Wolfe, J. B. & Kaplon, M. D. (1941). Effect of amount of reward and consummative activity on learning in chickens. Journal of Comparative Psychology, 31, 353–361. Woodruff-Pak, D. S. (1997). Clasical conditioning. In J. Schmahmann (Ed.), The cerebellum and cognition (pp. 342–366). New York: Academic Press. Wu, W., Black, M. J., Mumford, D., Gao, Y., Bienenstock, E., & Donoshue, J. P. (2003). A switching Kalman filter model for the motor cortical coding of hand motion. Proceedings of IEEE EMBS, Cancun, Sept. 2003. Yarbus, A. L. (1967). Eye movement and vision. New York: Plenum.



 Curious Emotions

Zachar, P. (2000). Child development and the regulation of affect and cognition in consciousness: A view from object relations theory. In R. Ellis & N. Newton (Eds.), The caldron of consciousness: Motivation, affect, and self-organization (pp. 205–222). Amsterdam: John Benjamins.

Index

A acetylcholine (ACh) 12, 75, 107, 109, 189 action 18–20, 42–45, 52–56, 131–137, 167–175, 189–208 action imagery 2–3, 9–10, 40, 42ff., 66, 70, 116, 134, 143–144, 160–161, 206, 217–219 action theory of intentionality 21, 42–43 and passim action versus reaction 18ff., Chapter 2 action affordances 2–4, 7, 11–13, 19–21, 25–29, 36–37, 70–71, 116, 121–123, 125, 129, 135, 147, 164, 173–174, 208, 217 action–consciousness connection 115ff. and passim addiction 39, 124, 212 adventure 15, 22, 111, 114, 117, 126 affect 1–45, 62ff., 93 and passim affective intentionality 1–9 and passim afferent 2–5, 9, 12–13, 36–43, 109, 116, 124, 145–148, 152, 172, 186, 216–219 agency 109ff. agency, primitive 138ff. aims of emotion 33–41 akinetic mutism 11 alexithymic 5, 31 amygdala 26, 62, 75, 107, 113–115, 143, 175, 207, 217 anger 3, 5, 15, 17, 20, 22, 30–35, 38, 62, 104–105, 112–127, 147, 162, 187, 220 anger, in infants 109, 116–117 anterior cingulate 2, 13, 43, 68–76, 107, 143, 169, 175 anxiety 15, 39, 119, 122–123

“appendage” theory 16ff. appetitive phase 15, 16, 107, 109, 117, 119 art 167–176 and passim artificial intelligence 61 as-if body loop 9ff. ATP 85ff., 94ff. attention 47–78 and passim attractor see basin of attraction auto-catalytic systems 197ff. B background conditions 19, 149, 190, 209, 212 basic emotions 15ff., 104 and passim basin of attraction 4, 77, 91, 93, 115, 122–123, 128, 196 behaviorism 13, 15, 85, 104, 106–110, 114, 218 bio-medical model in psychiatry 191 bipolar 38, see also depression body image 2ff. body schema 2ff. bonding 3, 5, 81, 84, 94, 104, 107–108, 119, 210 borderline disorder 159 boundaries, osmotic and interactive 103 boundary needs 103ff. brain-stem 3, 109–116 Broca speech area 116 C categories of utility 175 catharsis 182 causal closure 200ff. and passim

 Curious Emotions

cell 12, 19, 57, 202, 205, 207, 216 cerebellar-hypothalamic loops 25 cerebellar-thalamic loops 209 cerebellum 1–2, 11, 43, 113, 141–148, 208–209 chaos 105ff. choice 142ff. classical conditioning 4, 14, 15, 16, 106, 117 closed systems 105ff. cognitivism 25–35, 42, 51, 218 complexity 4, 9, 14,15, 16, 26, 57, 69, 84, 90, 92, 94, 96, 99, 111, 117–118, 126, 151, 156, 180, 186, 215, 219 condition of readiness 175–176 conditioning 4, 14, 15, 16, 106, 117 consciousness 1–8, 25–33, 63ff., 73–78, 128–130, 138–166 constructed self 164ff. consummatory drives 16, 22–23, 39, 79ff., 103ff. and passim consummatory phase see appetitive phase consummatory-drive reductionism 82ff. cooperation 15, 126 core consciousness 11 core self 133ff. cortical inhibition see inhibition culture 15, 69, 175, 183 curiosity 3, 5, 14–15, 22, 39, 56, 76–77, 88–93, 106, 115, 119, 124–126, 168–172, 210

