Philosophy, Psychology, and Psychologism: Critical and Historical Readings on the Psychological Turn in Philosophy

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

PHILOSOPHICAL STUDIES SERIES VOLUME 91

Founded by Wilfrid S. Sellars and Keith Lehrer

Editor Keith Lehrer, University of Arizona, Tucson Associate Editor Stewart Cohen, Arizona State University, Tempe Board of Consulting Editors Lynne Rudder Baker, University of Massachusetts at Amherst Radu Bogdan, Tulane University, New Orleans Marian David, University of Notre Dame Allan Gibbard, University of Michigan Denise Meyerson, Macquarie University François Recanati, Institut Jean-Nicod, EHESS, Paris Stuart Silvers, Clemson University Barry Smith, State University of New York at Buffalo Nicholas D. Smith, Lewis & Clark College

The titles published in this series are listed at the end of this volume.

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM Critical and Historical Readings on the Psychological Turn in Philosophy

Edited by

DALE JACQUETTE The Pennsylvania State University, University Park, PA, U.S.A.

KLUWER ACADEMIC PUBLISHERS NEW YORK, BOSTON, DORDRECHT, LONDON, MOSCOW

eBook ISBN: Print ISBN:

0-306-48134-0 1-4020-1337-X

©2003 Kluwer Academic Publishers New York, Boston, Dordrecht, London, Moscow Print ©2003 Kluwer Academic Publishers Dordrecht All rights reserved No part of this eBook may be reproduced or transmitted in any form or by any means, electronic, mechanical, recording, or otherwise, without written consent from the Publisher Created in the United States of America Visit Kluwer Online at: and Kluwer's eBookstore at:

http://kluweronline.com http://ebooks.kluweronline.com

That which renders Logic possible is the existence in our minds of general notions, — our ability to conceive of a class, and to designate its individual members by a common name...A successful attempt to express logical propositions by symbols...should be founded upon the laws of the mental processes which they represent... — George Boole, A Mathematical Analysis of Logic, 1847

No description of...mental processes which precede the forming of a judgement of number...can ever take the place of a genuine definition of the concept. It can never be adduced in proof of any proposition of arithmetic; it acquaints us with none of the properties of numbers. For number is no whit more an object of psychology or a product of mental processes than, let us say, the North Sea is. The objectivity of the North Sea is not affected by the fact that it is a matter of our arbitrary choice which part of all the water on the earth’s surface we mark off and elect to call the “North Sea”. This is no reason for deciding to investigate the North Sea by psychological methods. — Gottlob Frege, Grundlagen der Arithmetik, 1884

This page intentionally left blank

CONTENTS

PREFACE

ix

ACKNOWLEDGMENTS

xiii

INTRODUCTION: PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH Dale Jacquette PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO Rolf George

1 21

BETWEEN LEIBNIZ AND MILL: KANT’S LOGIC AND THE RHETORIC OF PSYCHOLOGISM Carl Posy

51

PSYCHOLOGISM AND NON-CLASSICAL APPROACHES IN TRADITIONAL LOGIC Werner Stelzner

81

THE CONCEPT OF ‘PSYCHOLOGISM’ IN FREGE AND HUSSERL J.N. Mohanty

113

PSYCHOLOGISM AND SOCIOLOGISM IN EARLY TWENTIETHCENTURY GERMAN-SPEAKING PHILOSOPHY 131 Martin Kusch THE SPACE OF SIGNS: C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM Vincent Colapietro

157

QUINEAN DREAMS OR, PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY Michael Bradie

181

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 195 Joseph Margolis vii

viii

CONTENTS

PROPOSITIONS AND THE OBJECTS OF THOUGHT Michael Jubien

215

THE CONCEPTS OF TRUTH AND KNOWLEDGE IN PSYCHOLOGISM John H. Dreher

229

PSYCHOLOGISM REVISITED IN LOGIC, METAPHYSICS, AND EPISTEMOLOGY Dale Jacquette

245

WHY THERE IS NOTHING RATHER THAN SOMETHING: QUINE ON BEHAVIORISM, MEANING, AND INDETERMINACY Paul A. Rhoth

263

COGNITIVE ILLUSIONS AND THE WELCOME PSYCHOLOGISM OF LOGICIST ARTIFICIAL INTELLIGENCE 289 Selmer Bringsjord and Yingrui Yang INDEX

313

PREFACE

Among the dichotomies that have divided philosophers, the rift between psychologism and antipsychologism represents some of the most heated metaphilosophical debate. The problem of whether and in what sense logic, mathematics, philosophical semantics, epistemology, and metaphysics are explanatorily related to psychology has been a fundamental watershed in the contemporary philosophy. The battlelines between psychologism and antipsychologism were first drawn in the mid-nineteenth-century. If logic, to take a conspicuous example, studies patterns of inference from thoughts to thoughts, then it has appeared to some theorists that logic is a branch of psychology that can best be understood in terms of the most advanced psychological science. Against this psychologistic view of logic, anti-psychologistic opponents have argued that logic is not a descriptive theory of how we actually think, but a prescriptive account of how ideally we ought to think. Logic on this conception is independent of the empirical facts of psychology. The inherently subjective nature of thought content appears diametrically opposed to the objectivity of the eternal truths of logic, and of philosophy of language and mathematics. To preserve the objectivity required of a rational a priori rather than empirical a posteriori science, antipsychologists have rejected the idea that philosophy is grounded in even the most rigorously scientific psychology. The psychologism-antipsychologism dispute can thus be interpreted as a deeper controversy about how philosophy can best be made scientific. There are two conflicting desiderata of science that provide a basis for the opposition between psychologism and antipsychologism. Science wants both to be objective and dependent on empirical facts. In physics, chemistry, biology, astronomy, and the other hard sciences, there is no collision of these values. It is peculiarly in the case of psychology, where the empirical facts of psychological experience have at least traditionally been regarded as essentially subjective, that a division has emerged between two opposed ways of trying to make logic and other philosophical subdisciplines (broadly, according to one ideology or another) ‘scientific’. The comparatively late development of psychology as a science as well as the subjectivity of psychological phenomena can be seen in this light as partly responsible for the dialectical confrontation between psychologism and

ix

x

PREFACE

antipsychologism. The two categories signify the legitimate but incompatible interests of these fundamentally irreconcilable requirements for a scientific psychology. If we could arrive at a satisfactory metaphysics of mind, then the apparently insurmountable impasse between psychologism and antipsychologism might simply disappear. Instead, we find only further manifestations of these two different ways of thinking about the empirical facts of subjective psychological occurrences reflected also in the philosophy of mind. Here they appear in longstanding oppositions between phenomenology and cognitive science, or between nonreductive intentionalist substance or property mind-body dualisms and eliminative or reductive behaviorism, materialism, functionalism, or computationalism in the cognitive psychological sciences. The disagreement over scientific ideals for psychology might be expected to fuel an inexhaustible dialectic between psychologism and antipsychologism. Such an interaction could provide the basis for a healthy and fruitful exchange in which competition from opposing sides could be harnassed for the sharpening of distinctions and refinement of arguments. To a limited extent, the opposition has continued and remains alive and well in the form of conflicts between realism and intuitionism or conceptualism, and between proponents and opponents of the program to naturalize or scientifically psychologize some of the traditionally nonpsychological philosophical disciplines like epistemology and metaphysics. In most ways, however, the psychologism-antipsychologism dispute has not exhibited this type of productive dialectical synergy. The rhetoric surrounding especially antipsychologistic philosophical discussions is revealing for its extraordinary degree of animus; it suggests the perception of a very ingrained division in outlook that cannot be overcome by a consideration of arguments with shared presuppositions, but that is directed polemically out of desperation at the presuppositions themselves. Psychologism has largely withered away under the criticism of historically influential antipsychologists. The objections have appeared both from within analytic and in the so-called continental schools. Among analytic philosophers, the most strident assault on psychologism originates principally with Gottlob Frege and his many followers, including Bertrand Russell, Ludwig Wittgenstein, Rudolf Carnap, and others; while in the nonanalytic European tradition, Edmund Husserl and Martin Heidegger are perhaps the most noteworthy ostensible antipsychologists. The friends of psychologism, whether or not they would be willing to identify themselves as such, have continued the struggle under a variety of different banners, which is itself an important feature of the rhetoric of psychologism and antipsychologism.

PREFACE

xi

To consider these problems, I invited a slate of distinguished scholars to present their perspectives on the history, philosophy, and rhetoric of psychologism. The papers with some overlap are presented roughly in historical sequence, by which the reader can trace certain themes through the development of the most significant episodes of the psychologismantipsychologism debate. The present collection of essays draws on three distinct sources of recent discussion of the philosophical problems of psychologism. The papers by Rolf George, Carl Posy, J.N. Mohanty, Joseph Margolis, and my Introduction were first published in a special issue of the journal Philosophy & Rhetoric, which I guest-edited in 1997, and which are reprinted here with the permission of Penn State University Press. Earlier versions of the essays by Michael Jubien, John H. Dreher, and myself were presented as feature contributions to an invited symposium on ‘Psychologism: The Current State of the Debate’ at the American Philosophical Association, Pacific Division, Albuquerque, NM, April 5-8, 2000. Finally, the essays by Werner Stelzner, Martin Kusch, Vincent Colapietro, Michael Bradie, Paul A. Roth, and Selmer Bringsjord and Yingrui Yang were specially commissioned for inclusion in this volume. Altogether, the expositions of critical and historical dimensions of psychologism offer a detailed picture of recent thinking about the problems and opportunities for philosophical understanding posed by various proposals for taking a psychological turn in philosophy. Dale Jacquette Mexico City - 21 December 2001

This page intentionally left blank

ACKNOWLEDGMENTS

I am grateful to the authors for their excellent contributions; to Rudolf Rijgersberg, my former editor at Kluwer, to my current editor Floor Oosting, and to Keith Lehrer, editor of the Philosophical Studies series, for supporting and encouraging my work on this project. I would like to thank Henry W. Johnstone, Jr., departed friend and founding editor of Philosophy & Rhetoric, for inviting me to guest-edit a special issue of the journal on The Dialectics of Psychologism, 30, No. 3, 1997. This volume constitutes the original nucleus of the present collection. The essays from that source, many with revisions, include: Dale Jacquette, ‘The Dialectics of Psychologism’, reprinted here as the Preface, v-viii; Rolf George, ‘Psychologism in Logic: Bacon to Bolzano’, 213-42; Carl J. Posy, ‘Between Leibniz and Mill: Kant’s Logic and the Rhetoric of Psychologism’, 243-70; J.N. Mohanty, ‘The Concept of “Psychologism” in Frege and Husserl’, 271-90; Joseph Margolis, ‘Late Forms of Psychologism and Antipsychologism’, 291-311; Dale Jacquette, ‘Psychologism the Philosophical Shibboleth’, 312-31. All are copyrighted 1997 by The Pennsylvania State University, and reproduced by permission of The Pennsylvania State University Press. Michael Jubien’s essay, ‘Propositions and the Objects of Thought’ is reprinted with the permission of Kluwer Academic Publishers from Philosophical Studies, 104, 2001, 47-62. Martin Kusch’s essay, ‘Psychologism and Sociologism in Early Twentieth-Century German-Speaking Philosophy’ is reprinted with the permission of Elsevier Science from Studies in History and Philosophy of Science, Vol. 30, No. 4, 1999, 651-85, where it originally appeared under the title, ‘Philosophy and the Sociology of Knowledge’. Dale Jacquette’s essay, ‘Psychologism Revisited in Logic, Metaphysics, and Epistemology’, is reprinted with the permission of Blackwell Publishers from Metaphilosophy, 32, 2001, 261-78. I wish to thank Scott K. Templeton and the graphics staff at The Onion for invaluable technical assistance. This book is dedicated to Tina with love.

xiii

This page intentionally left blank

DALE JACQUETTE

INTRODUCTION PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH The charge of psychologism has been made against my theory of knowledge. This is a word which has lately come into use and when it is spoken many a pious philosopher — like many an orthodox Catholic when he hears the term Modernism — crosses himself as though the devil himself were in it. — Franz Brentano, The Classification of Mental Phenomena (1911) Appendix XI, ‘On Psychologism’

1. A Psychologism Miscellany All theory is the product of thought, even when its subject matter is nonpsychological. Psychology, moreover, is widely recognized as a legitimate field of study for science and philosophy. Why then has ‘psychologism’ become the watchword of a fundamental antagonism from nineteenth-century to contemporary philosophy, particularly in logic, semantics, philosophy of language and philosophy of mathematics, but also in metaphysics and epistemology? There are many kinds of psychologism, and many ways of defining the concept of psychologism. Martin Kusch, in his recent book, Psychologism: A Case Study in the Sociology of Philosophical Knowledge, offers a taxonomy of distinct schools of psychologism in German philosophy between 1866 and 1931, of which additional variations can also be identified. What can so many different versions of psychologism have in common? To further complicate the picture, Kusch observes that most of the widely recognized critics of psychologism were themselves accused of being psychologistic by antipsychologistic opponents who were even more zealous in their intolerance. Why has ‘psychologism’ become such a dirty word that even avowed antipsychologists are subject to its anathema? As the name suggests, psychologism is not a branch of psychological science, but a philosophical ideology based on psychology. More 1 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 1-19. © 2003 All rights reserved. Printed by Kluwer Academic Publishers, the Netherlands.

2

DALE JACQUETTE

particularly, psychologism includes any attempt to ground philosophical explanation in psychological phenomena. Psychologism is a family of proposals for invoking different aspects of psychological occurrences in different ways to develop different styles of philosophical theory. The objections if not the vehemence with which antipsychologists frequently raise objections against psychologism can generally be attributed to the assumption that an empirical psychology of subjective thought cannot be expected to explain logically necessary objective truths, especially those of logic, semantics, and mathematics, but also of any field of discourse where a sharp distinction is supposed to hold between objective truths and subjective perceptions of the truth. Kusch describes the psychologism flora and fauna in the particular period in which he is interested, when he writes: An initial perspective on the inflation that the term ‘psychologism’ underwent between 1900 and 1930 can be gained by examining the (grammatical) attributes with which the term occurred. First of all, writers distinguished between different forms of psychologism according to the fields of philosophy and the human sciences in which psychologism needs to be combated. Thus one finds ‘psychologism’ qualified as ‘metaphysical’, ‘ontologica’, ‘epistemological’, ‘logical’, ‘ethical’, ‘aesthetic’, ‘sociological’, ‘religious’, ‘historical’, ‘mathematical’, ‘pedagogical’ and ‘linguistic’. Second, psychologism was also broken down into species according to the distinctive versions of psychologism that various schools were accused of proposing. Such adjectives included ‘empiricist’, ‘aprioristic’, ‘sensualist’, ‘rationalist’, ‘critical-teleological’, ‘evolutionary’, ‘pragmatist’ and ‘transcendental’. Third, versions of psychologism were distinguished on the basis of their age, their ‘degree of truth’ and the boldness with which they were supposedly put forward. That is to say, ‘psychologism’ could be ‘old’, ‘new’, ‘false’, ‘true’, ‘objective (intersubjective)’, ‘justified’ (wohlverstanden), ‘onesided/tendentious’, ‘extreme’, ‘moderate’, ‘universal’, ‘open’, ‘hidden’, ‘inverse’, ‘obvious’, ‘delicate’, ‘strict’ and ‘loose’. And finally, we also find distinctions between ‘intellectual’ and ‘emotional’ psychologism, as well as between ‘immanent’ and ‘transcendent’ psychologism. (108)

2. The Rhetoric of Antipsychologism The rise of antipsychologism is a chapter of philosophical rhetoric. Antipsychologism has had an enormous impact on the practice of philosophy and on philosophy’s self-image of its proper method and direction, and on its developing sense of what is to count as legitimate philosophical inquiry. Thus, in his recent study of Wittgenstein’s Place in Twentieth-Century Analytic Philosophy, P.M.S. Hacker explains that: “...twentieth-century analytic philosophy is distinguished in its origins by its

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

3

non-psychological orientation” (4). The rhetorical dimensions of antipsychologism are evidenced by the language of the arguments in which objections to psychologism are typically expressed. Although there are respectable philosophical criticisms of psychologism, there is also a large percentage of unreflective psychologismbashing in the rejection of psychologism without thorough-going criticism. Indeed, the reasons for rejecting psychologism often encountered in contemporary philosophical literature often amount to little more than an appeal to the authority of prominent antipsychologists in the history of philosophy who with or without good justification themselves have taken an influential stand against psychologism. Brentano, ex-priest and philosopher, does not exaggerate when he says that in his day too the philosophical badmouthing of psychologism had virtually assumed the proportions of religious zeal. It is not enough to discredit psychologism like any mistaken theory by demonstrating its disadvantages — we must exorcise psychologism from philosophy, hold it off with garlic, shoot it with a silver bullet and drive a stake through its heart. One of the most remarkable statements of antipsychologism appears in Arthur Pap’s influential study, Semantics and Necessary Truth: An Inquiry into the Foundations ofAnalytic Philosophy. What is especially noteworthy is that the statement occurs in the Glossary appended to Pap’s book as a definition of psychologism. Pap writes of “Psychologism” that it is “the tendency to confuse logical issues with psychological issues; e.g. if one tried to answer a question of logical validity by investigating actual beliefs (however, the meaning of this deprecatory word is unclear to the extent that the meaning of ‘logical’ is unclear)” (435). By definition, then, psychologism for Pap is a ‘confusion’ and the word alone is ‘deprecatory’. So it is no surprise to find Pap speaking negatively of psychologism throughout the body of the book. In an early section, he writes: “If [Immanuel] Kant’s conception of analyticity, then, is to be condemned as ‘psychologistic’, at least he will enjoy the company of many subtle contemporary analysts of reputation” (30). And near the end of the book, he similarly states: “Thus the old criticism, going back to presemantic days, of the Kantian analytic-synthetic distinction...as being relative to psychological conditions, cannot be simply exorcized as a symptom of ‘psychologism’” (246). Alongside Pap’s antipsychologistic propoganda, we may consider Herbert Feigl’s pronouncement that: “Ever since [Gottlob] Frege’s and [Edmund] Husserl’s devastating critiques of psychologism, philosophers should know better than to attempt to reduce normative to factual categories. It is one thing to describe the actual regularities of thought or language; it is an entirely different sort of thing to state the rules to which thinking or speaking ought to conform” (1963, 250). Alan Musgrave in a similar vein remarks that: “Nowadays only a few cranks officially subscribe to that view

4

DALE JACQUETTE

[psychologism]...There is progress in philosophy after all!” (1972, 593, 606). A final epitaph among indefinitely many that might be added is given by Gerhard Radnitzky when he relegates psychologism to the dustbin of ideas that have once and for all been refuted: “Thanks largely to the pioneering work of Frege and Husserl, psychologism in logic and metamathematics is largely a thing of the past: the attempt to reduce the norms of logic to laws of thought is now merely a historical curiosity” (1976, 505).1 In the semantic McCarthyite antipsychologistic hysteria of the times, there was no protection in declaring oneself a card-carrying antipsychologist by discovering and denouncing other thinkers as guilty of psychologism. The leading antipsychologists were themselves pilloried in print for latent psychologistic tendencies. Kusch quotes Rudolf Eisler in his 1907 Einführung in die Erkenntnistheorie, as saying: “...there is hardly an antipsychologist, from Kant up to [Hermann] Cohen and Husserl, who has not been accused of ‘psychologism’ by some even more extreme thinker. In the end one no longer knows just who is a psychologicist and who isn’t” (1907, 19; translated in Kusch 116).2 Later Kusch recalls that: Authors who were willing to regard themselves as ‘psychologists’, however, formed but a small minority in German-language philosophy between 1900 and 1930. As indicated earlier, most philosophers regarded psychologism as a gross philosophical error that needed to be ruthlessly identified in the thought of their contemporaries. We have already seen how all philosophical schools participated in this merry-go-round of charge and countercharge, and how practically every single German philosopher, dead or alive, was unmasked as a proponent of psychologism. Not surprisingly, this merry-go-round was possible only because the criteria for attributing a psychologistic stance to another philosopher were extremely flexible. While the different schools agreed on the fact that psychologism entailed a mistaken grounding of philosophy in psychology, they disagreed sharply as to what constituted such grounding. (115)

The antipsychologistic polemic is explicit in these objections. All the elements are present. There is an appeal to authority in the person of eminent antipsychologistic thinkers as having definitively exposed psychologism as a ‘confusion’ and ‘symptom’ of faulty reasoning, that has contributing to its becoming a mere ‘historical curiosity’. Other writers speak of theories that are ‘guilty’ of the ‘sin’ of psychologism, which they may ‘commit’. An equally significant mark of the pervasiveness of antipsychologism is the accusation of psychologism with no accompanying perjorative as condemnatory in and of itself. It is as though to classify a theory as psychologistic is enough automatically to disprove it, as though everyone already understands and agrees that psychologism is a defect or mistake, and that we should all be grateful for its having been thoroughly exploded in a heroic past age of philosophy.