D death instinct 83 depression 38, 123, 127, 130, 182, 194 despair 15, 41, 179 disgust 113 drive, definition of 97 drugs, as bottom-up causes 194ff. dualism 16, 145, 202 dynamical causal sequences 52–56, 99–102, 193–206

dynamical systems 16ff., 36, 41–45, 47ff., 79–118, 126, 146ff., 170ff., 189ff., 223ff. E early and late selection for attention 58ff., 69–78 early selection 58ff. ecstasy 194 efferent 2–9, 12–13, 36–43, 109, 116, 124, 145–148, 152, 172, 186, 216–219 elan vital 84 electrostatic reduction 89ff. embodied mind-body theory 14ff. and passim embodied self 138ff., 150ff. emotion versus motivation 27–28 emotional learning see learning enactive approach 1–20 and passim endergonic versus exergonic reactions 86ff. epiphenomenalism 16–17, 189, 199 equilibrium 89ff. event related potentials (ERPs) 8, 12, 36, 43, 63, 65, 66, 113, 207 exergonic see endergonic versus exergonic reactions existential anxiety see anxiety expectancy system see seeking system exploration 3, 21, 34, 77, 79, 88–91, 93, 96, 103–104, 108–110, 115, 119, 123, 168–171, 177, 181, 187, 210, 218 exploratory drives 107–108 extended reticular thalamic activating system (ERTAS) 207 extended self 131ff. extropy 89ff. eye movements, necessary for vision 172 eyelid conditioning 148 F fear 3, 15, 22, 26, 28–31, 36–38, 62, 96–97, 107, 113–123, 126–129, 160

Index 

feedback loops 19, 29, 58, 61–64, 69, 82–89, 100, 147, 192–194 feeling of familiarity 114ff. feeling versus emotion 6ff. and passim felt shift 164 felt sense/felt quality 20, 25, 32–37, 42, 80, 120–121, 125, 127, 131, 138, 144, 162–168, 171, 174–179, 217–221 fight or flight 19 focusing 34, 62, 81, 156–157, 172, 177, 179 free will 189 freedom 115ff. frontal cortex 13–14, 43, 64, 68, 71–72, 75, 116, 143, 145, 147, 152, 169–173, 175, 207–208, 217–218 frozen systems 105ff. frustration 20, 32, 108, 122, 127, 187, 220 G gating 79ff. Gestalt 70, 172–174, 216 global wave-forms 199ff. glutamate 26, 107 H habit 40, 43, 69, 142–143, 155, 219 handle (in Gendlin’s sense) 176ff. hard problem of consciousness 215 height, fear of 26, 28, 38, 121 hermeneutic circle 162 higher emotions 13ff., 31ff., 50–51, 79ff. and passim higher-energy basins of attraction 90ff. hippocampus 25, 66, 107, 113–115, 148, 175, 207, 209 histamine 75 homeostasis 22ff., 103ff. and passim hunger drive 17 hypervigilance 113 hypothalamocerebellar loops 148 hypothalamus 2, 26, 107–113, 143, 147, 175, 201, 207–208