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

5

The rhetoric of antipsychologism must further take account of the fact that in some philosophical circles ‘psychologism’ unalloyed has become a term of abuse, that deep-rooted hostility against psychologism, and even against the name ‘psychologism’ when applied to theories that many otherwise ardent antipsychologists find acceptable. Kusch explains that: “Aware of the danger that whatever they wrote on logic and epistemology would immediately be accused of the taint of psychologism, some philosophers tried to anticipate such criticism and to counter it by rhetorical means. [Anton] Marty went for an argumentum ad misericordiam by drawing attention to the fact that characterising a philosophical position as ‘psychologistic’ was to label it in a ‘disparaging’ way...” (117). The rhetoric directed against psychologism is interesting among other reasons for its precedent of substituting a mixture of appeal to authority and invective in place of open-minded philosophical argument. The very term ‘psychologism’ has become a philosophical shibboleth, an expression by which warring factions identify comrades and distinguish foes, like the ancient Israelites when they asked unfamiliar persons they encountered to pronounce this harmless word for a sheaf of grain.

3. Psychologism in a New Key To give a more balanced picture of the conflict, it should also be observed that some commentators have made responsible unbiased critical assessments of psychologism. Among the most important examples must be included Kusch’s sociological study of Psychologism, and the philosophical appraisals offered in Michael Dummett’s The Interpretation of Frege’s Philosophy and G.P. Baker and P.M.S. Hacker’s Frege: Logical Excavations. Kusch, for example, though not writing as a logician or philosopher, and targeting especially the impact of antipsychologism on the politics of German universities in the hiring practices of philosophy and psychology departments, reports that: “...I am more impressed by the force of [Brian] Ellis’s, [Theodor] Lipps’s and [Moritz] Schlick’s defences of psychologism, or by [Wilhelm] Jerusalem’s psychologism-cum-sociologism, than by Husserl’s or Frege’s attacks” (275). Baker and Hacker note the rhetorical overtones by which Frege tries to compensate for a distinct lack of argument against certain of the perceived implications of psychologism. They write that: Frege’s most general criticisms spill over into an onslaught upon idealism in general. Although his conclusions are unexceptionable, the supporting arguments are shallow, akin to a Johnsonian foray into the refutation of idealism. [Frege] found [the idealism implied by psychologism] absurd. No doubt it is. But his tirade simply relies on the assumption that there is

6

DALE JACQUETTE accessible to us an objective domain of mind-independent entities. (50-1)

After a painstaking investigation of Frege’s objections to psychologism, in the final paragraph of their chapter on ‘Psychologism and the Theory of Content’, Baker and Hacker maintain: The distinctive amalgam of Cartesianism and Platonism in Frege’s thinking must characterize any reasoning that could be called an extention [sic] of his arguments against psychologism. It would also deprive any such reasoning of real philosophical value. Both the Cartesian myths about the realm of the psychological and the Platonic myths about the realm of the logical generate deep philosophical confusions, and hence these central ingredients of Frege’s thinking are themselves in dire need of philosophical investigation and clarification. The only conceivable sources of light must be alien to his framework of thought. Consequently, we must conclude that Frege’s crusade against the incursions of psychology into logic is now largely obsolete. His way of drawing the distinction between logic and psychology is mistaken in detail and dangerous in its wider implications. Only somebody who shares a large measure of his Cartesian and Platonist mythology will find any seeds of the Tree of Knowledge scattered in his antipsychologistic polemic. (62)

Other assessments of psychologism and antipsychologism that do not rely on ad verecundium or diatribe, and that do not necessarily fall in lockstep with the predominant antipsychologistic mainstream opinion in the philosophy of logic, semantics, and mathematics, might easily be multiplied. Moreover, there is a countermovement that has sought to reinstate a modified concept of psychologism in contemporary philosophy. This has taken the form of an effort to naturalize epistemology and metaphysics in somewhat the fashion of the British empiricists John Locke, George Berkeley, David Hume, and John Stuart Mill, but recast in the vocabulary and involving the new methods and discoveries of cognitive psychological science. Analytic philosophy today is not averse to all attempts to incorporate the findings of empirical psychology. W.V.O. Quine is conspicuously in the vanguard of recent philosophical projects to introduce scientific psychology into philosophical epistemology. In his 1969 essay, “Epistemology Naturalized”, the original subtitle of which was “Or, the Case for Psychologism”, Quine observes that: “The old epistemology aspired to contain, in a sense, natural science; it would construct it somehow from sense data. Epistemology in its new setting, conversely, is contained in natural science, as a chapter of psychology” (83). Related philosophical proposals that seek to make psychology the foundation of traditionally purely rationalist a priori subdisciplines are indicated by the titles of such works as Jerry A. Fodor’s Psychosemantics and Patricia Smith Churchland’s Neurophilosophy.3 Dummett blames Frege and Husserl for not distinguishing sufficiently clearly between psychology and logic in such a way as to preclude the recent upsurge of neopsychologisms, when he argues:

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

7

Where both [Frege and Husserl] failed was in demarcating logical notions too strictly from psychological ones...These failings have left philosophy open to a renewed incursion from psychology, under the banner of ‘cognitive science’. The strategies of defence employed by Husserl and Frege will no longer serve: the invaders can be repelled only by correcting the failings of the positive theories of those two pioneers. (1991, 287)

Whether Frege and Husserl could or would have wanted to stave off the onslaught of cognitive psychology into the domain of philosophy, there have all along been psychologistic strains in many phases of modern philosophy’s development, from the late nineteenth century to the present day. In logic and mathematics, psychologism changed its name to intuitionism and constructivism, but continued to assume that thought processes and the limitations of ideas are the only basis for understanding logical and mathematical relations. In ontology and the metaphysics of universals, psychologism persisted in a different form under the less inflammatory but equally psychologically oriented label of conceptualism. In moral philosophy, value theory, and aesthetics, a kind of psychologism developed in different guises as naturalism, intentionalism, consequentialism, or emotivism. The distinction between psychologism and antipsychologism in its most general terms demarcates such a far-reaching schism that it is only reasonable to expect it to endure throughout the most challenging philosophical revolutions and intellectual upheavals. As Ernst Cassirer remarks: “...psychologism...still cannot be regarded as defeated. For although its form and justification have changed since Husserl’s sharp and trenchant criticism, we must note that psychologism has, to a high degree, the ability to appear in ever new guises” (1927, 32; translated in Kusch, 121). Quine’s suggestion that epistemology be naturalized was fully prefigured in his earlier philosophical semantics. In Word and Object, Quine develops a dispositional behaviorism of stimulus meanings by which language gains meaning in practical activity, without invoking abstract ‘thoughts’ or propositions like Fregean Gedanken, or objective entirely mind-independent Sätze an sich, as Bernard Bolzano advocated in his 1837 Wissenschaftslehre.4 The dialectic of psychologism and antipsychologism in this broad sense continues to transform itself in different ways, holding open the radical choice between including or excluding the content of thoughts and the empirical facts of psychology as a touchstone for philosophy. Psychology in the form of extratheoretical appeal to mental content, or as phenomenology, behaviorism, cognitive science, or Freudian psychoanalytic theory, stands ready to make different types of contributions to different types of philosophical explanation. The peaceful coexistence of new varieties of psychologism with the antipsychological heritage of Frege and Husserl in the current analytic philosophical climate is largely a result of two considerations. The first

8

DALE JACQUETTE

mitigating factor is that the term ‘psychologism’ by common consent has for the most part been consigned to quixotic attempts to explain the necessary truths of logic, semantics, and mathematics on the concrete empirical facts of subjective thought and mental content. In this application, psychologism remains as suspicious to most mainstream philosophers as it was for Frege, Russell, Wittgenstein, and their followers, and might better be called ‘mentalism’ than ‘psychologism’. The second mitigating factor is that the specific kinds of psychologism that have gained respectability in contemporary analytic philosophy are not forms of mentalism and are not supposed to explain the objective truths of logic and mathematics. At most, psychology is lately invoked to account for semantics in the sense of understanding the contextual pragmatics of meaning in language use for speech act theory, and of some of the softer topics in the philosophical gamut such as epistemology, as in Quine’s proposal for using cognitive psychological science to explain how epistemic subjects go about discovering and justifying some of their beliefs as knowledge. When Quine in Philosophy of Logic, for example, analyzes the logical truth of a sentence, he does so in terms of objective truth conditions and the preservation of truth under any uniform substitution of its extralogical terms. He says that: “Our new definition of logical truth, then, can also be put thus: a logical truth is a sentence that cannot be turned false by substituting for lexicon. When for its lexical elements we substitute any other strings belonging to the same grammatical categories, the resulting sentence is true” (58). When he speaks of naturalizing epistemology by making it a part of psychology, however, Quine limits his proposal to natural science: “We are studying how the human subject of our study posits bodies and projects his physics from his date, and we appreciate that our position in the world is just like his. Our very epistemological enterprise, therefore, and the psychology wherein it is a component chapter, and the whole of natural science wherein psychology is a component book — all this is our own construction or projection from stimulations like those we were meting out to our epistemological subject. There is thus reciprocal containment, though containment in different senses: epistemology in natural science and natural science in epistemology” (1969, 83). All this complicates the problem of psychologism. The rhetoric of antipsychologism is filtered through many layers of many different thinker’s understanding of different meanings of the word, shaped by their opinions about whether any particular form of psychologism is acceptable or unacceptable when applied to distinct kinds of philosophical subjects. To gain a better understanding of the complex rhetorical situation in the psychologism-antipsychologism dispute, we shall try next briefly to assess the most important philosophical arguments that have been raised against psychologism.

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

9

4. Arguments Against Psychologism...and Some Replies In this section, I shall not try to deal exhaustively or even especially systematically with objections to psychologism. My purpose is only to say enough to make it plausible to conclude that there may be more to the psychologism-antipsychologism controversy than has often been appreciated. I want to show that the dispute is not so one-sidedly in favor of antipsychologism as many antipsychologists have uncritically assumed. The decisive refutation of psychologism as an approach to the philosophy of logic, philosophical semantics of meaning, and philosophy of mathematics in my view has yet to be given. If my counterarguments are correct, then the debate will have to proceed along different lines and by means of more carefully elaborated arguments than those that I consider. I shall review without detailed historical scholarly apparatus the principal criticisms that have appeared against psychologism in the philosophical literature. I argue that all of the problems said to be entailed by psychologism can be related in different ways to philosophical disapproval of the subjectivity of thought as a foundation for objective scientific theory. Then I indicate some of the considerations by which problems about the subjectivity of thought can be disarmed, so that a more thorough reconsideration of the psychologism-antipsychologism imbroglio may be warranted in which the refutation of psychologism is no longer presupposed. Here are eight related arguments against psychologism. I have arranged them approximately according to my perception of their increasing strength, from less to greater difficulty as objections to the theory. In each case, I give a succinct statement of the criticism, explain how it is related to the general question of the subjectivity of psychological occurrences, and offer a brief reply in support of my claim that not every version of psychologism is necessarily the lost cause its opponents have polemically portrayed it as being. For simplicity, I limit consideration to arguments against efforts to reduce logic to psychology, with the understanding that similar criticisms with appropriate adjustments might be offered at least against semantics and mathematics, and philosophy of language and philosophy of mathematics. All of the arguments attempt to identify some essential difference between logic and psychology that prevents logic from being reducible to psychology.5 Argument 1

Logic is exact; psychology is inexact. The criticism is that logic cannot be reduced to psychology because of an inherent difference in their respective attainable degrees of precision. The argument directly reflects the fundamental opposition that is supposed to obtain between the objective

10

DALE JACQUETTE

eternal truths of logic and the subjective contingent truths of psychology. An objection of this sort may have held more sway in the days of psychology’s infancy, when there seemed little prospect of making psychology into an exact science. The terminology, experimental methods, and explanation of empirical discoveries in contemporary psychological science are in some ways as exact as in physics or chemistry. The conclusions psychologists reach may not be as universal as in the harder physical sciences, but this has nothing immediately to do with the question of their relative exactitude. If psychological laws are only statistical, that in principle does not place them at any disadvantage in comparison with the statistical laws of quantum physics. Statistics, moreover, is every bit as formally mathematically precise as the most rigorously developed formal symbolic logic. If we turn from cognitive science to phenomenology, we can interpret the efforts of philosophers like Husserl and Maurice MerleauPonty to achieve a philosophical theory of mind that, if successful, would also be as exact as symbolic logic, albeit according to a different paradigm of science with a different set of requirements for exactitude. Equivalent exactitude in any case is not a prerequisite for reducing one science to another. Biology is inexact when compared to chemistry. It is unclear whether the bacteriophages studied in biology are or are not living things, but it would be hard to find a comparable gray area of categories in chemistry. Yet most theorists agree that biology is fully reducible to chemistry. Nor does logic as it is understood today compare favorably as against psychology on other metatheoretical grounds. Logic is deductively incomplete according to proofs by Kurt Gödel and Alonzo Church, and there are undecided and even undecidable questions of modern logic, just as there are unsettled questions and unproven and undisproven hypotheses in psychology, as in other exact sciences at an early stage of their development.6 Finally, the question of just how exact logic is, and the benchmark of precision it affords, is softened by the fact that the exactness of logic is itself a quality judged by thought, and to that extent can only reflect the exactness of some psychological occurrences that must translate at least in principle into a corresponding exact psychological theory that must somehow be adequate to the explanation of such exact psychological phenomena. Argument 2 Logic is a priori; psychology is a posteriori. The empirical nature of psychology by contrast with the necessary truths of logic is sometimes offered as a decisive refutation of psychologism. Problems of exactness aside, if logic concerns a body of necessary judgments whose proof is independent of experience, and if psychology is or is about a body of logically contingent empirical judgments whose justification requires

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

11

experience, then there might be no prospect for reducing logic to psychology. This objection is indirectly related to the subjectivity of psychology as opposed to the objectivity of logic, because it is the subjectivity of thought by which it is immediately accessible only to the first-person subject of experience that entails that psychology unlike logic can only be known a posteriori. It is customary in explaining this type of objection to distinguish between contexts of concept acquisition and justification, and to admit that while all concepts might be acquired only a posteriori, concepts like those of logic do not stand in need of experience a posteriori for their justification. As with other criticisms of psychologism, the question of whether or not there is a difference between logic and psychology on the grounds of aprioricity or aposterioricity can be regarded as circular. If logic is part of or reducible to psychology, and if logic is a priori, then part of psychology is also a priori. The distinction between contexts of concept acquisition and of concept justification can further be invoked in defense of certain versions of psychologism. It is true that psychology is an empirical science that proceeds among other ways by an a posteriori method of hypothesis testing and confirming or disconfirming hypotheses through observation and experiment. This is not to say that psychology cannot acquire or discover by a posteriori investigations concepts that are only justified a posteriori. The problem of whether logic can be reduced to psychology if psychology as an empirical science does not provide a priori justifications again begs the question of whether psychology is a purely empirical discipline that does not contain logic with its a priori justifications of logical truths as one part of its formal theory of the workings of the mind.7 Argument 3 Logic is prescriptive; psychology is descriptive. This argument seeks to establish a stronger distinction between logic and psychology. Psychology as an empirical science describes psychological occurrences. Logic is different because it does not merely describe how we reason, but prescribes standards for correct thinking, or of how we should or ought to reason. Again, whether or not psychology is purely descriptive depends on whether or not we have good independent reasons for deciding whether or not logic is a part of psychology. If logic is a part of psychology, and if logic is prescriptive, then psychology is also not purely descriptive, but partly prescriptive. Still, it might be said that psychology as a science cannot imply standards for correct reasoning. There is an implicit complaint against the subjectivity of psychology in this objection, in the idea that psychology can only describe the phenomenological contents of individual minds. Needless to say, this is just one way of thinking about psychology. But it also appears in the case of more ‘scientific’ cognitive psychology, behaviorism,

12

DALE JACQUETTE

neuropsychology, and the information sciences that such disciplines can at best teach us how we think, whereas logic looks very much to be the normative set of guidelines for how we ought to think.8 By way of reply, it is worth remarking first that phenomenological and scientific psychology are not purely descriptive, but, like other sciences, they try to explain the phenomena by adducing lawlike regularities and general principles or laws. Insofar as psychology succeeds in advancing general principles, it is more like traditional logic than a purely descriptive science such as morphology and taxonomy. It is standard to maintain that we should expect a complete psychology but not a complete logic to include all the errors of reasoning a subject happens to make. Yet the traditional study of logic also includes discussion of invalid inferences and formal and rhetorical fallacies. Moreover, logic can be understood as descriptive of how some reasoning occurs, at the very least of the reasoning of certain logicians. There is no need to deny the hypothetical imperative or prescriptive element of logic as it is usually presented. If we want to reason correctly, then we ought to accept and to make our reasoning conform, say, to modus ponendo ponens, and to reject and avoid the fallacy of affirming the consequent. There are similar marriages between pure science and recommendations for practice in other areas of knowledge that pose no obstacle to theoretical reduction. Engineering and medicine are two obvious examples. In engineering one learns how to build a bridge ‘correctly’ so that it will bear its load properly, withstand windsheer and other kinds of stress, minimize metal fatigue and the like, just as in logic one learns how to reason ‘correctly’ to draw sound and avoid fallacious inferences. Similarly in medicine. When we learn to build bridges or perform heart surgery or reason correctly, we also learn what not to do relative to presumed practical purposes. We want the bridge to safely span the gorge and not to collapse, the patient to recover with improved health and not die on the operating table, and our reasoning to expand our knowledge and improve our decisionmaking ability, and not to lead us from truth to falsehood. Now I think that most theorists would be reluctant to conclude that engineering and medicine are not reducible to physics and biology, but rather that their respective practices combine these sciences with an assumption about the goals they can be assumed to help their practitioners achieve. Why should things be different with respect to logic as an applied discipline grounded in the psychology of reasoning? Argument 4 Logic is universal; (human) psychology is (human) species-specific. The attempt to distinguish logic from psychology by discerning a relevant species-relative difference between them is perhaps clear enough. The objectivity of logic and subjectivity of psychology is explicit at the

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

13

differential species level. The intent of the argument is more difficult to assess. What is it supposed to mean to contrast the universality of logic with the human species specificity of psychology? Does psychology apply only to humans, while logic is the same for humans, apes, and giraffes — or perhaps for angels and extraterrestrials? Animal psychology is of course a legitimate subdiscipline of contemporary scientific psychology, and much that we have come to know about human psychology depends on or was originally suggested by observation and experiment involving nonhuman animals, some of which concerns nonhuman problem solving and extralinguistic reasoning abilities. It is not clear furthermore whether and to what extent our standard logics can be regarded as appropriate for other species who do not have and are unable to use our or any other language. The situation in logic today in any case is rather different than in the heyday of antipsychologism. There is no longer thought to be just one monolithic universal logic that is assumed to be correct for all reasoning. Rather, logicians have developed families of different logics, some of which are not even mutually compatible, that logicians tend to regard as individually suited to particular kinds of reasoning about particular kinds of topics. There are Boolean and non-Boolean logics, extensional and a variety of intensional logics, modal logics, deontic logics, bivalent and many-valued logics, and so on. In this regard, logic has followed the proliferation of nonEuclidean geometries like the Riemannian and Lobachevskian, where in the past the parallel assumption was equivalently that there could be only one ‘correct’ geometry. It would not shock most logicians or philosophers of logic nowadays, if alien beings were discovered who did not use any of our logics but had a different and possibly more advanced logic distinct from any that we have used. The aliens’ unique reasoning might still recognizably constitute a logic, that we might learn about only by studying their psychology expressed in their language and behavior, as a reflection of their nonhuman psychology. Argument 5 Logic is discovered, not invented; logic is therefore presupposed by its discovery in thought, contrary to the claims of psychologism. The objectivity of logic as something already existing to be discovered as opposed to the subjective invention of logical systems to express the eternal truths of logic is obvious. I think, nevertheless, that the presupposition of the objection is as questionable as it is question-begging. It is not clear in light of the many different systems of logic available that logic is not a product of certain kinds of thought rather than a body of necessary truth waiting to be discovered by and expressed in thought. This indeed is one of the most intriguing problems for an undogmatic philosophy of logic to investigate.