I identity, psychophysical see psychophysical identity illusory-choice model 142ff. image schema 70ff. imagery 3, 9–12, 22, 40–52, 66, 70–71, 116, 125–126, 134, 141, 143, 147–148, 159–162, 164, 168ff., 206, 217–219 imitation learning 10ff. implicit knowledge/perception 10, 18, 32–34, 48–49, 59–73, 116, 135, 154, 157, 163 inattentional blindness 47–52, 63–78 and passim infant development 91–92 inhibition 10–12, 42–43, 66, 74–76, 86–90, 97, 116, 134, 145–148, 189, 200, 208–209, 213, 217 inhibitory neurons see inhibition inspiration 15, 119, 124–128, 171, 181, 184, 190, 210–211 intentional versus mechanical explanation 190ff. intentionality 1–9, 15–17, 21–22, 25–45, 47, 120, 138, 165–167, 170, 207, 209, 221 interoception 2–9, 20–21, 42–46, 158ff., 218–232 intertheoretic reduction 82ff. intrinsic versus instrumental values 110ff. introspection 31–33, 158ff.

J joy 15, 90, 97, 123 just so stories 4, 95

K knowledge argument 97, 100 Krebs energy cycle 97ff.

 Curious Emotions

L late selection for attention 58ff., 69–78 learning 4, 15–16, 22, 39, 68, 70, 75, 85, 96, 104, 106, 108, 110, 112, 123, 147, 182–183, 196 life wish 95ff. linear versus dynamical causal sequences 52–56 living systems, definition of 18ff. loneliness 15, 127 love 15, 25, 32, 104, 122–125, 127–128, 158, 176–177, 181ff. lust 5, 13, 15, 119, 127–129, 141, 162, 170, 189, 194, 205, 213, 219 lust system see lust M mapping (bodily) 3, 7, 36 marasmus 123 medulla 113 mental causation 140ff. mentality see intentionality micro-reductionism 16, 84–89 midbrain 3, 63, 70, 107, 113, 116, 207–209 mirror neurons 10ff. mismatch negativity 148 modularity 50ff. monoamines 26, 107 moral agency 40 motivation 27–30 and passim motivation versus emotion 27–28 motor cortex 2, 13, 43, 113, 142–148, 208–209 motor imagery see action imagery multiple realizability 5, 146 N narcissism 161 narrative self 137 natural attitude 5, 98, 159, 168, 220 natural selection 4, 75–76, 87–94, 108, 128–129, 220 neuropeptides 26, 68, 107

neurotransmitters 15, 26, 43, 50, 63, 68, 72, 75, 84, 92, 106–110, 113, 124, 127, 141–142, 147, 152, 209, see also specific neurotransmitters, e.g., dopamine, acetylcholine, etc. noema 156 noesis 156 non-consummatory motivation 104ff. norepinephrine 75, 91–92, 109, 126, 198, 209 novelty 115ff. nurturance 5, 79, 81, 91, 96, 103–110, 119, 127–129, 210 O occipital lobe 12–14, 25, 50, 58–59, 63–64, 66–72, 74–76, 115, 143, 147–148, 169, 173, 207–208, 217 operant conditioning 15 opioids 107ff. orbitofrontal cortex 2 organism-environment system 20ff. oxytocin 107ff. P PAG see periaqueductal gray pain 35–36, 96, 98, 107–108, 119–121, 127, 132, 151, 182, 202 panic 122 parietal lobe 12–14, 63–68, 70–72, 143, 169, 173, 175, 207–208, 217–218 perception 167–176 and passim periaqueductal gray (PAG) 2, 5, 26, 63 107ff., 133–137, 147–148, 207–209 permanent vegetative states 11 personality 39, 130–143, 150–165, 183 personality change 39, see also personality phantom limb 3ff. phenomenal consciousness 99 phenomenological reduction 162 phenomenology 1–9, 25–42, 95–99, 128ff., 150–166, 176–181, 215ff.