14

DALE JACQUETTE

Regardless of how opinions may vary on this issue, the criticism does not establish any relevant distinction by which logic must be independent of psychology. The reason is that the truths of psychology themselves are not invented, but discovered by the honest hardworking efforts of the phenomenologist or more conventionally scientific psychologist. Psychology as much as logic is the objective study of thought, wherein logic as a part of psychology, for all that this objection has to say, might well be the objective study of particular kinds or particular aspects of thought. Argument 6 Logic is presupposed by any theory, including psychology, rather than the other way around, so that logic cannot be reduced to psychology. Here the difference between logic and psychology is supposed to be that logic is objective in a very ultimate sense, as preceding other disciplines, including psychology. The idea that logic must come first as a kind of queen of the sciences is in many ways attractive. But if we take seriously and openmindedly the suggestion that logic might be a part of psychology, then the only conclusion to follow is that one part of psychology must come before the other parts, or that there must be a kind of priority in the order by which the principles of psychology are established. An analogy with chemistry serves to demonstrate a similar lesson. If we think of chemistry primarily as the science of molecular behavior, then we cannot regard atomic theory as part of chemistry, since molecules ontically presuppose atoms. One might wrongly suppose that chemistry for this reason cannot subsume atomic theory because atomic theory is presupposed by chemistry. But if atomic theory is rightly understood as a part of chemistry, then chemistry as a whole will meet its explanatory burden, provided that the atomic theory it contains first explains the chemistry of atomic structures, on which molecular chemistry can build its theory of more complex substances. The same relation might then hold of logic subsumed by psychology. Logic articulates the principles by which sound reasoning takes place, by which means the particular types of reasoning presupposed by the development of the remaining parts of psychology among other theories is provided. Argument 7 Logic concerns identical objects of distinct thoughts, and so cannot be reduced to the contents of individual psychologies. This is a very popular objection to psychologism. If psychologism were true, and logical entities were merely psychological, then any two different psychological subjects would immanently possess distinct logical entities by virtue of having different thoughts. The subjectivity of psychology by this argument entails

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

15

that psychologism leads immediately to idealism, relativism, and even solipsism or skepticism about the existence of other minds. The problem is most easily illustrated in the case of numbers. If numbers are immanently psychological entities, then there is no single entity such as the number 2 for two different psychological subjects to think about — I will have the number 2 that belongs to my thought, and you will have a different number 2 that belongs to your thought. This implication seems untenable, because we suppose that logic and mathematics has to do with entities and relations that are independent of any particular subject’s thought, but that can be thought about as the same things by many different thinkers. The idealism and relativism that might be thought to follow from psychologism according to the criticism are evident in this description. Solipsism may further be implied by the fact that if even abstract logical and mathematical entities are subjective, then the privacy and epistemic inaccessibility of the contents of thought by which one mind is sealed off from other minds, then there seems to be nothing outside the mind on which there can be a definite meeting of different minds, and thus no justification for belief in the existence of other minds. Perhaps the best way to meet the objection is firmly to renounce any formulation of psychologism with such idealistic implications. To suppose that logic can be explained in psychological terms is not necessary to try to reduce logical entities to psychological entities. A thought about the Statue of Liberty does not make the Statue of Liberty a mental entity, and a psychological theory that explains the workings of perception need not entail that the objects encountered in perception are subjectively psychological entities. Why then should a psychological theory of logic entail that logical entities are subjectively psychological entities? Argument 8 Logic is objective; psychology is subjective. The claim that logic cannot be reduced to psychology because logic is objective and psychology is subjective is finally the simplest and most direct statement of the underlying criticism that in different forms drives the rhetoric of antipsychologism. Frege perhaps offers the most famous expression of this objection when in the Introduction to Die Grundlagen der Arithmetik, he offers this methodological heuristic: “Never let us take a description of the origin of an idea for a definition, or an account of the mental and physical conditions on which we become conscious of a proposition for a proof of it. A proposition may be thought, and again it may be true; let us never confuse these two things. We must remind ourselves, it seems, that a proposition no more ceases to be true when I cease to think of it than the sun ceases to exist when I shut my eyes” Later, Frege summarizes his antipsychologistic resolve: “In the enquiry that follows, I have kept to three fundamental

16

DALE JACQUETTE

principles: always to separate sharply the psychological from the logical, the subjective from the objective...” The subjectivity of psychology has often been raised as the final line of defense against psychologistic reductions of an mind-independent objective logic. But while the contents of psychological experiences may be subjective, psychology as the scientific study of thought need not be. The objectivity of logic is moreover rendered at least somewhat doubtful by the fact that different logicians sometimes disagree about issues that to an outsider would appear to be among the most fundamental and unquestionable propositions of logic. An example is the dispute between classical and intuitionist logicians, in which classical logicians accept while intuitionist logicians for plausible if not universally admitted reasons deny the logical truth of the biconditional thesis that a sentence is true if and only if the negation of its negation is true. Intuitionists agree only with the conditional that if S then not-not-S, but not with the converse that if not-notS is true then S is true. Nor is it obvious that one side or the other in disagreements about logic must be wrong. There are different logics for different analytic purposes in modeling different kinds of reasoning about different kinds of things. As a result, the antipsychologist objection that psychologism is beyond redemption in the philosophy of logic because logic is inherently objective and psychology is inherently subjective appears inconclusive. I make no claim that the above list of objections is complete. Nor do I suppose that my counterobjections to the standard objections against psychologism that I have considered are decisive. It may be possible even where the defense of psychologism appears quite strong to reinforce a particular antipsychologist criticism in such a way as to lend it greater credence as a refutation of psychologism. To choose just one example, it might be granted that psychology can be as exact as logic, but that logic still cannot be reduced to psychology because the precise degree or kind of exactitude demanded by logic cannot be provided by psychology. The present point is only that there is more to the psychologismantipsychologism dispute than antipsychologists have generally acknowledged, basking in the warm assurance that Frege, Husserl, and other opponents have defeated psychologism so thoroughly that the entire approach is dead, buried, and impossible to resurrect.

5. Toward a Renewed Psychologism-Antipsychologism Dialectic If I have succeeded to any degree in making psychologism look less disastrous than its critics have tried to portray it, then there may be some hope of reawakening the controversy between psychologism and

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

17

antipsychologism. My desire is to rekindle interest in the topic and in a small way to help begin to restore the balance of dispute to what I believe should remain an open, lively, and more actively considered controversy. There is much to be gained in my view from constantly reevaluating the challenges posed by the psychologism-antipsychologism antithesis. My reason for thinking this goes beyond a general suspicion of fundamental philosophical disagreements ever being definitively resolved in one way or the other. A more particular encouragement is implied by the fact that psychologism demands further reexamination in light of the greening of psychology as a scientific discipline. The changing shape of scientific psychology, and new directions in phenomenology and intentionalist philosophy of mind, may combine to bring about a greater acceptance of objection-resistant formulations of psychologism, that may eventually support a more enthusiastic appraisal of psychologism versus antipsychologism. If renewed attacks against psychologism determine that its prospects cannot be revived under any aegis, then it will be just as well to satisfy ourselves that this is so. But if there remain good arguments for refined versions of psychologism that can withstand the best assaults of antipsychologism, then we may look forward to a fertile interaction between psychologism and antipsychologism that will continue to inspire a refinement of both positions. Only then can we overcome the stagnation that has been fostered by the rhetoric according to which any form of psychologism must be a hopelessly mistaken relic of nineteenth-century philosophical logic that in more enlightened times has been irretrievably defeated. Department of Philosophy The Pennsylvania State University, USA NOTES 1

Quoted in Philipse (1989), 58. See also Mulligan (1995), esp. 19-23. 3 Fodor (1988); Churchland (1986). 4 Quine (1960); Frege (1977); Bolzano (1972). 5 Compare Kusch (1995), 30-94; Rath (1994). 6 Gödel (1931); Church (1935). 7 Boole (1958): “The design of the following treatise is to investigate the fundamental laws of those operations of the mind by which reasoning is performed; to give expression to them in the symbolical language of a Calculus, and upon this foundation to establish the science of Logic and construct its method; to make that method itself the basis of a general method for the application of the mathematical doctrine of Probabilities; and, finally, to collect from the various elements of truth brought to view in the course of these inquiries some probable intimations concerning the nature of the human mind” (1). Boole sees no incompatibility 2

18

DALE JACQUETTE

between this psychologistic approach to logic, and the generality and certainty of the principles of logic as laws of thought, which he contrasts with inductive empirical methods in the natural sciences: “On the other hand,” he adds, “the knowledge of the laws of the mind does not require as its basis any extensive collection of observations. The general truth is seen in the particular instance, and it is not confirmed by the repetition of instances...For we not only see in the particular example [of a logical law] the general truth, but we see it also as a certain truth,— a truth, our confidence in which will not continue to increase with increasing experience of its practical verifications” (4). 8 Haack (1978), 238-42, distinguishes between the strong psychologistic claim that logic is descriptive of mental processes, and weak psychologisms by which logic is prescriptive of how we ought to think. She maintains that the problem of psychologism stands in need of reevaluation, and rejects strong psychologism while cautiously endorsing some of the advantages of weak psychologism over antipsychologism.

REFERENCES

Baker, G.P. and Hacker, P.M.S. Frege: Logical Excavations. Oxford: Basil Blackwell, 1984. Bolzano, Bernard. Theory of Science: Attempt at a Detailed and in the Main Novel Exposition of Logic with Constant Attention to Earlier Authors [Wissenschaftslehre, 1837]. Ed. and trans. Rolf George. Berkeley, University of California Press, 1972. Boole, George. An Investigation of the Laws of Thought of which are Founded the Mathematical Theories of Logic and Probabilities. New York: Dover Publications, Inc., 1958. Brentano, Franz. Psychology from an Empirical Standpoint. Ed. Oskar Kraus. Trans. Linda L. McAlister. London: Routledge & Kegan Paul Ltd., 1973 [includes Appendices from the 1911 revision of Brentano’s Psychologie vom empirischen Standpunkt (1874), published as Von der Klassifikation der psychischen Phänomena (The Classification of Psychological Phenomena)]. Church, Alonzo. “An Unsolvable Problem of Elementary Number Theory”, Bulletin of the American Mathematical Society, 41 (1935): 332-33. Churchland, Patricia Smith. Neurophilosophy: Toward a Unified Science of the Mind-Brain. Cambridge: The MIT (Bradford Books) Press, 1986. Dummett, Michael. The Interpretation of Frege’s Philosophy. Cambridge: Harvard University Press, 1981. Dummett, Michael. Frege and Other Philosophers. Oxford: The Clarendon Press, 1991. Eisler, Rudolf. Einführung in die Erkenntnistheorie. Leipzig: Verlag von Johann Ambrosius Barth, 1907. Feigl, Herbert. “Physicalism, Unity of Science and the Foundation of Psychology”, in Paul A. Schilpp, ed., The Philosophy of Rudolf Carnap (LaSalle: The Open Court Publishing Co., 1963). Fodor, Jerry A. Psychosemantics: The Problem of Meaning in the

PSYCHOLOGISM THE PHILOSOPHICAL SHIBBOLETH

19

Philosophy of Mind. Cambridge: The MIT (Bradford Books) Press, 1988. Frege, Gottlob. The Foundations of Arithmetic [Die Grundlagen der Arithmetik: Eine logisch mathematische Untersuchung über den Begriff der Zahl (1884)]. Trans. J.L. Austin. Oxford: Basil Blackwell, 1974. Frege, Gottlob. “Thoughts”, in Frege, 1977: 1-30. Frege, Gottlob. Logical Investigations. Ed. P.T. Geach. Trans. Geach and R.H. Stoothoff. Oxford: Basil Blackwell, 1977. Gödel, Kurt. “Über formal unentscheidbare Sätze der Principia mathematica und verwandter Systeme I”, Monatshefte für Mathematik und Physik, 38 (1931): 173-98. Haack, Susan. Philosophy of Logics. Cambridge: Cambridge University Press, 1978. Hacker, P.M.S. Wittgenstein’s Place in Twentieth-Century Analytic Philosophy. Oxford: Basil Blackwell, 1996. Kusch, Martin. Psychologism: A Case Study in the Sociology of Philosophical Knowledge. London: Routledge, 1995. Mulligan, Kevin. Review of Martin Kusch, Psychologism: A Case Study in the Sociology of Philosophical Knowledge. Metascience 8 (1995): 1726. Musgrave, Alan. “George Boole and Psychologism”, Scientia, 107 (1972): 593-608. Notturno, Mark A. Ed. Perspectives on Psychologism. Leiden: E.J. Brill, 1989. Pap, Arthur. Semantics and Necessary Truth: An Inquiry into the Foundations of Analytic Philosophy. New Haven: Yale University Press, 1958. Philipse, Herman. “Psychologism and the Prescriptive Function of Logic”, in Notturno, 1989: 58-74. Quine, W.V.O. Word and Object. Cambridge: The MIT Press, 1960. Quine, W.V.O. “Epistemology Naturalized”, in Quine, 1969: 68-90. Quine, W.V.O. Ontological Relativity and Other Essays. New York: Columbia University Press, 1969. Quine, W.V.O. Philosophy of Logic. Englewood Cliffs: Prentice Hall, Inc., 1970. Radnitzky, Gerhard. “Popperian Philosophy of Science as an Antidote Against Relativism”, in R.S. Cohen, et al., ed., Essays in Memory of Imre Lakatos. Dordrecht: D. Reidel Publishing Co., 1976. Rath, Mathias. Der Psychologismusstreit in der deutschen Philosophie. Freiburg: Alber Verlag, 1994.

This page intentionally left blank

ROLF GEORGE

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

I In a review of Bolzano’s Wissenschaftslehre (WL, 1837), a pseudonymous Dr. P. Menelaos remarked that “throughout, the author assumes the old, strictly objective or dogmatic, viewpoint, in contrast to the contemporary, which is based on the psychological self-consciousness of the thinking mind” (Bolzano 1972, XXIX). It must be granted that Bolzano did not think much of grounding logic in psychological self consciousness, and that he introduced mind independent, timeless, “sentences in themselves,” “objective” propositions, among which relations of consequence, consistency, probabilification, etc. obtain. He is now remembered mostly for this theory, which, as a counterpoint to psychologism, earned him the admiration of Husserl (1900, 225 f.), and the titles of a logical Plato and a Bohemian Leibniz. By contrast, Bolzano’s specific logical teachings, though of great and even current interest, have received less attention. It is not the focus of this paper, however, to deal with them. Dr. Menelaos is correct in saying that Bolzano, unlike many of his contemporaries, did not seek to ground logic in psychology, but it is a mistake to think that he paid no attention to the “manifestations” or “appearances” of propositions in the mind, as the contents of thoughts that come and go: substantial portions of his Wissenschaftslehre are given over to this concern. A second misleading suggestion is that there was an accepted “contemporary” way of thinking about logic in psychological terms, when there were in fact deep divisions on just this issue. Thirdly, it is suggested that Bolzano reverted to an older style of doing logic, not long before then superseded by the modern psychological approach. Although Bolzano gives much space to pointing out, largely as a matter of courtesy, that many earlier logicians must have thought of something like propositions (as mind-independent entities), it is manifest that none of them had

21 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 21-49. © 2003 All rights reserved. Printed by Kluwer Academic Publishers, the Netherlands.

22

ROLF GEORGE

explicitly postulated them. He did not, as is claimed, return to a previous practice, but pioneered a new one. Indeed, there was not all that much previous practice. During the 18th century logic had fallen into a state of disrepair; the Enlightenment was perhaps an age of reason, but it surely was not an age of logic. A reconstruction and rehabilitation of logic had to start very nearly from the ground up. The logic books that were then published were meant to be propaedeutic to philosophical studies, more school book than treatise. Logic’s decline had been brought on, in part, by an increased insistence on its connection to psychology, only much later disparagingly called psychologism, in part by some other factors. I will now chronicle some highlights of this development.

II It is tempting to think that psychologism made its entry into logical theorizing as a matter of reasoned insight, as a better grounding of the subject than what had gone before. That would be a mistake. It was as much a matter of style, fashion and ideology as of theory. I ask, therefore, the reader’s indulgence as I begin with a ramble through some aspects of the sociology and politics, if I may use that term, of logic in the late 17th and 18th centuries, without, however, neglecting theory altogether. Older logics of the conservative variety, into the 17th century, usually began with a few perfunctory remarks about thinking and the laws of thought, but then got right down to business, discussing terms, propositions, syllogisms, without much attention to the ontological status of the entities they discussed. One cannot, without anachronism, describe them as either psychologistic or anti-psychologistic — that was not an issue. The pedagogy of logic divided the subject into the logica docens and the logica utens. There is, on the one hand, the subject as it is taught, remembered and practiced in disputations, and on the other hand its use in everyday life, in legal practice, and the like. There was no attention to what later came to be called “natural logic”, that is, the inferential habits of the untutored mind. The distinction between the latter and “artificial logic”, the codification of the logical canon, came later and is found, for instance, in the Logic of Port Royal. Arnauld says this: Conceiving, judging, reasoning, ordering are all done quite naturally, and sometimes done better by those ignorant of the rules of logic than by persons instructed in these rules. So logic does not teach us how to conceive, to judge, to reason, or to order; for nature in giving us reason gave us the means to perform these operations. Logic consists, rather, in reflecting on these natural operations.1

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

23

He goes on to say that this reflective knowledge of logic assures that we reason well, and that it has therapeutic value if things go wrong. There is, thirdly, the theoretical, scientific interest: logic investigates the natural operations of the mind. It is to the mind as physics is to material bodies (ibid.), it is descriptive of the operations of the mind and, in a manner of speaking, is a branch of psychology, though that word is not used. Arnauld repeatedly warns against the conceit that knowledge of logic, that is knowledge of the workings of the mind, will always improve reasoning: “Though there are some whom logic helps, there are others whom it harms” (204) and: Though an argument may contain features not in conformity with the rules, if our native understanding says that the conclusion is validly drawn, then our common sense should be respected (204).

If the job of logic as a science is to describe the operations of the mind when it reasons, the question at once arises if the traditional treatment of the subject, the Barbara - Celarent routine, does a good job of it. This was soon denied. It was Locke who even more sharply focused attention on the acquisition of ideas, and the actual workings of the mind in cognizing and inferring, and who then extolled the virtues of untutored and thus unstifled reasoning. While Arnauld had done an exemplary job laying out the theory of the syllogism, for instance, Locke discussed the subject with derision: God has not been so sparing to men to make them barely two-legged creatures, and left it to Aristotle to make them rational...I would fain see the man that was forced out of his opinion by dint of syllogism...Native rustic reason...is likelier to open a way to, and add to the common stock of mankind, rather than any scholastic proceeding...I doubt not, nevertheless, but there are ways to be found to assist our reason in this most useful part... [but they] will scarce be found, I fear, by those who servilely confine themselves to the rules and dictates of others. For beaten tracks lead this sort of cattle...whose thoughts reach only to imitation, not where we ought to go, but where we have been (Essay 4.17.3/6/7).2

Traditional Logic was still part of the Oxford routine, and those engaged in its teaching were not pleased to be likened to livestock, nor could they take kindly to Locke’s claim that logic and the liberal sciences have “given reputation” to the “abuse and the mischiefs of confounding the signification of words” and this because in the “admired Art of Disputing... victory does not go him who has truth on his side, but the last word in the dispute” (Essay 3.10.6/7).3 Consequently, in 1703 the Principal of St. Edmund’s Hall and the Master of Balliol put a motion to the heads of Oxford colleges to censure the Essay and ban it from Oxford, perhaps intending a bonfire like that of 1683 when the Convocation burnt Hobbes’s Leviathan and De Cive in the Schools’ quadrangle. The reason for the proposed ban was, as one con-

24

ROLF GEORGE

temporary source had it, that “there was a great decay of logical exercises in the University which could not be attributed to anything so much as the new philosophy, which was too much read.”4 The Warden of Wadham stood up for Locke, and the motion failed, but the Heads of Houses informally instructed tutors to forbid students the Essay. Locke was amused. Alexander Pope later picked up the image of logicians as a herd: Thick and more thick the black Blockade extends A hundred Head of Aristotle’s Friends... Each staunch Polemick, stubborn as a Rock Each fierce Logician still expelling Locke. (Dunciad iv, 191 ff.)

Leibniz pleased himself imagining that he could actually persuade Locke, or at any rate his stand-in Philateles of the New Essays, not, indeed, of the usefulness of syllogistics, but of logic properly conceived and carried out, a project to which he repeatedly pledged himself. Philalethes: I begin to form for myself a wholly different idea of logic from what I formerly had. I regarded it as a scholar’s diversion, but now I see that, in the way you understand it, it is like a universal mathematics. Would God that it might push on to something more than it yet is, in order that we might be able to find thereby these true helps of reason...which would raise men far above their present condition (1913 4.17, p. 569 f.)

These are the concluding remarks of an insightful discussion of logic and its role, written down in 1704, but not printed until 1765. Had they been published earlier, they might have changed the course of thinking on logic. As it was, Locke had a head start of two generations. Locke did not mean merely to criticize traditional logic. He and those following his lead intended the new philosophy to take its place; Hume’s Treatise, too, was meant to supplant all logic teaching. In some circles logic was no longer thought even to include the formal teachings of the schools and the received canons. Rather, in Hume’s words, “the sole end of logic is to explain the principles and operations of our reasoning faculty, and the nature of our ideas” (Treatise, xix). After laying out his associationism and theory of causation, he gives eight rules like Mill’s Canons in spirit if not in detail “by which to judge cause and effect,” and continues: Here is all the logic I think proper to employ in my reasoning; and perhaps even this was not necessary, but might have been supplied by the natural principles of our understanding. Our scholastic headpieces shew no such superiority above the mere vulgar in their reason and ability, as to give us any inclination to imitate them in delivering a long system of rules and precepts to direct our judgment in philosophy (ibid. 1.3.15, 175).

A proper theoretical account of the workings of the mind will show that natural reason needs to be liberated rather than trained; the corrupting

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

25

weight of scholastic authority must be got off the back of the common man so that his thinking may be restored to its natural state. This goes beyond even what Locke (who took aim mostly at the practice of disputations), and certainly what Arnauld had said: the received logic is now not just a superfluous codification of what comes naturally, but a detriment to sound thought. It belongs to a pompous and affected culture and belongs to the things he really detests like “celibacy, fasting, penance, mortification, selfdenial, humility, silence, solitude, and the whole train of monkish virtues” which also follow from the denial of “natural, unprejudiced reason” (Principles of Morals 9.2, 1975, 270). Hume’s opinion was common, reflected, for instance in the article Logique in the Encyclopèdie, which was not written by d’Alembert, who could have done justice to the formal aspects of the subject and might have advanced it, but by Diderot, who could do neither and who echoed the Lockean and Humean sentiments: ...in order for our mind to work well, it is not necessary to study how it works. The mind is an instrument which God made, and made well. It is most useless to discuss metaphysically what our understanding is and how and of what pieces it is composed. It is as if one set oneself the task of dissecting the human leg in order to learn how to walk. Our reason and our legs fulfill their functions perfectly well without all this anatomy and these prologues; it suffices to use them, without asking more of them than they are capable of. In addition, if the mind could not work well without the help of artificial logic, then it also could not be certain that the rules which it has established were good ones. For the rest, we will show that syllogisms are by no means necessary for discovering truth. (Diderot 1969).

Diderot claims that the subject, traditionally treated, is surcharged with terms and barbaric phrases, ridiculous and invariably disgusting. The way it is still treated in the schools “plays no small part in strengthening the contempt in which many people have always held this science” (ibid.). The nine columns of Diderot’s article contain only polemic, no logic. No one ignorant of the subject can here discover what it is. The heroes, predictably, are Locke and other champions of untutored reason. The article on syllogism (where again we remain ignorant of what a syllogism is) is even more extreme: Montaigne not only bore contempt for the rules of argumentation (as did Locke), but claimed that ordinary logic was only useful for creating dirty, fetid [crottés et enfumés] pedants . “The greatest mark of wisdom — he says — is a constant enjoyment; a state like that of the things beyond the moon, always serene. These baroco and baralipton, which make their henchmen so dirty and fetid, these are not wisdom; they only know by hearsay how it brings about a state of serenity which calms the storms of the soul and teaches it to laugh at hunger and fever, not by means of imaginary epicycles, but by natural and probable reasons.”5

26

ROLF GEORGE

So much for syllogisms. In Goethe’s Faust we find a poetic rendition of the same sentiment. Mephistopheles advises a naive and eager student above all to drill himself in logic as a preparation to all other studies: At first collegium logicum. There will your mind be duly braced and well in Spanish boots enlaced so that more slowly than before thought creeps to execute its chore and not go zigzag, here and there like jack-o-lanterns in the air ....(Faust 1.1.4).

The lesson: logic slows thought and kills spontaneity. Hegel, for different reasons, carried forward the Humean disdain for logic with an example not unlike Diderot’s: That one could learn to think through logic, usually thought to be its benefit and therefore purpose — as if one could only learn to digest and walk through the study of anatomy and physiology — this is a prejudice that has long been exploded (1812, V; 1978, 6).

Human reason is a natural thing, no more capable of improvement by instruction than a healthy digestion. Bolzano later took Hegel to task, making the obvious point, already suggested in Arnauld, that as long as one is in good health, no physiological knowledge is needed to aid digestion. Knowledge of emetics, etc. is however helpful, if the bowels act up. Just so with logic (WL §9).6 Enlightenment logic bashing was not universal, however. Traditional instruction in the subject went on, and numerous texts and handbooks continued to be published. Christian Wolff, in obvious reference to Locke, took issue with the superiority claim for natural logic: Some think that natural logic is far superior to the artificial, and some even want nothing whatever to do with the artificial and think that the natural variety suffices for all tasks of the understanding (1965, 244).

He notes that “these days much is presented as artificial logic that does not deserve the name” (ibid.), but that the good stuff has good uses as a corrective, i.e., in detecting fallacies and similar tasks, but also in perfecting and facilitating reasoning. Similarly, a couple of generations later Thomas Reid, who opposed Hume in most things, observed sensibly: Men rarely leave one extreme without running into the contrary. It is no wonder, therefore, that the excessive admiration of Aristotle, which continued for so many ages, should end in an undue contempt, and that the high esteem of logic, as the grand engine of science, should at last make way for too unfavorable an opinion, which seems now prevalent, of its being unworthy of a place in a liberal education. Those who think according to the fashion, as the

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

27

greatest part of men do, will be as prone to go into this extreme as their grandfathers were to go into the contrary. (1843, 592 f.)

Reid, nonetheless goes on to praise Locke’s contribution to the subject. The relation between logic as it was traditionally understood and the laws of the mind Locke and Hume thought they had discovered had to be an uneasy one. If, as in Hume, ideas are copies of impressions, or recombined copies, they share some of their sensual properties, being round and coloured or loud etc., even if in some sense “fainter” than their original. It is not clear how these entities could, and perhaps impossible that they should, play the role of subjects or predicates of judgments, which they must if, for instance, geometrical truths are “relations of ideas.” It is not surprising, then, that in Hume there is no good and clear theory of judgments, or of argument. Laws of association, on the other hand, can plausibly be explained as one image calling up another. Interesting associationist theories can be developed, but logic, as traditionally understood, lies beyond the horizon of this sort of inquiry. Associationism cannot be a foundation, even if one takes the task of logic to be the tracing of thought. Despite this, as noted, the term ‘logic’ continued to be applied to the epistemological inquiries here undertaken.

III Extolling natural reason at the expense of logical discipline was only one reason for the decline of logical culture in the 18th century. A related one is suggested in the passage from Locke, cited above: Native rustic reason...is likelier to open a way to, and add to the common stock of mankind, rather than any scholastic proceeding...

Here as in many other sources, logic, if it is to be useful, must be an engine of discovery and add to the “common stock of mankind.” It was common to distinguish a branch of logic called ars inveniendi or art of discovery, from the ars iudicandi or art of judging the validity of evidence and proofs derived from some other source. Logic as traditionally taught was adequate enough to do the latter, if the proofs in question were syllogisms and immediate inferences in the standard mode. But this, as Locke would have it, is to follow the “beaten track” of evaluating an argument already discovered by someone else. The distinction had a long history. Arnauld traces it back to a Ciceronian admonition: Let us, then, completely forego that art which has all too little to offer when proofs are being sought, and over much when they are being evaluated (De Oratore II.38.160, Arnauld 1964, 236).

28

ROLF GEORGE

Heimsoeth noted that “art of discovery” [Erfindungskunst] had become a fighting slogan in the Renaissance: The new art was claimed superior to traditional logic, which supposedly was of use only as an organon for proving and presenting already discovered truths.7 The fight in question was against the scholastic and Aristotelian tradition of logic. A later source remarks that “Aristotle did not even touch upon many necessary and useful things, e.g. the way of discovering truths.”8 Descartes was among those who did touch on these matters, viz. in the Discourse on Method: I observed in respect to Logic that the syllogisms and the greater part of the other teaching served better in explaining to others those things that one knows... than in learning what is new (1955, 91).

While it contains some good principles, logic as well promotes some that are “hurtful and superfluous” (ibid). Francis Bacon says in the Novum Organum, that the “logic we now have” does not help us in discovering new things and “serves rather to consolidate error than to assist in the search for truth” (Novum Organum 1.11,12).9 Unlike Locke and his followers, however, Bacon did not think that unaided human reason was up to the task of augmenting the sciences, since the “spirit of man... is variable, confused, and as it were actuated by chance” (N.O. 1.42). After he details the “idols of the tribe,” that is, the shortcomings of human understanding found in all of us (N.O. 1.45-51) he sums up: Such are the idols of the tribe, which arise either from the uniformity of the constitution of man’s spirit, or its prejudices, or its limited faculties, or restless agitation, or from the interference of the passions, or the incompetency of the senses, or the mode of their impressions (N.O. 1.52).

The natural, psychological tendencies of the mind therefore stand in need of a remedy which the old logic is unable to provide. A new logical regimen of carefully devised rules of induction, experimentation and record keeping must be instituted as a control: I supply the mind with such rules and guidance that it may in every case apply itself aptly to the nature of things (N.O. 1.127).

This new method of scientific discovery (“a kind of logic, though the difference between it and the ordinary logic is great” (Bacon 1939, 15)) can be devised only by attending to the actual operations and shortcomings of the understanding and depends on psychological insight. The traditional exercises could not impress a Lord Chancellor who thought the virtue of a reformed logic to lie not merely in veritas, but also in utilitas, and who died, a pioneer in the field of cold storage, trying to freeze a chicken.

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

29

Christian Wolff attempted to rehabilitate traditional logic, specifically syllogistics, by showing that it was itself the sought after engine of discovery. His argument is a narrow adaptation of the analytic-synthetic method first articulated by Pappus and common in the preceding century: the sentence to be proved is first “dissolved” or “analyzed” in such a way that premisses are found that are needed for its proof. This is continued until one arrives at premisses whose truth is already established. The “synthesis” is then merely a restatement in opposite order of the steps uncovered in the “analysis”.10 Wolff first gives a not altogether successful proof in syllogisms of the theorem that the interior angles of a triangle equal to two right angles (Euclid, Elements, Prop. 32). He then asks how this proof might be discovered and responds by repeating the syllogisms he had just used for the proof — a reversed version Pappus’s procedure. In sweeping generalization he then proclaims that “through the common arguments all truths are discovered” (Wolff (1965) 175, Ch.3, §24), and: ...through these arguments [Schlüsse] everything is discovered that human understanding makes evident, and everything is demonstrated to others who want to be convinced of its certainty (171, Ch. 3, §20).

Syllogisms represent “the natural sequence of thoughts.”11 If correct, this would show that the classical canons were in themselves a proper account of the scientific mind at work, and indeed that syllogistics is itself the underlying natural logic.12 Wolff’s argument did not catch on. In brief: if the art of discovery is thought part of logic, an opening exists for psychology in logic. For Locke’s, simply, the “natural logic” inherent in the mind and the tracks it follows are superior to the scholastic canons when it comes to finding new truths. Plainly, new truths are always discovered, but not thanks syllogisms. For Bacon the mind is but an “uneven mirror” of nature, whose faults could not be overcome except by a psychologically insightful new logic. And even Christian Wolff, in his apology for traditional logic, had to argue that syllogisms represent the natural sequence of thoughts. Eventually the art of discovery was split off from logic. Largely through the efforts of Kant it became a separate concern and was absorbed into what is now called epistemology. Until then, and as long as it was thought integral to logic, psychology inevitably intruded into that part of the field.13

IV Not directly connected with the increasing influence of psychology, but in part responsible for the sad fate of logic in the 18th century, was the diminishing public role of logic instruction. A few words about this are in

30

ROLF GEORGE

order. In 1654, Seth Ward wrote that students entered university (Oxford in this case): ...not that they be engaged in...experimentall things [he had mentioned Chymistry, Agriculture and Mechanics], but that their reason, and fancy, and carriage, be improved by lighter Institutions and Exercises, that they may become Rationall and Gracefull Speakers, and be of acceptable behaviour in their Countries (Ward 1654, 50).

These “lighter Institutions” were meant to include logic. As the Earl of Clarendon later remarked: ...though we do not make our conversation in Syllogisms, and discourse in Mood and Figure, yet our Conversation and our Discourse is much the more reasonable, and the better formed, by the Experience we have had in that Art, and in which we may have spent some Time very merrily. And I must say again, this most useful Art was never well taught or learned but in the Universities (1727).14

It was recognized, of course, that bearing and conversational savvy were enhanced not just by logical knowledge, but by the culture of disputation that went with the instruction. We should note also that Clarendon thought students engaged in the study of the subject “merrily”, sounding the “logic is fun” theme. Students evidently did not find logic as entertaining and not as useful toward the end of the century as a few decades before, preferring Locke. This development occurred as well on the continent. If we look at the Logic of Port Royal with this in mind we note that Arnauld stressed the public utility of logic properly taught. One of his complaints was that logic usually was not practical enough: it was freighted with abstract and useless precepts, second intentions, universals a parte rei and the like, and harnessed to “examples which are unexciting and never encountered elsewhere” (1964, 21). As a consequence, of a thousand young men who learn logic “no more than ten know anything of the subject six months after the course is finished” (ibid.). He thought that a book containing the practical and applicable part of logic could be written in a day, and taught in five. In the event, he wrote it in five days (or so he claims, (5)), but teaching it took only one day for each of the four parts, one less than projected. Surely an impossibility showing that logic and truth do not always go together. Arnauld’s book was translated and reprinted many times and remained a much used text for over a century. He himself did not sacrifice much logical doctrine to practical usefulness, but others were not so scrupulous. Wolff reports that his teachers taught him to despise the “common arguments; ...I and others laughed about them without comprehension [so that] they have become subject of universal derision” (1965, 173: Ch 4. §22). Yet, despite his view of the centrality of syllogistics and its role in discovery, he

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

31

surrenders to the mood of the time, omitting discussion of the second and third figures (to say nothing of the fourth) because he does not want to “hold back the beginner with superfluous matters, especially with those made so hateful these days” (1965, 168, Ch.4, §14). As well, many other traditional matters bit the dust, despite which the book is a useful, systematic, and concentrated account of many subjects in the broader reaches of logic, viz. definition, persuasion, disputation, the critical reading of books and so forth. First published in 1712, it went through several editions and was widely used. King Frederick William I. ordered all preachers “to arrange their holy sermons in accordance with this logic,” and his son, the Great Frederick, commanded its use in the drill of officer candidates (Arndt 1965, 96). Notwithstanding the royal promotion of Wolff, logic contnued to decline. While there are still substantial portions of logical doctrine in Wolff, other authors promised in their very titles that their logic books will in effect contain no logic. Here is Christian Thomasius: Introduction to the doctrine of reasoning, which shows in a simple way, comprehensible to all reasonable people, of whatever station or sex, to distinguish the true, the probable and the false from each other, and to discover new truths without syllogisms (Thomasius 1691). Toward the end of the century we find more books promising easy and non-technical instruction like Logic for People Who do not Want to Study.

V Kant lectured on logic throughout his career, using as his text George Friedrich Meier’s Auszug aus der Vernunftlehre (Meier 1752). As was customary, Meier gives much space to subjects like dogmatism, scepticism, prejudice, testimony, clarity of concepts, and not very much to ratiocination. The rules of logic, he claims, must be derived “from the experience and effects of human reason, from the nature of human reason, and from the basic truths on which all human knowledge rests” (Akad. 16.5). He acknowledges natural reason, but unlike Locke and Hume, and in line with most practitioners, thinks that school logic can improve it. Kant justified the choice of the book in his “Announcement of the Arrangement of his Lectures” for the Winter Semester 1765/66 (Akad. 2.303-13). He noted that many students quit their lectures because they are overwhelmed with abstractions. To halt the exodus he intends to begin with “a short introduction to empirical psychology.” He reasons that the: ...student whose enthusiasm has evaporated toward the end of the empirical psychology (which really need not be feared if one proceeds in this way) will have learned something that is comprehensible because easy, pleasant because interesting, and useful because of its many applications in life (Akad. 2.309 f.).

32

ROLF GEORGE

The logic lecture will be based on this psychological introduction and will aim at the improvement of sound reason, which should precede all science and learning as the: ...quarantine (if I am allowed this expression) in which the apprentice must be kept if he wants to migrate from the land of prejudice and error to the domain of enlightened reason and science (Akad. 2.310).