Index 

play 3–5, 11, 14–15, 22, 27, 30, 41, 77, 79–81, 88, 90–93, 96, 103–112, 115–119, 123, 129, 210 pons 26, 107, 113, 209 pragmatic understanding 2 preconscious 17–18, 21, 33–45, 56, 63–64, 67–68, 155, 217 predatory behavior 117 predictions of enactivism 9–13 prefrontal cortex 116, 143, 147 premotor cortex 2, 13, 208 prereflective 36 primary emotions 15ff., 104 and passim priming 79ff. prolactin 107ff. psychophysical forms 99ff. psychophysical identity 16ff. purposeful intentions 140ff. Q qualia, emotional 20ff. R rage 107, 120, 129 reactive theories 14ff., 48–69, 163, 195, 221 readiness potential 144ff. reflection 157 reflective self 138ff. reinforcement 15, 85, 89, 93–94, 106–107, 110, 120 religion 124 respiratory energy cycle 94ff. S sadness 15, 31, 182, 187 satiation 15, 79–85, 88–91, 94, 96, 98–99, 101, 103, 108–111, 115, 120, 184–186 scales of organization 198 scapegoats 175 secondary emotions see “higher” emotions, primary emotions

seeking 5, 17, 52, 58, 60, 68, 90–93, 106–112, 115, 117–119, 124, 126–129, 146, 161–162, 196 seeking system see seeking selective attention versus arousal 75 self 131ff. self-actualization 96–97 self-for-others 160 self-organizing 4, 6, 8–9, 14, 16ff., 27–43, 51–65, 74–76, 79ff., 131ff., 167ff., 190ff. and passim self-representation 135 sensation 42–46, 57–62, 69–73 sense of self 133ff. sensorimotor action imagery 1–9, 42–46, 158 and passim sensorimotor versus proprioceptive imagery 9ff. and passim sensory areas 12, 68, 147 separation distress 15, 104, 107, 119, 122–123, 127–129 serotonin 75, 91, 95, 193–193, 200–201, 209, 212–213 sexual desire see lust sickness, feeling of 120 sleep versus waking 91 snakes, fear of 38 social bonding see bonding socialization 15, 22, 103 startle response 122 stimulus generalization 15 stimulus-response paradigm 16ff., 24ff. and passim stria terminalis 107 stroke recovery 202, 205 subatomic energy shells 85ff. subjectivity 131 Substance P 107 supplementary motor area (SMA) 2 symbolization-vehicles 33–41, 158–166, 167ff. T temporal lobe 12, 70–71, 143, 147, 207 thalamic reticular nucleus 75, 76

 Curious Emotions

thalamocortical loops 51, 113 thalamus 2, 29, 51, 64–67, 91, 107, 113–115, 143–147, 173, 207–208, 213 thermodynamic systems, open versus closed 196ff. thrill seeking see adventure top-down causal systems 193–207 and passim tragic paradox 182 transplantation of brain cells 205 triggers of emotion 33–41

V V-4 perceptual area 12 valence 35, 170 value-intensifying versus end-attaining motives 182ff. vasopressin 107ff. ventral tegmental area 107 vicarious pain 184 viscera see interoception voluntary action 142ff. W

U unconditioned emotions see primary emotions unconscious emotion 27–41, 158–166, 176–180 and passim

wellness, feeling of 120 Wernicke’s speech area 116 widely distributed brain activity 106, 142, 143 will 133

16, 74,

In the series Advances in Consciousness Research the following titles have been published thus far or are scheduled for publication: 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