Kant had taught logic for the first time in the Winter of 1755/56 and began, at that time, to make notes in his copy of Meier. He stopped doing so in the decade of the 60s, when he had fallen under the spell of Hume or, as he would have it, when Hume had “awakened him from his dogmatic slumber” (Prolegomena, Preface, Akad. 4.260).15 Although there are no notes of his own, we know from student transcripts of his lectures that Hume effected his understanding not only of causality, but also of logic. In 1762/63, according to Herder’s fragment, he taught that Aristotle’s logic “had done the greatest damage...All logics [i.e. logic books] have caused damage... To study scholastic logic is torture...The task of logic is to abstract rules from common cognition” (Akad. 24.4/5). A decade later he notes that “corrupt taste” was responsible for keeping the memory of Aristotle alive and says that “in logic...one studies one’s own subject, the laws of the understanding as well as those of sensibility” (Akad. 24.333/4). At about the same time another student wrote this down: Aristotle ... harmed philosophia more than he helped it (Kant 1992, 23). It took great effort to forget [Aristotle’s] false propositions, to give the understanding its natural perfection again, and to investigate its true rules (ibid. 16). Locke’s book de intellectu is the ground of all true logica (ibid. 24).

But after another decade, at about the time of his writing the Critique of Pure Reason, Aristotle is no longer chastised, though the “slavish” adherence to his doctrine is mentioned (ibid. 263). Hume and Bayle are now antilogici. A transcript of 1792 reports a decisive change of view that must have taken place a decade earlier: We have no one who has exceeded Aristotle or enlarged his pure logic (which is itself fundamentally impossible), just as no mathematician has exceeded Euclid (ibid. 438).

Locke is now dethroned: “He speaks [in the Essay] of the origin of concepts, but this really does not belong to logic, but rather to metaphysics” (ibid. 439). In his published Logic of 1800 (actually a book compiled by Gottlob Benjamin Jäsche from student transcripts and Kant’s notes), the same position is maintained: Aristotle is “the father of logic” and Locke not a logician at all (ibid. 534 f.).

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

33

In the Critical period Kant distanced himself from the rustic reason school of Locke, Hume, Condillac et. al., and their preoccupation with psychology. From the early remark “The logician must know the human soul... Not without psychology” in 1762 (Akad. 24.3) he moves to saying, in the first Critique, that “pure logic derives nothing from psychology.”16 This change of mind brought with it a change in the subjects covered. In a transcript from the early ’70s the four figures of the syllogism are not even mentioned (Kant 1992, 227-29), in keeping with his essay on “The False Subtlety of the Four Syllogistic Figures,” of 1762, which justifies the omission (Akad. 2.57). Later they are treated, in some cases rather extensively, even if the derived figures suggest acts of mind that do not actually take place since (as he had said) “reason only infers in the first figure and cannot do otherwise” (Akad. 24.474). But this no longer matters; psychology no longer dictates. The major and most influential German logics of the next generation were written by Kantians. Most notable is J.G.C.C. Kiesewetter’s Grundriß einer allgemeinen Logik nach Kantischen Grundsätzen (1796), which contains a solid and extensive introduction to traditional logic and went through several editions. J.F.A. Calker could say with good reason that “Kant gained a pervasive influence upon the history of logic” (1822, 189) and Bolzano’s thorough coverage of historical antecedents refers more often — by a very large margin — to Kant and Kantian logicians than to any other source (1837; vol. 4. 687-717 of the edition of 1931). This is a reminder that Kant’s work and reputation played a role in the continuation and indeed increased activity of ordinary school logic, conceived as a propaedeutic to all other study. I will therefore briefly sketch the treatment of the subject in the later Kant and his students.17 Kant made several claims and distinctions that were followed by all Kantian logicians of the next generation. I note, first, his division of logic into “organon” and “canon”. Niceties aside, the first stands for the older ars inveniendi, art of discovery. It occurs occasionally in phrases like “heuristic organon” (Refl. 1600, Akad. 16.31). Logic, he says, serves as a critique of the understanding, but not for creation (Hervorbringung) and is thus not an organon (Refl, 1602, Akad. 16.32), a point he repeatedly makes in his published work. In the Jäsche logic he says that an organon is a “directive as to how a certain cognition [knowledge, Erkenntnis] is to be brought about” (1992, 528, Akad. 9.13). But this presupposes a grasp of the special science in which the discovery is to be made. There is no general art of discovery, and since logic is to be general, paying attention only to form and not to content, it cannot be an organon, but only a canon, that is, a decision method (diiudicatio, Refl. 1585, Akad. 16.26). This is the old ‘ars iudicandi’ , containing the “necessary laws and conditions of the correct use [of the understanding] (1992, 529, Akad. 9.13). The effect of this is that the contentious art of discovery is removed from the field of logic.

34

ROLF GEORGE

A further point Kant makes is that the notion of a natural logic is incoherent. The study of the actual inferential habits of the mind of a person has to be an empirical inquiry. Kiesewetter puts the point this way: One divides logic into natural and school logic (naturalis and artificialis). In the first case it is viewed as a natural ability (habitus), in the latter as a science. Only the latter is our concern, since the former belongs in psychology (1824, 7).

The “artificial logic” of earlier logicians now becomes “pure logic”, accessible a priori and containing the “necessary” part of the logical machinery that performs the function of “natural” logic. The content of pure logic is the same as that of artificial or school logic, but there is now a new grounding of the supposed necessity of its laws. In the context of Critical Philosophy, logic, under the name of transcendental logic, plays a founding role in the very constitution of reality. In the view of Kant and the Kantians, it can therefore not itself be an empirical science, but is, rather, a condition of all science. This is why the laws of logic are necessary and apply to all things, while the discoveries of psychology can only be contingent. There is, however, an ambiguity in the concept of a natural logic. When Kant claims that there is no such thing, he says no more than that logical doctrine cannot be discovered through psychological investigation. He does not mean to deny that the mind is programmed to follow the rules of logic, and that they are the software on which it runs. In this second sense the artificial or pure logic codifies the activity of the sober mind in reasoning. Unfortunately, the mind does not always obey the laws of logic. It is sometimes “pathologically” determined. Passion, lack of attention, forgetfulness, prejudice can get in the way. There should therefore be another branch of logic, or logic instruction, the so-called applied logic. Kant discusses it in the First Critique (A54/B78), and Kiesewetter describes it thus: Applied general logic must know the subjective conditions of the human use of the understanding, and therefore it needs as auxiliary sciences empirical psychology, anthropology, critique of pure reason (1824, Vol. 2, 2).

This part of logic aims to deal with the proper implementation and unhindered application of the rules of “pure” logic. In sum, Kant and the Kantians held that logic is not an art of discovery, but a “canon”, that there is no “natural logic” or, rather, that this is a misnomer since empirical investigation cannot yield logical insight. They claim that logic properly conceived is not based on psychology but that it is “pure”, that is, a priori, and that there is an applied branch dealing, basically, with obstructions that can get in the way of logical thinking.

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

35

The working out of this program is disappointing, particularly the way in which the claim for the necessity of the laws of logic is supported. It is worth a moment’s digression to understand what is not argued here. It is not merely claimed that one’s own inferential habits must appear valid to oneself, that even the discovery of an inferential error, accepting that a mistake was made, presupposes acknowledgment of the law that has been violated. Though there is a sort of subjective necessity to the logical routines I unearth when I reflect on my reasoning, Kant and his followers wanted a stronger conclusion, namely, that the a priori laws of pure logic are objectively necessary and universal, i.e., intersubjective. But their arguments disappoint. The law of contradiction, variously phrased, plays a central role in all these logics as the law, it is claimed, that validates all others. How is the necessity of this law argued? Jakob says this: The understanding cannot think anything that contradicts itself. Contradictory representations cannot be united in one consciousness...Contradictory representations are representations, one of which denies the other. Hence they destroy each other when one attempts to unite them (Jakob 1800, 32).

Kiesewetter, after claiming that “the understanding itself is the source of pure general logic” (1824, 9) goes on to say that “a whole that is inconsistent [das sich widerspricht] cannot be united in a unity of thought” (14). In a note he remarks “round and circular can be thought together, round and quadrangular cannot. One representation destroys the other” (48). Born gives an incomprehensible reason why we cannot think the contradictory: Neither the matter nor the form can be thought by itself, because the former is merely in the object, and the latter merely in the subject: from this we can understand [but how?] why it is that the contradictory can simply not be thought. For the contradictory consists in the combination of predicates, one of which is diametrically opposed to the other so that positing one of them thwarts positing the other (1791, 7).

Wilhelm Esser’s System der Logik is a clever and sensible book that acknowledges Kant’s role in reshaping logical theory while maintaining that logic is by no means a finished and complete science as he had claimed (1823, 47). Esser, like Born, sensed that an argument was missing to show that we cannot think the inconsistent. He notes that logic is not a branch of psychology but that the formulation of logic requires nonetheless a single psychological fact, namely that we think (8). The sketchy argument that follows can perhaps best be cast (much amplified) in Kantian (or perhaps Carnapian) terms, as follows: The self, not the noumenal but the phenomenal, is itself a constructed object. It is a coherent, minimally a consistent, set of cognitions. But many

36

ROLF GEORGE

representations occur in consciousness that do not cohere with the rest. They must be excluded, classed as dreams, illusions, figments, while the remainder becomes “my experience.” Put crudely, my self is (or I am) a large, perhaps the maximal, consistent set of representations culled from a more inclusive set. When an object appears to have two contradictory attributes, one of them must be rejected: they cannot both be experienced by me, because consistency is the very core of what gives my apperception its transcendental unity. It is a clever, if confused, argument, throwing together what I experience with what I can think about. While it may be granted that I cannot experience inconsistency in objects, I surely can form inconsistent concepts, like that of a regular dodecahedron with hexagonal sides. I could not sit on one, because none can exist, but I can form (and just did form) the concept. Bolzano, for instance, insisted that one must be able to form such concepts, and that they are concepts “without objects”. How else could one discover that there are no regular dodecahedrons with hexagonal sides? How could I investigate the problem of odd perfect numbers if I could not think that thought? (It is not yet known whether there are any). The Kantian attempt to escape psychology was unsuccessful, first because people do in fact at least entertain inconsistent concepts. If they were psychologically incapable of doing so, could they construct indirect proofs or argue that there are no square circles? Further, there is the problem of error. If someone actually managed the thought that dodecahedrons had hexagonal sides, it would have to be despite and in the face of the fact that this is inconsistent. The usual answer to this problems was that in such a case the concepts are not “clear.” Hence the claim that inconsistency cannot be thought must come with the proviso that the concepts in question are clear, thus importing a further piece of psychology into logic. I mention, finally, the dubious claim, surely also psychological if anything is, that different logical operations are performed by different faculties of the mind: concepts, judgments and immediate inferences are the province of the understanding, syllogisms that of reason, and induction falls under the faculty of judgment. Despite all this, Kantians continued to claim that pure logic as they constructed it is independent of psychology. The program was laudable, but its execution did not succeed.

VI Johann Friedrich Herbart was the first logician I could find to describe the objects of logical inquiry in almost Fregean terms, and logic as concerned with the objects of thought, not with thinking:

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

37

All our thoughts can be viewed from two sides: on the one hand as activities of our mind, on the other with a view of what is thought through them. In the latter respect they are called concepts [Begriffe], which word, by designating something that is grasped [das Begriffene] enjoins us to abstract from the way in which we receive, produce or reproduce the thought...It would be a bad mistake to expect from logic any trace of an investigation of [mental laws]...In logic it is necessary to ignore everything that is psychological (1813, 67f.).

A criticism of Kiesewetter et. al. at once follows: Logical relations hold between what is thought; nothing is implied about our ability to think these contents. This is a good start. Herbart simply states that a and non-a are contradictories, one and only one of which will hold. (I believe that he thought a to be an attribute of something, and non-a its complement, but it does not much matter for our present concern.) He does not attempt to link this with a mental capacity. It follows from his view that there cannot be two like concepts, because “they would not differ in what is thought, that is, differ as concepts” (ibid.). This, it turns out, is the antecedent clause of Frege’s claim that: ...if the content of the sentence 2+3=5 is exactly the same, in the strictest sense, for all those who recognize it to be true, this means that it is not the product of the mind of this person and the product of the mind of that person (1979, 4).

The consequent, i.e., that a sentence is not the product of the mind, is not found in Herbart, who simply admonishes us to “abstract” from the fact that a sentence is before the mind, and to concentrate on its “object”, whose nature is not further discussed. Had he been more radical he might have avoided the residual psychologism found in his syllogistics. He thought that not only the first, but also the second and third, figures represent a natural progression of thought. Reduction to the first figure is a “conceit”: Reduction has not improved the presentation of the progress of thought through which conclusions are drawn in actuality (1813, 103).

In a world of propositions and arguments, of what is thought, there can of course be no “progression” of thoughts; there is no mind that moves from one sentence to another. Hence relations of logical consequence, etc. must be described without reference to mental ongoings. The avowed and committed psychological school carried on in the person of Friedrich Eduard Beneke (and some others). Beneke, an altogether admirable man, was deeply disturbed by the isolationist and antiscientific attitude of German philosophers, and the deep decline of the subject in his country. Abroad, so he says, philosophers — the Scots, the French, the Italians — read each others’ work, but Germany has left this community and is separated from it by insurmountable barriers. “In Germany there are a half

38

ROLF GEORGE

dozen heads that form schools here and there and read nothing but their own writings”(1832, 8). He quotes Mme. de Staël: “In Germany you are out of work if you don’t study the universe” (10). No philosophical journal in the last twenty years has lasted beyond the first issues, there have been no translations, not even extensive reviews, of foreign works; the number of books is declining and total bankruptcy of philosophy must be feared (9). This is a contemporary account of a period which later reconstructions made out to be a high point in the history of German philosophy. Beneke can only lament that natural scientists “already look with contempt” upon philosophy (9). He himself wants to follow the scientific method and “naturalize”, as we now say, not only epistemology, but also logic. And this meant, of course, that he wanted to develop it on the basis of psychology. He thought that psychology had attained the status of a science, and that it would shortly equal or even surpass the attainments and rigor of the physical sciences. What he had in mind was a then common attempt to give mathematical models of the force of representations, and the resulting forces when they are combined or replace each other. This new scientific approach would make possible what could not be done when the science, in the preceding century, was in a rhapsodic and poetical phase, namely, to ground logic upon it. From a later vantage point, we can easily detect the strands of psychology in the logic of Kant and his followers, and in Herbart. But for Beneke even the attempt to develop logic apart from psychology is misguided. “Logic was justly called the heart of philosophy, but its healthy life depends upon the psychological - genetic approach to its problems” (21). His psychologism is not, as it were, inadvertent. Despite this, his logic of 1842 is a creditable work and does not differ all that much from his contemporaries. In particular, it is astounding how little is actually drawn from psychology. He engages in a polemic with Herbart which rests once again on the inclusion of the “art of discovery” in logic, notwithstanding Kant’s efforts to draw a line between them. According to Beneke, Herbart had made syllogistics the core of logic. He therefore had to set aside and ignore just those reasonings — the most important — that actually go beyond what is given in the premisses. Discoveries are the result of combination and recombination of concepts (a kind of lateral thinking) and stand apart from the purely analytical procedures of traditional logic. “From this follows the important claim that logical thinking has no part in all those inferences...through which our knowledge is actually enlarged” (1842, 23, 265f.).

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

39

VII The first logician to conceive of logic as a matter wholly apart from psychology was Bolzano. He did, however not neglect the old concerns. Of the four volumes of his Wissenschaftslehre only the first two (and not all of them) deal with the objective world of propositions in themselves. The third is epistemology, dealing with the manifestation of propositions in the mind: a judgment, in contrast to a proposition, which is abstract and mind independent, now is a mental episode whose “matter” is a proposition in itself. In this part of the work he discusses all those issues that tended to be mixed into the discussion of logic itself: clarity and obscurity of representations, knowledge and error, as well as the “art of discovery” which now has its proper place as a part of epistemology. The last volume, finally, is given over to the presentation of a science in the form of a treatise of the subject. This is the old “methodology”, the theory of combining discovered truths into the system of a science. By a proposition [Satz an sich] Bolzano means “any statement that something is or is not the case, regardless whether someone has put it into words, and regardless even whether it has been thought” (WL, §19). He does not call this a definition, but an orientation and, indeed, thinks that he cannot provide a definition that meets his exacting standards. He denied that propositions “exist”, and claims that none of the of the ordinary nouns indicating existence are applicable: they have neither Sein, nor Dasein, nor Existenz, nor Wirklichkeit; he wants to say merely that “there are” propositions (ibid.). The distinction between these statements is that “exists” is construed as a predicate, usually implying location in space and time, while “there is” roughly corresponds to the quantifier, indicating that the expression has a referent (WL §§137, 142). He wants to commit himself to no more than e.g. the mathematician who speaks about a formula that generates all prime numbers. The example is chosen because such a formula is so far unknown, and no one has actually thought it. Possibly there is no such formula. In this case the designator ‘formula that generates all prime numbers’ is “objectless”. Nonetheless, the mathematician who asks if there is such a formula indicates thereby that he thinks that there are propositions. Again, Bolzano wants to be committed to no more than a man who thinks that there truths not yet known. In this way we quite commonly use locutions suggesting that there are propositions.18 Eventually he claims that: Once it is agreed that it is necessary or even simply useful to speak of truths in themselves, i.e. of truths irrespective of whether they have been recognized by somebody, and especially to speak of the relations among them, it will not be denied that the concept of propositions in themselves deserves to be introduced into logic (WL §20).

40

ROLF GEORGE

It is important to note that he will deal “especially” with the relations between propositions. To this we must add his equally prominent concern about their internal structure. I am inclined to say that, except for the rather casual remarks about their being, these are his predominant concerns in the first two volumes.19 Just a few words about the theory Bolzano develops. In contrast to all other theories before then, and most that followed, he defined the terms of propositions (he called them representations or ideas [Vorstellungen], but they are not in any mind) as any part of a proposition not itself a proposition. There are difficulties with this I shall ignore, and only point out that the order of definition is the opposite of what was common. The usual routine was to deal with terms first on the grounds that they are earlier in the order of cognition, making the order of definition in logic track that of knowledge acquisition. This was still done by Lotze: we first have sensations, then we attach names to them, then we combine them into judgments — the common line that judgments come about by a mental act of combination (Lotze 1888, 13 ff.).20 A further enormously important innovation is the introduction of variables, allowing a clear definition of logical form: the form “with respect to variable ideas A and b” of the sentence ‘A has (property) b’ (as Bolzano writes) is the set of sentences generated from ‘A has b’ by making all grammatically appropriate substitutions on ‘A’ and ‘b’. Grammatical appropriateness is not defined, recursively or otherwise, but remains a seat of the pants thing. With the use of variables certain logical properties can now be defined: a sentence M is analytic with respect to a variables i, j if every substitution for them generates a true sentence (WL §148). This generalizes the Kantian definition beyond the subject-predicate form to sentences of all forms. But it allows cases where only some, but not all, extralogical terms are varied. For instance ‘A depraved man does not deserve respect’ is analytic with respect to ‘man’. A sentence is “logically analytic” if it is true under variation of all its extralogical terms. Consequence, in anticipation of Tarski (1936), is famously defined thus: A sentence P follows from M,N,O with respect to i,j,k if every substitution on these “variable ideas” that makes M,N,O true also makes P true (WL §155). If all the extralogical terms are varied, then the consequence is logical, otherwise enthymematic.21 There is no talk here of the mind “progressing” from premisses to conclusion. One other issue must be mentioned. Bolzano thought that there is a relation of ground and consequence between true propositions. No definition of this relation is attempted; it is, rather, introduced by way of examples. Suppose that it is hot, and that a thermometer registers high. While either of these (suitably formulated) may be deduced from the other, the first is the ground, the second the consequence and not the other way round. The schoolmen had similarly distinguished demonstrationes quia from

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

41

demonstrationes propter quid. If the propositions are empirical, then ground and consequence correspond to cause and effect. In mathematics it is not so simple. He holds that two circles intersect because they have a point in common; it is wrong to say that they have a point in common because they intersect. This is not as absurd as it seems, having to do with the order of demonstration in a mathematical system, which must reflect an objective order of propositions.22 This objective order, he conjectures at one point, may be none other “than the ordering of truths which allows us to deduce from the smallest number of simple premisses the largest possible number of the remaining truths as conclusions” (§221). Bolzano not merely postulated propositions and truths in themselves, but also an objective order of them, a guiding thought that motivated his foundational work in mathematics and lead to his singular achievements in analysis, geometry and arithmetic. On the role of psychology in logic he had remarkably little to say. In contrast to Frege and Husserl, there is no polemic to speak of against those who had thought to found logic upon psychology. Early in the Wissenschaftslehre the question is raised if logic is an independent science, specifically if, as many have claimed, psychological theses [Lehrsätze] can have a place in it. With typical thoroughness Bolzano first asks what it means for a science to be independent of all other sciences. A science, for Bolzano, is a group [Inbegriff] of a certain type of proposition, viz. those that “deserve to be collected together in a treatise” (§1). A science A is independent of another B, if no proposition from B needs to occur in A. Given this, it turns out that for instance geometry, the very paradigm of a rigorous stand-alone science, includes propositions from arithmetic and analysis. Many of its proofs require theorems that have nothing to do specifically with space, but with magnitudes in general. Just so we must expect that certain non-logical theses will be found in logic (§13). He will not deal with the question from which other sciences they are borrowed, but we may suppose that arithmetic is again in his mind, having a role in probability theory and elsewhere. The only science specifically mentioned that we must expect to provide theses of importance for logic is psychology. This should not surprise. As mentioned at the beginning, after Bolzano has finished the development of his abstract logic, he turns in the third volume to epistemological matters, the art of invention of old, discussing the problems of how truths may be discovered, how causes for given effects, effects for given causes may be found, and so forth. In §13, he says: Logic is to teach us rules by which our knowledge can be organized into a scientific whole. To do this, it must also teach us how truth may be found, error discovered, etc. It cannot do the latter without attending to the way in which the mind acquires its ideas and knowledge. The proofs of its rules and theories must therefore make reference to the faculty of representation, to memory, the association of ideas, imagination, etc. But the human mind and

42

ROLF GEORGE its faculties are the subject of an already existing science, namely empirical psychology. From this it follows that logic is dependent at least on psychology, and that it must forego the reputation of being a completely independent science.

Does Bolzano here commit the sin of psychologism? I believe not. It has become customary to refer to the discovery project here at issue as Bolzano’s epistemology, and this is surely correct. He follows, but argues anew for, the old custom of making discovery one of the tasks of logic. This is no more than a terminological issue; in the properly logical part, the theory of concepts, propositions and arguments, psychology plays no role. For the most part, Bolzano ignored other philosophers’ attempts to ground logic in psychology, attending to what is left of their claims when the psychological overburden is stripped away. For example, he discusses at some length the common treatment of the so-called fundamental or most general laws of all thought, among them the “law of contradiction”. One version of this is that an object does not both have and fail to have an attribute “at the same time.” Bolzano remarks that this evidently only applies to objects that are in time, which is an undue restriction. But even for temporal objects, the proviso “at the same time” is misleading. He suggests, instead, to locate time determinations in the subject of the sentence: if Caius was not skilled at t1, but skilled at t2, then the attribute ‘skilled’ is applied to two different subjects. But in the law of contradiction the topic must be precisely one subject, Caius-at-t1. Hence the rider “at the same time” is superfluous and misleading. If this correction is made, the law of contradiction becomes “If an object has an attribute then it does not lack it.” The reformulated “law of identity” becomes “An object has the attributes it has.” If properly construed, these sentences are surely true. But, in my view, one can by no means say that they contain “the ground of all truth in thinking, and are sufficient to determine the truth and correctness of all our thoughts.”23 For they contain, as far as I can see, neither the objective, nor even some subjective, ground for the deduction of any truth worth talking about, let alone of all truths...I shall later explain what I mean by the objective ground of a truth... (§45).

And so he does (§168). If some claim is advanced as the ground of all truths, then these truths must follow from it, and nothing of note follows from the laws of identity and contradiction. This was clear to him because he clearly understood, and had advanced the first clear and viable definition of, logical consequence. Bolzano’s treatment of Kiesewetter on the subject of the basic laws is most kind. Kiesewetter had said (I now use Bolzano’s paraphrase) that “a manifold [Mannigfaltiges] that contradicts itself cannot be united in a unity of consciousness, is not thinkable” (§45.4). But, says Bolzano, to be

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

43

thinkable can only mean to be non-contradictory, so that the law of contradiction means that whatever is contradictory is contradictory, “a mere tautology” (ibid.). His strategy, here as elsewhere, is to ignore the psychological thesis that is put forth, and to attend only to the logically significant claim that might be left if the psychology is stripped away. In this case he finds nothing, more often his charitable construals unearth hidden sense of which even the authors must have been unaware. Bolzano’s overall strategy, in his “pure” logic, is to develop the theory, acknowledge antecedents, but not to be overly critical of other philosophers’ foibles (except for Hegel).

VIII A typology of psychologisms must note, first of all, that the term does not occur exclusively in connection with logic. It has been used, in contrast to the “sensualism” of Locke and Condillac, to describe the philosophy of Kant,24 and has also been applied to the philosophy of Beneke and Fries, this time in contrast to the “criticism” of Kant (Erdmann 1904, 2.396 ff.). In these contexts it was not a term of derogation, as it became with Frege and Husserl. As a derogatory term it also occurs in sociology, for instance, to describe the (presumably illicit) attempt to reduce social phenomena to psychological ones. As concerns the intrusion of psychology into the field of logic, our survey teaches us to demarcate several varieties of psychologism. There is first the eliminative psychologism of Locke and his followers, whose aim it was to replace logic with the empirical investigation of inferential habits, or perhaps with nothing at all. From this we may distinguish the unintended psychologism of Herbart, for instance, who let psychological thought creep into the discussion of logic even as he earnestly tried to avoid it. Then also, there is the psychologism of transcendental philosophy, where psychology intruded even as it was hotly denied and cleverly argued that it did not. In all these cases “psychology” refers to rather vague ruminations about the workings of the mind when it thinks. Except perhaps for associationism, it is not even good science. We might use the term scientific psychologism to describe the project intended by Beneke and Fries, to base logic on a properly constructed scientific psychology. A common element in all but the eliminative variety is their reductionism, viz. of logical relations to psychological ones. This less detailed and general understanding of the term is the now common one. Traditionally connected with abstract logic, earlier thought part of it, were further concerns, namely its corrective or therapeutic role, its role in discovery, and its connection with actual thought and inferential habits. In these allied concerns, as Bolzano pointed out, psychology clearly plays a

44

ROLF GEORGE

role though it is out of place in the business of validating the propositions of logic — or metalogic. Historically, postulating a realm of propositions or Gedanken was the only successful strategy for focusing on logic itself and apart from the thought episodes in which propositions might manifest themselves in a mind. Ironically, this abstract approach enormously improved even the discussion of the three applied aspects. For example, Bolzano was able to show that no previous attempt at understanding what a judgment is had been successful or even illuminating. He introduces the term by way of an “orientation”, that is a series of contexts, asking to consider the expressions “to assert, to decide, to believe, to opine, to hold true” and to focus on what their meanings have in common: this common element, he says is “to judge” (urteilen). Further, every judgment contains a proposition, it exists in time (in a mind), and a distinction must be made between judging, and representing, i.e., referring to, a proposition: a proposition has a different presence in the mind when I judge (it) than when I refer to it, and so on (WL §34). There are difficulties with this, but Bolzano’s “orientation” is far superior to what had gone before. He shows that the usual approach tended to rely on metaphors: a judgment results when two representations are “brought together”, “combined” “united in one consciousness” and the like, and then notes that none of the earlier accounts actually succeeded in distinguishing judgments from representations, concepts or terms (WL §35). Propositions are called the “matter” of judgments — a view that has its own problems, but I shall not dwell on them. I hope to have clarified in this paper at least some of the strands of psychologism that ran through the history of logic between Bacon and Bolzano. Much had to be left out. My thesis — if I may be said to have argued one — has been that there were different kinds of intrusion of psychology into logic, some due to a conception of logic that included much of what is now assigned to other fields, others due to cultural and ideological persuasions, and still others to the obsession that logic is the science of thinking. Department of Philosophy University of Waterloo, CANADA NOTES 1

Arnauld 1964, 29 f. I do not know if Arnauld was the first to think of the subject in these terms. It is certain, however, that his division became very influential. There are a number of logic texts in Germany, for instance, that bear titles like “Kunstlehre des Denkens.” 2 Locke was not consistent in his opinions about the function and utility of syllogisms. In the Second Vindication of the Reasonableness of Christianity of 1697, he calls it “the true touch stone of right arguing.” For more on this cf. Fraser’s note in Locke (1959), Vol. II, 397 f.

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

45

Charles W. Hendel discusses the influence of Arnauld on Locke and Hume in the Introduction to the Art of Thinking, Arnauld (1964), xviii-xxii. 3 It is worth noting that there was no single, monolithic logical tradition in European philosophy. On the continent, Peter Ramus’s ferocious criticism of Aristotle, and the Humanist’s attacks on the inelegance of the received logic (as compared with Cicero), lead to a decline in the teaching and practice of the subject. Oxford, on the other hand, soldiered on with disputations and syllogisms, taking note, it seems, of these developments, but showing little enthusiasm in following them (Cf. Ashworth 1988). 4 Tyrell to Locke, April 1704. The incident is described in Cranston (1957), 465 ff. 5 I thank Paul Rusnock for calling my attention to, and translating the passages from Diderot. 6 Bolzano’s Wissenschaftslehre (1837) will be cited as WL. References are to the paragraphs, which can be found in Bolzano 1837 (and reprints), 1969, 1972. 7 Heimsoeth (1912), 202 f. Cited from Arndt (1965) 35. 8 Zedler’s Grosses vollständiges Universal-Lexicon, Leipzig 1738. Quoted from Arndt (1965), loc. cit. 9 The attack on logic did not seem truly convincing without an assault on the character of Aristotle as one who “seems to have had a greater passion for fame than for truth” (ibid.) and who was most unfair to his predecessors and contemporaries. Here is Thomas Reid’s summary of Bacon N.O. 1.67 in Reid (1843), 554: “After the manner of the Ottoman princes, says Lord Verulam, he thought his throne could not be secure unless he killed all his brethren...The faults we have mentioned are such as might be expected in a man who had the daring ambition to be transmitted to all future ages as the prince of philosophers;...and who was not very scrupulous about the means he took to obtain this end.” It is astonishing that Fraser attributes this view to Reid himself (Locke 1959, Vol. II, 391), who rather took exception to the common Aristotle bashing. 10 See H.W. Arndt’s outstanding introduction to Wolff’s book (Arndt 1965) 44. Arndt refers to Descartes, the Port Royal Logic and other sources to show that the analytic-synthetic method was widespread in the 17th and even early in the 18th century. Wolff refers to it in Sect. 885 of his Latin Logic of 1728. Pappus’s mathematical writings were first published in Latin Venice 1594. 11 Wolff (1841) 136, quoted from Arndt (1965) 20. 12 Wolff had earlier thought that syllogisms had nothing to do with discovery, but Leibniz, it appears, put him on a different track when he wrote him in a letter “I would not dare to claim, absolutely, that the syllogism is not the medium for the discovery of truth.” (Non ausim absolute dicere, syllogismum non esse medium inveniendum veritatem.) Letter to Wolff of Feb. 21, 1705. Leibniz (1860), 18. Cf. Arndt (1965), 20. 13 The term ‘epistemology’ occurs first in 1856 James Ferrier’s Institutes of Metaphysic, 48: “This section of the science is properly termed the Epistemology. It answers the general question, “What is Knowing and the Known?” or more shortly, “What is Knowledge?” Similarly, the German ‘Erkenntnislehre’ or ‘Erkenntnistheorie’ does not make its appearance until the 1830s; cf. Fries 1971, X, and the introduction to that edition 16. Note that, for example, Malebranche’s epistemological and metaphysical treatise Recherché de la vérité bore the Latin title Tractatus de arte inveniendi, an obvious reference to the discovery project. The work of Hume and Condillac (whose Logique contains no logic) and that of many other epistemologists, can be seen as standing in the tradition of the ars inveniendi. 14 I thank my colleague Jennifer Ashworth for contributing this information from the inexhaustible fund of her knowledge of Oxford logic teaching (and many other things). 15 It is worth noting that after 1762 or so he also did not write down a lot of reflections on metaphysics or other abstract matters. His more recent biographers think that his attention was riveted on one Maria Charlotta Jacobi, the reason perhaps for his remark, next to a paragraph in Baumgarten’s Metaphysica, that “the absolute perfection of the world does not consist in the multitude, order, and variety of substances, but in the greatest and purest sensual pleasure” (Refl. 3804, Akad 17.298). In that period he wrote down far more remarks, by the

46

ROLF GEORGE

numbering and paragraphing of the Academy Edition, about sex and the relation between the sexes than on logic and metaphysics combined. A sample: “The path to a good marriage goes through promiscuity” (Akad. 20.69) and “we want them chaste, but ourselves are not” (Akad. 20.86), which gives women (specifically apparently Mme. Jacobi, a married woman) the right to escapades. 16 A54/B78; cf. also what he says in the Jäsche logic, Kant 1992, 532. 17 The logics of Kiesewetter and Jakob have nearly as much claim to Kantian authenticity as does the Jäsche logic, which was published under the master’s name. Both were Kant’s students and neither thought much of Jäsche’s efforts. In any case, a good deal of what we know about Kant’s logic comes from student transcripts. 18 In (1837, §21.3), Bolzano paid special tribute to Leibniz for having made very nearly the same point in his Dialogue “De connexione inter verba et res” of 1677, a short segment of which is worth repeating: “A: Do you agree that this [the truth that the circle encloses more area than any other plane figure] is true even if you were never to think of it? B: Certainly, and even before geometricians had proved it or men observed it. A: So you think that truth and falsehood are in things, and not in thoughts? B: By all means. A: But can a thing be false? B: I do not think that things are false, but thoughts and propositions about things...[B gets entangled] B: You have confused me. A: Yet we must attempt to reconcile the two views. Do you think that all the thoughts that can possibly be made are actually formed? Or, to put it more clearly, do you think that all propositions are thought? B: I do not. A: It seems then that truth is a quality is a quality of propositions or thoughts, but of possible thoughts, so that what is certain is only that if anyone should think this way or in the opposite way, his thought would be true or false” (Leibniz 1969, 182 f.). Bolzano notes that Leibniz used ‘proposition’ and ‘possible thought’ interchangeably (1837, §21.3). 19 I do not find it in the least unreasonable to emphasize relations without much concern for the ontological status of the relata. It is, in fact, a common thing, for example when it comes to money. We find no difficulty in thinking about money earning interest, we understand positive and negative balances, deficit and debt, purchasing power and so forth without worrying about the ontology of the units of currency among which this panoply of relations hold. In fact, there is nothing to them except these relations. I note in passing that the bills and coins in the bank’s vault are not money (not M1 nor M2, etc.); some of them become money when an account is debited — perhaps it is better to say that they betoken that an account has been debited. Units of currency (not the coins but what they stand for) are abstract objects, failing most reality tests one can think of. Indeed, there are abstract objects all around us — the usual stock examples, sets, numbers, properties are distinguished only by the long history of philosophizing about them. I say this to generate sympathy for Bolzano’s repeated entreaties that postulating sentences in themselves is no big deal. What really matters is their structure and the relations among them and to other things (their truth conditions, for instance). 20 The first section of Ch. 1 of his logic is titled “The formation of impressions into ideas” and goes on from there. This view, and Frege’s criticism of it is well laid out in Carl (1994). 21 I have argued this in George (1972. For more on Bolzano’s logic see George (1984, 1986), Berg (1962), Morscher (1973) and numerous references in Bolzano (1969) 22 This is discussed at somewhat greater length in the Introduction to Bolzano (1972), xxxvii. Cf. also George and Rusnock (1996), 463. 23 Bolzano had a relaxed approach to quoting. This is more paraphrase than quotation, apparently from Reimarus (1776), §17. 24 The OED cites Ueberweg. Cf. the entry “Psychologism” in the second edition.

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

47

REFERENCES

Arnauld, Antione. The Art of Thinking. Translated and edited by James Dickoff and Patricia James. Indianapolis: Bobbs-Merrill, 1964. First published 1662. Ashworth, E. Jennifer. “Oxford”, in Grundriß der Geschichte der Philosophie (Ueberweg), Vol. 3. Edited by Jean Pierre Schobinger. Basel: Schwabe, 1988, 6-9. Bacon, Francis. Novum Organum in The English Philosphers from Bacon to Mill. Edited by Edwin A. Burt. New York: Modern Library, 1939. Beneke, Friedrich Eduard. Kant und die philosophische Aufgabe unserer Zeit; eine Jubeldenkschrift auf die Kritik der reinen Vernunft. Berlin: E. S. Mittler, 1832. Beneke, Friedrich Eduard. System der Logik als Kunstlehre des Denkens. Berlin: Duemmler, 1842. Berg, Ian. Bolzano’s Logic. Stockholm: Almqvist and Wiksell, 1962. Bolzano, Bernard. Wissenschaftslehre (WL). Sulzbach: Seidel, 1837. 2nd edition Leipzig 1929-31, reprinted Aalen (Scientia) 1970. Bolzano, Bernard. Gesamtausgabe. Stuttgart: F. Frommann, 1969 —. Bolzano, Bernard. Theory of Science. Translated and edited by Rolf George. Berkeley and Los Angeles: University of California Press, 1972. Born, Friedrich Gottlob. Versuch über die ursprünglichen Grundlagen des Menschlichen Denkens. Leipzig: Barth, 1791. Reprinted Brussels (Culture et Civilisation) 1969. Calker, Johann Friedrich Augustran. Denklehre: oder Logik und Dialektik. Bonn: Eduard Weber, 1822. Carl, Wolfgang. Frege’s Theory of Sense and Reference. Cambridge: Cambridge University Press, 1994. Clarendon, Edward Earl of. “Dialogue...Concerning Education” in A Collection of Several Tracts. London: Woodward and Peele, 1727, 31348. Cranston, Maurice William. Locke. New York: Macmillan, 1957. Descartes, René. The Philosophical Works of Descartes. Edited by Haldane and Ross. New York: Dover, 1955. Diderot, Denis. “ Logique” in Encyclopedie, ou, Dictionnaire raisonne des sciences. New York: Pergamon, 1969. First published 1751 ff. Erdmann, Benno. Geschichte der neueren Philosophie. Leipzig: Breitkopf, 1904. Esser, Wilhelm. System der Logic. Elberfeld, 1823. Frege, Gottlob. Posthumous Writings. Edited by P. Long, R. White, and R. Hargraves. Oxford: Blackwell, 1979.