HARDCASTLE, Valerie Gray: Locating Consciousness. 1995. xviii, 266 pp. STUBENBERG, Leopold: Consciousness and Qualia. 1998. x, 368 pp. GENNARO, Rocco J.: Consciousness and Self-Consciousness. A defense of the higher-order thought theory of consciousness. 1996. x, 220 pp. MAC CORMAC, Earl and Maxim I. STAMENOV (eds.): Fractals of Brain, Fractals of Mind. In search of a symmetry bond. 1996. x, 359 pp. GROSSENBACHER, Peter G. (ed.): Finding Consciousness in the Brain. A neurocognitive approach. 2001. xvi, 326 pp. Ó NUALLÁIN, Seán, Paul Mc KEVITT and Eoghan Mac AOGÁIN (eds.): Two Sciences of Mind. Readings in cognitive science and consciousness. 1997. xii, 490 pp. NEWTON, Natika: Foundations of Understanding. 1996. x, 211 pp. PYLKKÖ, Pauli: The Aconceptual Mind. Heideggerian themes in holistic naturalism. 1998. xxvi, 297 pp. STAMENOV, Maxim I. (ed.): Language Structure, Discourse and the Access to Consciousness. 1997. xii, 364 pp. VELMANS, Max (ed.): Investigating Phenomenal Consciousness. New methodologies and maps. 2000. xii, 381 pp. SHEETS-JOHNSTONE, Maxine: The Primacy of Movement. 1999. xxxiv, 584 pp. CHALLIS, Bradford H. and Boris M. VELICHKOVSKY (eds.): Stratification in Cognition and Consciousness. 1999. viii, 293 pp. ELLIS, Ralph D. and Natika NEWTON (eds.): The Caldron of Consciousness. Motivation, affect and self-organization — An anthology. 2000. xxii, 276 pp. HUTTO, Daniel D.: The Presence of Mind. 1999. xiv, 252 pp. PALMER, Gary B. and Debra J. OCCHI (eds.): Languages of Sentiment. Cultural constructions of emotional substrates. 1999. vi, 272 pp. DAUTENHAHN, Kerstin (ed.): Human Cognition and Social Agent Technology. 2000. xxiv, 448 pp. KUNZENDORF, Robert G. and Benjamin WALLACE (eds.): Individual Differences in Conscious Experience. 2000. xii, 412 pp. HUTTO, Daniel D.: Beyond Physicalism. 2000. xvi, 306 pp. ROSSETTI, Yves and Antti REVONSUO (eds.): Beyond Dissociation. Interaction between dissociated implicit and explicit processing. 2000. x, 372 pp. ZAHAVI, Dan (ed.): Exploring the Self. Philosophical and psychopathological perspectives on selfexperience. 2000. viii, 301 pp. ROVEE-COLLIER, Carolyn, Harlene HAYNE and Michael COLOMBO: The Development of Implicit and Explicit Memory. 2000. x, 324 pp. BACHMANN, Talis: Microgenetic Approach to the Conscious Mind. 2000. xiv, 300 pp. Ó NUALLÁIN, Seán (ed.): Spatial Cognition. Foundations and applications. 2000. xvi, 366 pp. GILLETT, Grant R. and John McMILLAN: Consciousness and Intentionality. 2001. x, 265 pp. ZACHAR, Peter: Psychological Concepts and Biological Psychiatry. A philosophical analysis. 2000. xx, 342 pp. VAN LOOCKE, Philip (ed.): The Physical Nature of Consciousness. 2001. viii, 321 pp. BROOK, Andrew and Richard C. DEVIDI (eds.): Self-Reference and Self-Awareness. 2001. viii, 277 pp. RAKOVER, Sam S. and Baruch CAHLON: Face Recognition. Cognitive and computational processes. 2001. x, 306 pp. VITIELLO, Giuseppe: My Double Unveiled. The dissipative quantum model of brain. 2001. xvi, 163 pp.