48

ROLF GEORGE

Fries, Jakob Friedrich. Grundriß der Logik (1827) and System der Logik (1837). ed. Gert König and Lutz Geldsetzer. Aalen (Scientia) 1971. George, Rolf. “Enthymematic Consequence”, American Philosophical Quarterly, 9, 1972, 113-16. George, Rolf. “Bolzano’s Consequence, Relevance and Enthymemes”, Journal of Philosophical Logic, 12, 1983, 299-318. George, Rolf. “Bolzano’s Concept of Consequence”, Journal of Philosophy, 1986, 558-64. George, Rolf and Rusnock, Paul. Review of J. Alberto Coffa, The Semantic Tradition from Kant to Carnap, Philosophy and Phenomenological Research, 56, 1996, 461-68. Hegel, Georg Wilhelm Friedrich. Wissenschaft der Logik (1812), in Gesammelte Werke Vol. 11, Hamburg (Meiner) 1978. Heimsoeth, Hans. Die Methode der Erkenntnis bei Descartes und Leibniz, Gießen, 1912. Herbart, Johann Friedrich. Lehrbuch zur Einleitung in die Philosophie (1813, 1821 1834, 1837), Sämtliche Werke, ed. Karl Kehrbach und Otto Fluegel, Langensalza, 1891. Reprint Aalen (Scientia) 1964. Hume, David. Enquiries. Edited by P.H. Niddditch. Oxford: Oxford University Press, 1975. Husserl, Edmund. Logische Untersuchungen. Vol 1, Halle: M. Niemeyer, 1900. Jakob, Ludwig Heinrich. Grundriß der allgemeinen Logik und kritische Anfangsgründe der allgemeinen Metaphysik. Halle: Hemmerde and Schwetschke, 1900. Reprinted Brussels (Culture et Civilisation) 1981. First ed. 1788. Kant, Immanuel. Logic. Edited and transalted by Robert S. Hartman and Wolfgang Schwarz. New York: Dover, 1988. Kant, Immanuel. Lectures on Logic. Translated and edited by J. Michael Young. Cambridge: Cambridge University Press, 1992. Kant, Immanuel. Gesammelte Schriften, Akademieausgabe, Berlin und Leipzig: Walter de Gruyer, 1902; cited as Akad m.n, where m is the volume, n the page. Kiesewetter, J.G.C.C. Grundriß einer allgemeinen Logik nach Kantischen Grundsätzen. 2 Vols (pure and applied logic). Leipzig: Köchly, 1824/25. Reprinted Bussels (Culture et Civilisation) 1973. First ed. 1793. Leibniz, Gottfried Wilhelm. Briefwechsel zwischen Leibniz und Wolff, ed. C.J. Gerhard. Halle, 1860. Locke, John. An Essay Concerning Human Understanding. Ed. Alexander Campbell Fraser, 2 Vols. Oxford: Oxford University Press 1894. Reprinted New York: Dover, 1959. Lotze, Hermann. Logic in Three Books: of Thought, of Investigation and of Knowledge. Translated by several and edited by Bernard Bosanquet.

PSYCHOLOGISM IN LOGIC: BACON TO BOLZANO

49

Oxford: The Clarendon Press, 1888. Reprinted New York: Garlandm 1980. First German edition 1874. Meier, George Friedrich. Auszug aus der Vernunftlehre. Halle: Gebauer, 1752; reprinted in Kant, Akad 16. Morscher, Edgar. (1973): Das logische An-sich bei Bernard Bolzano. Munich: Pustet, 1973. Platner, Ernst. Philosophische Aphorismen, Leipzig: Schwicker, 1793. Reprinted Brussels (Culture et Civilisation) 1970. Reid, Thomas. “Analysis of Aristotle’s Logic” published as an appendix to Essays on the Intellectual Powers of Man, ed. G.N. Wright. London: Tegg, 1843, 553-600. First published as an appendix to Home, Henry, Sketches of the History of Man. Edinborough, 1774. Reimarus, H.S. (1776): Vernunftlehre. Hamburg (Bonn). Tarski, Alfred. “On the Concept of Logical Consequence”, in Tarski, Logic, Semantics, Metamathematics, translated by J.H. Woodger. Oxford: Oxford University Press, 1956, 408-20. German version in Actes Congres International de Philosophie Scientifique VII, Paris 1936. Thomasius, Christian. Einleitung in die Vernft-Lehre, worinnen durch eine leichte, und allen vernünftigen Menschen, waserley Standes oder Geschlechtes sie seyn, verständliche Manier der Weg gezeiget wird, ohne Syllogistica das wahre, wahrscheinliche und falsche von einander zu unterscheiden und neue Wahrheiten zu erfinden. 2nd ed. Halle 1699. Ward, Seth. Vindiciae Academiarum containing some briefe Animadversions upon Mr. Websters Book stiled The Examination of Academies. Oxford: Leonard Lichfield, 1654. Wolff, Christian. Eigene Lebensbeschreibung, mit einer Abhandlung über Wolff, ed. H. Wuttke. Leipzig, 1841. Wolff, Christian. Vernünftige Gedanken von den Kräften des Menschlichen Verstandes und ihrem richtigen Gebrauch in Erkenntnis der Wahrheit. Gesammelte Werke. 1. Abteilung, Vol. 1. Hildesheim: Georg Olms, 1965. With an indtroduction by Hans Werner Arndt, 5-102.

This page intentionally left blank

CARL J. POSY

BETWEEN LEIBNIZ AND MILL: KANT’S LOGIC AND THE RHETORIC OF PSYCHOLOGISM

INTRODUCTION 1. Background In April of 1865 John Stuart Mill published a highly polemical study called An Examination of Sir William Hamilton’s Philosophy and of the Principal Philosophical Questions Discussed in his Writings. It was a sustained attack on the metaphysical, psychological and logical views of one of the nineteenth century’s prominent Anglophone philosophers indeed a man, in Mill’s words, of enormous “reputation and authority.” The book — Mill’s third longest work — was by his own admission the “chief product of the two years” from 1862 to 1864 (see his Autobiography, 1969, 161). Three more English editions and a French translation followed within six years.1 The book occasioned a maelstrom of polemical thrusts by Hamilton’s supporters and parries by Mill and others. It was a major philosophical event. Today this book is forgotten; it is scarcely read and rarely cited. Moreover, it is neglected precisely because it was an unqualified rhetorical success. Mill wished to discredit Hamilton; and that he did, absolutely.2 Because of this book, Hamilton’s reputation vanished; and with it evaporated all but the most arcane interest in the instrument of its demise. However, Mill’s rhetoric in An Examination did bequeath an enduring legacy to the philosophy of logic. For we live today with a semantic dichotomy that Mill imposed in order to trip-up Hamilton. The fact is that throughout the book Mill pursued a two pronged rhetorical strategy: On issue after issue, he found Hamilton to be not only wrong, but internally inconsistent.3 When he came to the foundations of logic — the study, in those days, of conception, judgment and inference — in order to ground this charge of inconsistency, Mill posited a rivalry between two opposing semantic schools. The two sides are what we would today call referential and cognitive semantic theories. Though famous for his humanized (specifically “psychologistic”) views about logic, when it comes to 51 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 51-79. © 2003 All rights reserved. Printed by Kluwer Academic Publishers, the Netherlands.

52

CARL J. POSY

semantics, Mill himself favored the referential theory, as do many modern philosophers. He then proceeded to berate Hamilton both for choosing the cognitivist side — an untenable “conceptualism” in Mill’s terms — and for simultaneously couching his views in a theoretical language which, Mill insists, smacks of the rival referential position. This rhetorical device, as I said, has influenced the course of logical theory: The referential/cognitive divide is now seen as a standard rivalry in the philosophy of logic. The cognitive position has become the modern “assertibility theory” championed today by Michael Dummett. The rival referentialism has turned into modern “realism” in one form or another. Indeed, the particular version favored by Mill is a forerunner of current “anti-descriptivist” theories of reference. Only in the last few years have philosophers begun to challenge the Millian assertion that referentialism and cognitivism are incompatible semantic theories.4

2. What I Shall Do in this Essay Part I of this essay concerns the semantic foundations of logic. I shall begin with a brief precis of Mill’s assault on Hamilton’s conceptualism and on Hamilton’s supposedly inconsistent conflation of conceptualism with referential truth. This will be Section (1). But then, in section (2), I shall show you that referential and cognitive semantic theories can indeed coexist, and that this fact was known well before Mill’s time. In particular, I shall argue that Leibniz already claimed that the conceptualism which Mill attacks is perfectly consistent with a referential semantic theory. Now Mill himself would reject the Leibnizian view of the foundations of logic. For Leibniz constructed his logical theory from an inhuman, God’seye, perspective; and that clashes with Mill’s own “humanized”, psychologistic approach to these matters. In section (3), however, I will argue that Kant humanized the Leibnizian position — indeed, that on each of the relevant issues, he agreed substantially with Mill — yet Kant’s mature philosophy still maintained the cognitive-referential equivalence in semantics. So the rival conceptions of truth which Mill deemed inconsistent were consistently held by Leibniz in the abstract and by Kant in a humanized empirical adaptation. That shall be my first main point about the rift in the philosophical foundations of logic that was impelled by Mill’s attack on Hamilton. My second point is that Mill’s rhetoric does highlight a certain tension within his own “humanized” philosophical view of logic. This is the topic of Part II of the present essay. I will begin by showing you in section 4 that Kant himself toyed for a while with the Millian separation of referential and assertibilist truth. But I shall point out that Kant found it impossible to sustain that separation while maintaining a humanized approach to semantics. This will explain why Mill founded his own humanism in a non-

KANT’S LOGIC AND PSYCHOLOGISM

53

semantic psychologistic fashion. But in section 5 I will point out that there is a certain philosophical tension in Mill’s non-semantic psychologism. Then in section 6 I shall show that though Kant is sensitive to Mill’s psychologistic considerations, once again he does not fall into the Millian tension.

I. SEMANTICS AND LOGIC 1. Mill versus Hamilton Let us turn then to Mill’s attack on Hamilton’s philosophy of logic, specifically to his criticism of Hamilton’s theories of conceiving, judging and reasoning. In the mid-nineteenth century, as I said, these were still the central concerns of logical theory, and according to Mill, it is the simple definitions of correct concept, true judgment and valid reasoning that run Hamilton aground. Hamilton, says Mill, takes a concept to “exist as a separate and independent object of thought” and thus defines conceiving as a special grasp of these separate abstractions. (Mill 1872, 402) This is the heart of what Mill calls the “conceptualist” point of view, and it is, in his view, the common theme of all Hamilton’s mistakes; not only about concepts, but about judgment and inference and logic as well.5 Worse yet, says Mill, Hamilton also regularly couches his wrong conceptualist doctrines in formulations which confuse those doctrines with true accounts of the same matters. Let us look at these briefly in order: Concepts:

A concept, according to Mill, is a “general notion.” Indeed it is ordinarily an “agglomeration of more general attributes.” But Hamilton adds to this view that claim that a concept is “limited to the thought of what cannot be represented in imagination, as the thought suggested by a general term” (footnote to Hamilton’s edition of Reid; quoted in Mill 1872, 411). This, according to Mill, is a false theory. The truth, as Mill would have it, is that the representation of individuals is always primary. A concept is simply part of the presentation of an independent individual thing, a thing “depicted in imagination” (1872, 391n). As for the general components of a concept, says Mill, “we neither conceive them, nor think them, nor cognise them in any way, as a thing apart, but solely as forming, in combination with numerous other attributes, the idea of an individual object” (1872, 393). All of this “nominalist” emphasis on the individual is supposed to be antithetical to Hamilton’s conceptualism. Yet, quite inconsistently, Hamilton also ultimately accepts a “nominalist” doctrine which sounds almost identical to this Millian position; for Hamilton states that “we find it

54

CARL J. POSY

altogether impossible to represent any of the qualities expressed by a concept, except as attached to some individual and determinate object...” (Hamilton, Lectures III, 125, quoted in Mill 1872, 304). So Mill finds Hamilton’s theory of conception to be an unacceptable conceptualism wrapped in an ill-fitting nominalist mantle. Judgments: Hamilton holds — again wrongly, says Mill — that judging is entirely a “notional operation, consisting in the recognition of some relation between concepts,” the relation, namely, “that one of the concepts includes the other as part of itself” (1872, 427). It is, in the end, no more than conceptual analysis. Indeed, Mill is quick to remark that under Hamilton’s account of judging all true judgments must perforce be analytic (1872, 427-8). And we can make only general judgments. Moreover, Mill notes that in order to allow for any but the most trivial knowledge, Hamilton must also claim that concepts are overfilled with information. He must, for instance, include in the concept of a class “all the attributes which we have judged, and still judge, to be common to the whole class.” How else can he reconcile his conceptualist notion of judgment with the many truths we learn about the world? In Mill’s opinion, this is a mistake that is built upon a prior mistake. Reality, according to Mill, always outruns our meager ability to define concepts.6 On this point Mill is a forerunner of modern an “antidescriptivism”: He insists that the components of our judgments denote their objects quite independently of our ability to describe those objects by means of general concepts. And indeed, our judgments may in fact be wrong, and our concepts can miss their mark (see, e.g., Mill 1841, bk.1 chap. 5). Hamilton’s conceptualism, according to Mill, cannot properly accommodate this possibility. But, says Mill, once again Hamilton is not merely wrong in his views about judgment and truth, he is also inconsistent; for he defines a judgment “as the result of a comparison of concepts, either between themselves, or with individual objects” (1872, 424; emphasis added). This notion of testing concepts against objects is quite incompatible, says Mill, with Hamilton’s conceptualist stand. Once again we have conceptualist doctrine (in this case about judgment and truth) dressed in what Mill considers to be incompatible referentialist garb. Inference: Mill says that Hamilton holds the same variety of contradictory views regarding reasoning: On the one hand, Mill points out that Hamilton’s conceptualism forces him to claim that reasoning is no more than “the

KANT’S LOGIC AND PSYCHOLOGISM

55

comparison of two notions through the medium of a third,” that is, simple conceptual manipulation. Indeed, he must relegate logic to the narrow study of “pure consistency,” an empty “sportive” use of the intellect. Yet on the other hand, he does admit that “Reasoning is a source from which we derive new truths” (1872, 445). Because of the generality of his view of inference, Hamilton believes that the laws of logic must be so wide as to apply even to noumena, were there any such things. This is a view which is at best otiose,7 but in fact again, inconsistent.8 So, in the end, Mill complains that “[i]t would hardly be believed, prior to a minute examination of [Sir William Hamilton’s] writings, how much vagueness of thought, leading to the unsuspecting admission of opposite doctrines in the same breath, lurks under the specious appearance of philosophical precision which distinguishes [Hamilton]” (1872, 430). The first of those “opposite doctrines,” — and the correct semantic theory, according to Mill — is, as I said, a referential theory. Mill himself sums it up as follows: A concept to be rightly framed must be a concept of something real, and must agree with the real fact which it endeavors to represent, that is the collection of attributes composing the concept must really exist in the objects marked by the class-name and in no others. A judgment, to be rightly framed must be a true judgment, that is the objects judged of must really possess the attributes predicated of them. A reasoning, to be rightly framed, must conduct to a true conclusion, since the only purpose of reasoning is to make known to us truths which we cannot learn by direct intuition. (1872, 430)

This is a straightforward and modern picture: The concepts which compose a judgment refer to objects and to the properties those objects might have. This notion of reference is the quintessential semantic relation. But that judgment is true only if the object denoted by the judgment’s subject term in fact has (ontologically, so to speak) the property denoted by the predicate. No knowledge or cognitive state intervenes. The semantic relation — reference — connects judgments to the world, but it is the ontological properties of the world and its components which ground the truth. The world, we may say, goes about its own business, and our epistemic task is to describe it perceptually and to determine by means of valid arguments the facts that we do not observe. Moreover, this correct theory, Mill would claim, is opposed at every turn to that subjective “conceptualism” which bases truth upon our cognitive states and to the corollary theory which asserts that cognition somehow determines reality. Hamilton, says Mill, unflaggingly held to the discredited conceptualist view, but perversely refused to admit that this view conflicts with the correct referential picture.

56

CARL J. POSY

2. Leibniz’s Conceptualism Let us turn now to Leibniz and let me show you that he espoused conceptualist doctrines quite similar to Hamilton’s views, but that he combined these consistently with an ontological nominalism and a referentialist semantics. Leibniz had a single governing image which dictated his doctrines about concepts, judgments and logic. He pictured a vast lattice-work of concepts.9 At the top of this conceptual hierarchy is a set of maximally general, primitive predicates.10 One moves “downward”, so to speak, by combining these basic predicates or their negations into increasingly specific compound concepts. The higher, more general, components are “marks” (Merkmale) of the more specific compound concept and are said to be “contained” in it. A simple predicative judgment is just the application of a higher concept to a lower, more specific one. Singular and General Concepts:

At the very bottom, says Leibniz, are the maximally specific combinations, concepts which are fully determined with respect to each elementary predicate. Serially order these combinations and you get a sequence of “stages,” each complete in itself. That serial arrangement is an individual concept, representing an object (or, as Leibniz calls it, an individual substance). Such a concept is a vast conjunction containing each predicate or its denial for every state of the object. Nothing further can fall under it; it is the ultimate infimum species. Thus such a concept can only occur in the subject place of a simple predicative judgment. It can never be predicated of anything “more specific”, for there is nothing more specific. But when a judgment does have such a complete concept as its subject, it is a singular judgment. The only true individuals are those which are described by complete concepts dictating all the properties of the individual at each stage of its development. “It is the nature of an individual substance or complete being to have a concept so complete that it is sufficient to make us understand and deduce from it all the predicates of the subject to which the concept is attributed” (Leibniz, 1686, VIII). This conceptual completeness is the unity that, for Leibniz, marks a true object. To use Leibniz’s famous example, everything about Julius Caesar (from his birth, to crossing the Rubicon, to his assassination on the Ides of March) is coded in his individual concept.11 This is how conceptualism achieves singularity. A general concept, by contrast, is one which lacks this completeness in one or both dimensions; that is, it leaves some elementary predicate undetermined at some stage. Such a concept, at best, describes a mere compound or “aggregate” and not a true object.

KANT’S LOGIC AND PSYCHOLOGISM

57

Aggregates have an externally imposed, ontologically inferior, “phenomenal” unity; they are not true objects. Leibniz is an ontological nominalist. “Being,” for Leibniz, belongs only to those truly unified entities which are described by complete concepts.12 Nevertheless, the important point for us now is that the difference between singular and general concepts is a question of degree, not of kind: General concepts are simply less complete than singular concepts. Grasping a concept of either variety simply amounts to having a full awareness of all of its Merkmale. And the general concepts are indeed derivable simply by omitting one or more determinate predicates from a complete singular concept. Thus, Leibniz does foreshadow the Hamiltonian doctrine that concepts are composed of more “general attributes”, and he does believe that grasping a concept amounts to grasping its collected Merkmale. Moreover, a Leibnizian concept — be it singular or general — does indeed contain all of the information that there is about its object, just as Hamilton will claim for his own notion of concept a century and a half later. Judgments and Truth: For Leibniz, a simple predicative judgment is true only when the analysis of its subject concept produces the predicate concept: “All humans are rational” is true because the concept human contains the concept rational. And “Caesar crossed the Rubicon” is similarly true because unpacking Caesar’s full individual concept reveals that to be the case.13 That is, as Leibniz would say, the judgment is true when the subject concept contains the predicate concept. Leibniz himself described this “containment theory” as an “intensional” theory of predication, because the set of general concepts that are contained in a given concept is called the “intension” of that concept. This is a classic “conceptualist” view of truth; for it is the component concepts of a judgment which ground the judgment’s truth. In this theory, all truths are indeed analytic in quite a strict sense of analyticity.14 Once again, we have a clear forerunner of Hamilton’s semantics. Logic and Ontology: The great Leibnizian logical principles of the identity of indiscernibles, of sufficient reason and of excluded middle are actually ontological principles about the most general properties of objects. The identity of indiscernible follows from the fact that if two individual concepts are alike in all details, their respective objects must be identical. Indeed, though each individual object has its own particular nature, following the scholastics, Leibniz called this the objects “haecceity” yet there is no hidden aspect of the object’s haecceity beyond what is contained in the individual concept (Leibniz 1686, VIII). The principle of sufficient reason simply records the fact that all

58

CARL J. POSY

changes in an object’s status are already encapsuled in — and thus explained by — the object’s individual concept. Much the same can be said of the logical principle of excluded middle. It, too, follows from the notion of a complete concept. For any given individual concept and any simple or complex predicate, either the predicate belongs to the concept or its negation does. Since, according to Leibniz, a full individual concept will include the relation of the individual with all other elements and aspects of the universe, the law of excluded middle will hold for all judgments about any aspect of reality. Anyone who grasps or could grasp such a concept could in principle determine the answer to any question about the world. So Leibniz, just like Hamilton, espouses a conceptualism which governs both his notion of truth and his most general logical principles. But, at the same time, Leibniz also says that the we should “content ourselves, with seeking truth in the correspondence of the proposition in the mind with the things in question” (Leibniz, 1765, Bk. 4, chap. 5, 3). His is thus a referential semantic theory which bases truth on objects and their interrelations. This, of course, is precisely the conception of truth that Mill favored, — and it is the conception of truth that Mill thought to be incompatible with the intensional theory that Leibniz also held. However, Leibniz’s espousal of a referential theory is not a aberration, nor is it incompatible with this intensionalism. Leibniz can merge these two conceptions of truth because he simply equates what I above called the “ontological” predication with conceptual containment itself. According to Leibniz, this ontological relation which grounds the truth of a judgment is just the relationship “by virtue of which one idea is or is not included within the other” (1765, Bk. 4, chap. 5, 2). For Leibniz it is an “ontological” aspect of an object that its concept be analyzable in the way that it is. Moreover, Leibniz’s view of the semantic “reference” relation itself — the relation between the components of a judgment and their referents — is equally straightforward. When the component in question is a general concept then the referential relation is in fact simple identity. If the component in question is an individual concept, it is related automatically — indeed isomorphically — to its corresponding object. There is no haecceity, recall, beyond the information coded in the individual concept. Thus, in this Leibnizian picture a concept simultaneously can be an agglomeration of general features and the complete representation of a class and even of an individual object. Moreover, truth can be intensional and at the same time referential. All of this is held together by the single premise that there can be no aspect of an object that eludes its individual concept. The intimacy between being and judgment also guarantees for Leibniz, as we have seen, that the laws of logic can be most general and still apply to objects. So these apparently inconsistent doctrines do cohere in a single, consistent, philosophically respectable semantic theory.