33 YASUE, Kunio, Mari JIBU and Tarcisio DELLA SENTA (eds.): No Matter, Never Mind. Proceedings of Toward a Science of Consciousness: Fundamental approaches, Tokyo 1999. 2002. xvi, 391 pp. 34 FETZER, James H. (ed.): Consciousness Evolving. 2002. xx, 253 pp. 35 Mc KEVITT, Paul, Seán Ó NUALLÁIN and Conn MULVIHILL (eds.): Language, Vision and Music. Selected papers from the 8th International Workshop on the Cognitive Science of Natural Language Processing, Galway, 1999. 2002. xii, 433 pp. 36 PERRY, Elaine, Heather ASHTON and Allan H. YOUNG (eds.): Neurochemistry of Consciousness. Neurotransmitters in mind. With a foreword by Susan Greenfield. 2002. xii, 344 pp. 37 PYLKKÄNEN, Paavo and Tere VADÉN (eds.): Dimensions of Conscious Experience. 2001. xiv, 209 pp. 38 SALZARULO, Piero and Gianluca FICCA (eds.): Awakening and Sleep–Wake Cycle Across Development. 2002. vi, 283 pp. 39 BARTSCH, Renate: Consciousness Emerging. The dynamics of perception, imagination, action, memory, thought, and language. 2002. x, 258 pp. 40 MANDLER, George: Consciousness Recovered. Psychological functions and origins of conscious thought. 2002. xii, 142 pp. 41 ALBERTAZZI, Liliana (ed.): Unfolding Perceptual Continua. 2002. vi, 296 pp. 42 STAMENOV, Maxim I. and Vittorio GALLESE (eds.): Mirror Neurons and the Evolution of Brain and Language. 2002. viii, 392 pp. 43 DEPRAZ, Nathalie, Francisco J. VARELA and Pierre VERMERSCH: On Becoming Aware. A pragmatics of experiencing. 2003. viii, 283 pp. 44 MOORE, Simon C. and Mike OAKSFORD (eds.): Emotional Cognition. From brain to behaviour. 2002. vi, 350 pp. 45 DOKIC, Jérôme and Joëlle PROUST (eds.): Simulation and Knowledge of Action. 2002. xxii, 271 pp. 46 MATEAS, Michael and Phoebe SENGERS (eds.): Narrative Intelligence. 2003. viii, 342 pp. 47 COOK, Norman D.: Tone of Voice and Mind. The connections between intonation, emotion, cognition and consciousness. 2002. x, 293 pp. 48 JIMÉNEZ, Luis (ed.): Attention and Implicit Learning. 2003. x, 385 pp. 49 OSAKA, Naoyuki (ed.): Neural Basis of Consciousness. 2003. viii, 227 pp. 50 GLOBUS, Gordon G.: Quantum Closures and Disclosures. Thinking-together postphenomenology and quantum brain dynamics. 2003. xxii, 200 pp. 51 DROEGE, Paula: Caging the Beast. A theory of sensory consciousness. 2003. x, 183 pp. 52 NORTHOFF, Georg: Philosophy of the Brain. The brain problem. 2004. x, 433 pp. 53 HATWELL, Yvette, Arlette STRERI and Edouard GENTAZ (eds.): Touching for Knowing. Cognitive psychology of haptic manual perception. 2003. x, 322 pp. 54 BEAUREGARD, Mario (ed.): Consciousness, Emotional Self-Regulation and the Brain. 2004. xii, 294 pp. 55 PERUZZI, Alberto (ed.): Mind and Causality. 2004. xiv, 235 pp. 56 GENNARO, Rocco J. (ed.): Higher-Order Theories of Consciousness. An Anthology. 2004. xii, 371 pp. 57 WILDGEN, Wolfgang: The Evolution of Human Language. Scenarios, principles, and cultural dynamics. 2004. xii, 240 pp. 58 GLOBUS, Gordon G., Karl H. PRIBRAM and Giuseppe VITIELLO (eds.): Brain and Being. At the boundary between science, philosophy, language and arts. 2004. xii, 350 pp. 59 ZAHAVI, Dan, Thor GRÜNBAUM and Josef PARNAS (eds.): The Structure and Development of Self-Consciousness. Interdisciplinary perspectives. 2004. xiv, 162 pp. 60 DIETRICH, Eric and Valerie Gray HARDCASTLE: Sisyphus’s Boulder. Consciousness and the limits of the knowable. 2005. xii, 136 pp. 61 ELLIS, Ralph D.: Curious Emotions. Roots of consciousness and personality in motivated action. 2005. vii, 238 pp. 62 DE PREESTER, Helena and Veroniek KNOCKAERT (eds.): Body Image and Body Schema. Interdisciplinary perspectives on the body. ix, 327 pp. + index. Expected July 2005 63 BARTSCH, Renate: Memory and Understanding. Concept formation in Proust’s A la recherche du temps perdu. x, 151 pp. + index. Expected July 2005