KANT’S LOGIC AND PSYCHOLOGISM

59

Human Perception and Truth:

But, we can hear Mill object, those Leibnizian objects are scarcely the things to which we refer. Leibnizian individual substances are monads, indivisible, unextended, atemporal beings. It is Julius Caesar the monad — and not his body — that is described by his individual concept. And only God can grasp that individual concept, or any individual concept, in all its infinite complexity. That, in fact, is Leibniz’s paradigm of “intuition:” God’s direct grasp of all the components of a concept, no matter how complex that concept might be. (See, Leibniz, 1684.) Indeed, from Leibniz’s perspective, we humans have almost no direct knowledge of those true individuals.15 We depend on sensory perception which, according to Leibniz, is merely confused thought. Individual perceptions, he held, are concepts, but deeply inferior ones. For an individual perception is always perspectival: it never presents its object all at once, and it is never complete either synchronically or diachronically. Our human perceptions present only composite (and therefore phenomenal) entities. And empirical concepts, which we create from repeated perceptions, inherit this inferiority. They cannot have more in them than experience has provided. We can have “intuitive” grasp of only the most elementary and most general items, never of a true object. Mill will complain about this Leibnizian theory — quite as he did of Hamilton — that the logic applies to noumena. Hamilton can take little comfort in this Leibnizian demonstration that conceptualism can coexist with a respect for the individual and with a referential semantics. If this is how we reconcile these semantic poles, by demoting perception to the status of confused thought, by expanding logic to encompass humanly inaccessible monads, by denying ourselves any access to legitimate truth, then both the cognitive and the referential semantics — however theoretically compatible they might be with one another — are of little interest to us. Here, enters Kant to humanize Leibniz’s semantic views.

3. Kant’s Critical Semantics The first thing we need to note about Kant is that throughout his career he worked fundamentally within the Leibnizian semantic framework.16 This means in particular that he retained the idea that true objects are complete and a-perspectival.17 And it means that he maintained a cognition-based notion of predication. (An object, Kant says in the Critique of Pure Reason, is subsumed under a concept when the representation of that object is “homogeneous” with the predicate concept [A137/ B176].) But Kant’s central move in his “Critical” philosophy is anti-Leibnizian. Kant lets the notion of human perception play the role in his own philosophy which God’s grasp of an individual concept played for Leibniz. This means

60

CARL J. POSY

that our human empirical judgments based upon those perceptions count as genuinely true and that the subjects of those judgments, perceived empirical things, count as legitimate objects. This new “humanism” dictates Kant’s views on our issues of conception, judgment and logic. I will briefly highlight each of these Kantian views. The result in the end will be a unified theory which meets Mill’s demands, but that still equates cognitive truth and referential truth. Indeed, it will be a theory which foreshadows both the modern referencebased semantics and modern assertibilism. It will be, in the end, a theory challenging not only Mill, but the semantic orthodoxy that has flourished since Mill’s time. A. KANTIAN CONCEPTS

Humanly graspable concepts are for Kant, as they are for Mill, always “general notions.” There are no infima species. (See Kant, 1800, 10 and Critique of Pure Reason A655-6/B683-4.)18 Thus for Kant, my judgment “The cherry tree in my front yard is blooming,” is not automatically a singular judgment. Even were I to incorporate the recognition that there is only one cherry tree in my yard, still that would not render the judgment singular. Only when I accompany it with a verifying perception do I have what Kant calls a “singular use” of that judgment (see 1800, 1, n2);19 for singularity is granted only by perceptions and not by the concepts in a judgment. But, more than that, for Kant, again, just as for Mill, our use of concepts rests heavily upon our ability to depict an individual thing “in imagination.” This is true for the role of concepts in perception; it carries through to more general empirical conception, to abstract mathematics, and even to the nonempirical Ideas of Reason. Indeed, Kant has a very detailed theory of imagination which underlies each of these cases: (1) Perception

Consider the perception that verifies my judgment about the status of my cherry tree. Kant will insist that to say that the manifold of data is arranged in a tree-like fashion I must impose a certain order and unity among its parts. I must group the upper branches, for instance, with the lower trunk as part of one structure, and the pink blossoms and brown bark as belonging to the same object. I must also discount the fact that I noticed the top of the tree before the bottom; for that order is a subjective component of my mental state. These decisions are dictated by the concept tree, and they require that I project all the different ways in which I and others might view this tree and all the various additions that I might have to my knowledge about this tree.

KANT’S LOGIC AND PSYCHOLOGISM

61

In some of these I might view it from another angle or side. In some the bottom will be noticed first; in some the top. That is the sign that this ordering is merely subjective. In some of these projections I might dig beneath the surface and expose the roots; and in some of those the root-ball might have one color, in others, another. Nevertheless, in all, the shape and colors of the exposed parts of the tree will remain as they now appear. That constancy is the sign of an objective grouping. Of course I cannot yet observe any of these projected changes of perspective or additions, and I certainly cannot survey the whole lot of them. So this task is accomplished by the use of “productive imagination.”20 And since each such projection is itself incomplete and completable in different ways, the productive imagination gives me incomplete, “schematic” imaginary tableaux. This is how, for Kant, my fleeting perspectival perceptions can fulfill the Leibnizian requirement for the grasp of a true individual. For Kant, it is not completeness, but completeability that characterizes that cognitive grasp. To be sure, what I see before me now does not tell me how many rings the tree has and thus it does not suffice to distinguish this tree from other potential trees with a different number of inner rings. Nonetheless, actually seeing the tree and locating it in my full perceptual field does guarantee that I could obtain that information if the need arose; and my imaginative projections shape the form that that information can take. My projection of further experiences makes my present state of mind into a legitimately objective presentation. (2) General Empirical Conception

Imagination plays a role here as well. ordinarily we consider a concept (such as the concept tree) by focusing on a single paradigm instance. The instance may be actual or imagined. The paradigm instance represents the set of all possible tree perceptions, including cherry trees, elm trees and cypresses. Concentrating on that instance and abstracting from the characteristics peculiar to this particular tree (its color, size, variety and location) amounts to projecting (again schematically) to the elements of that set. Here the only constant elements would be those which are dictated by the concept tree; for example, that it have wood and foliage, that there is a root system, and that there will be growth rings.21 (3) Mathematics

Imagination is also at the heart of that mix of perception and abstract thought that is Kant’s theory of mathematics. Qua mathematician I once again start with a particular perception and abstract from any features that characterize the specific (actual or imagined) object at hand. In this case I will abstract even from the fact that it is a tree that is perceived and

62

CARL J. POSY

concentrate, for instance, only on the cylindrical shape of its trunk. I might manipulate the image in my mind and draw conclusions from the results of that manipulation. But now the conclusions will be valid for all possible cylinders. Once again, I cannot survey that class, but only schematically project it. (4) Ideas of Reason

The clearest case here is the role of imagination in subsuming an ethical maxim under the idea of the moral law. Kant maintained that the notions of the ethical self and of the moral law are Ideas of Reason, which are not bound by the spatio-temporal limits of human perception. Kant’s test for a morally good maxim is the universalization procedure by which I determine that there would be no contradiction in will or conception if all agents were always to act on this maxim. However, to do this I must consider what the world would be like if everyone were to adopt this maxim. In doing so I must assume that the individuals are free from the bond of empirical causality that characterizes our spatio-temporal world. Now, even if I assume that the maxim is universally adopted, still it can be applied in many different situations. As an ethical deliberator I must survey the array of all these possible situations, and then determine whether any of them (or any that involve myself) are inconsistent. Once again, the full survey cannot be accomplished, and a schematic (and thus an imaginative) depiction of a paradigm case will be necessary.22 So Kant does indeed have a highly developed theory about how all of our conceptual activities rest on our human ability to imagine individual objects and situations. His is, just as Mill’s is, a theory of conception based upon human imagination. B. EMPIRICAL JUDGMENTS AND EMPIRICAL TRUTH

Having spoken of concepts, let us turn to judgments and truth. Here, just as Leibniz did, Kant too will have a theory of truth that combines cognitive and referential components simultaneously. But, as I said, he will do this in a humanized fashion. (1) Assertibilism

Kant has, as I said, a cognition-based theory of truth for elementary judgments. But it is, as I also said, human perception, and not conceptual analysis, that underlies the truth of a synthetic judgment like “The cherry tree is blooming.” The result is that he has moved from the Leibnizian “intensional” truth theory to one much closer to a modern “assertibility” theory, a theory according to which the truth of a judgment is linked to

KANT’S LOGIC AND PSYCHOLOGISM

63

justification (or “warranted assertibility”) by humanly available evidence. John Dewey coined the term warranted assertibility, and the notion goes back in the “Pragmatist” tradition at least to the writings of C.S. Peirce. (See, e.g., Dewey 1938.) But this notion of truth has been exploited most effectively by the school of intuitionistic mathematics founded by L.E.J. Brouwer and has been developed in that vein recently by Michael Dummett. We find clear evidence of Kant’s Assertibilism in his treatment of the “Antinomies” in the “Dialectic” of the Critique of Pure Reason. Thus in the “First Antinomy” Kant has the transcendental realist argue that the “Thesis” claim — the claim that physically occupied space extends infinitely — cannot be true. The premise of this argument is that we finite humans cannot synthesize (and thus cannot perceive) an infinite physical expanse. In a corollary manner, the “Antithesis” claim — that the spatially occupied universe cannot have a finite extent — rests upon the fact that the empty space which would border such a finite world must be infinite (and thus again unsynthesizable). Arguments ad ignorantiam, such as these, are valid only under an assertibility theory of truth. Similar such arguments play a crucial role in some of Brouwer’s most important mathematical papers. (See, e.g., 1975.) Assertibility theory also plays a central role in Kant’s reasoning in the “Analytic” section of the first Critique, where he establishes our right to a priori knowledge of the categorial principles. Those projections of future experience, for instance, which accompany my perception of that tree are not merely plays of imagination. They embody predictions about the variable and constant aspects of my future experience. If my judgment that the tree is in bloom is justified (and thus, true, according to assertibilism) then these presupposed predictions must be justified as well. The categories are designed to guarantee that those predictions can be justified. Thus the category of “quantity” and the related Principle of Extensive Magnitude guarantee that predictions about the constancy of shape and size will be justifiable. Quite similarly it is the category of “cause” and its related principle that help me distinguish between objective and subjective temporal orderings, and that guarantee that I will be able to justify predictions about objective time sequences. This argument, which depends centrally on the equation between truth and justification, is the central core of the “Transcendental Deduction,” and the “Analytic of Principles.” Indeed, assertibilism is an irreducible premise of Kant’s “Analytic”. (2) Reference

For Kant, as we have just seen, Leibniz’s criterion of predicative completeness has given way to “completeability,” and Leibniz’s simple intensionalism, his containment theory of truth, has become an elaborate assertibility theory, that permeates Kant’s Critical writings. But there is one Leibnizian principle that Kant keeps wholly intact: Kant, just like Leibniz,

64

CARL J. POSY

equates the cognitive notion of truth with a referential semantics. “Truth,” he affirms, is “agreement with the object”.23 And the purpose of our human intuition, he says, is to “give” that object. An intuition presents its object, and through that presentation it grounds the truth of our judgments about that object. Thus, for Kant, empirical intuitions are not merely the basic units of evidence; they are the vehicles of reference. And while I spoke of justifying my decision to connect the top of the tree with the trunk, Kant speaks of their “combination in the object” (Critique of Pure Reason B142). That is what underlies the truth of my judgment about the shape of that tree, and that is why such judgments carry “existential import.”24 Moreover, Kant describes the success of the categories in bolstering that truth as their “objective validity” — their application to objects. But even more telling for our present purposes, Kant’s view about how empirical concepts connect with objects is exactly what Mill would want. “[W]hat useful purpose,” Kant asks: ...could be served by defining an empirical concept, such, for instance, as that of water? When we speak of water and its properties, we do not stop short at what is thought in the word, water, but proceed to experiments. The word, with the few characteristics which we attach to it, is more properly to be regarded as merely a designation than as a concept of the thing, the so-called definition is nothing more than a determining of the word. (Critique of Pure Reason, A728/B756)

This is again an “anti-descriptivist” theory of reference. Kant’s point is that we can learn and revise our catalogue of properties while referring to the same object or kind. We may even make mistaken claims about objects and natural kinds. The object or kind — and thus what is true about them — cannot be captured within some conceptually closed definition. Leibniz’s God, with his grasp of complete concepts might be able to define the objects of His judgments, but we humans can only refer. And so, if our human judgments are to be legitimate truth bearers, they must derive their truth referentially — from the objects, so to speak. Thus, we find in Kant, as we did in Leibniz, a union of the referential with the cognitive semantics. Moreover, the two, allegedly rival, semantic approaches coexist in Kant’s philosophy for the very same reason that they do in Leibniz’s; for Kant accepts the Leibnizian confluence of knowledge, truth and ontology: There is no more to the haecceity of the objects than is given in the cognitions of those objects. To be sure, the objects are now extended empirical objects, and the cognitions are now our human epistemic states; but still, in Kant’s view, there are no hidden, unknowable components to an object’s empirical nature. And thus there is no more to truth, defined referentially, than what we can know. This is the heart of Kant’s Transcendental Idealism.

KANT’S LOGIC AND PSYCHOLOGISM

65

C. LOGIC

When we come to Kant’s views on logic we find that referentialism does not merely coexist with Kant’s assertibilism, it actively cooperates. Kant’s assertibilism has a logical consequence — indeed a very modern one — but only in virtue of reference. The logical effect of assertibilism is what we today call intuitionistic logic. Michael Dummett, building on mathematical intuitionism, has argued that the assertibility theory of truth yields a non-classical, “intuitionistic” logic, a logic which rejects the law of excluded middle, (and with it, indirect proof and other marks of classical logic). (See, e.g., Dummett, 1978). The reason for this is that under assertibilism a proposition can be counted as false only if it has been positively refuted. Thus, since some propositions may well elude both proof and refutation, for the intuitionist these propositions will be neither true nor false. Intuitionistic logic provides a powerful tool for explaining many parts of the first Critique. Consider again, for instance, Kant’s reasoning in the “First Antinomy.” We can formalize the thesis as

where the variables range over spatial regions, E(y,x) means that y and x are equidistant from some fixed origin point, and F(y,x) means that y is further removed from that point than is x. And we can formalize the antithesis by

We then can note that the disjunction of (1) and (2) is classically valid, but intuitionistically invalid,25 which corresponds precisely to Kant’s claim that for the transcendental idealist the disjunction of (1) and (2) is not a logical truth.26 However, the path that Dummett sets out from assertibilism to intuitionistic logic only works on the pessimistic assumption that there are, or at least might be, some eternally undecided judgments, some judgments that never get justified but are never positively refuted. For if we could assume, optimistically, that every proposition can eventually be either proved or refuted, then there are really no compelling grounds to refrain from endorsing the law of excluded middle and full classical logic.27 Here the referential component of Kant’s semantics is crucial. It provides the appropriate epistemic pessimism about empirical science. “In natural science,” Kant tells us, “there is endless conjecture, and certainty is not to be counted upon.” But the reason for this attitude is that “the natural appearances are objects which are given to us independently of our concepts, and the key to them lies not in us and our pure thinking, but

66

CARL J. POSY

outside us; and therefore in many cases, since the key is not to be found, an assured solution is not to be expected” (Critique of Pure Reason, A480/ B508). Empirical truths elude us not because (or not merely because) our cognitive epistemic abilities are finite, but because the objects might not be found.28 It only is because our intuitions are “receptive” and because their denotata are “outside” us, that we have this uncertainty.29 Only if we were to view the world from a Leibnizian,“God’s-eye,” perspective could we assume the sort of optimism that would justify the principle of excluded middle and classical logic. That, indeed, is Kant’s reductio of Transcendental Realism. For the realist combines the view that the haecceity of objects is defined by our human abilities and limits with the epistemic optimism of a “God’s eye” perspective. So for Kant — again just as Mill would want — logic is a science with special application to empirical judgments. It is a science that supposes those judgments to be true in virtue of the fact that their terms refer to empirical objects. Yet, for Kant, all of this is quite consistent with (and indeed supportive of) a cognition-based theory of truth. And that consistency is quite opposed to the basic assumption of Mill’s attack, the assumption that these two notions of truth are incompatible. This is the first main point that I promised to make.

II. PSYCHOLOGISM Let us turn now to the question of psychologism. Here is where I believe that there is a tension within Mill’s own thought. And here is where I want most stringently to contrast his views with those of Kant. Psychologism, as we are concerned with it here, is an attempt to include the facts of human cognition as part of the foundations of logic. The psychologistic philosopher will try to show that though logic is our means of moving beyond the individual deliverances of our human experience to a global notion of validity (to “the logical structure of the world” so to speak), nevertheless that science is conditioned by the nature of those individual human experiences. Assertibilist semantics when combined with a similarly “humanized” notion of reference is one way of achieving this humanized effect in the philosophy of logic. That is a combination that Mill rejects. I want to show you next that Kant too, for a while, rejected that combination. Such a separation between sensibly based assertibilism and ontologically weighted reference characterized his view of semantics in the Inaugural Dissertation of 1770 (On the Form and Principles of the Sensible and Intelligible World), the last of his “pre-critical” works. It is instructive to look at this not merely as an interesting stage in Kant’s intellectual development, but —

KANT’S LOGIC AND PSYCHOLOGISM

67

more to our purposes — as showing us the difficulties one will encounter in trying to humanize the foundations of logic while maintaining this separation between reference and empirical assertibilism.

4. Kant’s Pre-Critical Semantics The “sensible world” that Kant mentions in the title of his Inaugural Dissertation is the world of “phenomena” that are presented by our intuitive faculty. The intelligible world is the noumenal world, given, as he says, by the intellect. Though the title and structure of this work promise a strict separation between these two faculties the Dissertation, indeed, already contains Kant’s anti-Leibnizian distinction between human sensible intuition and abstract conception the fact is that for Kant here they thoroughly cooperate in producing human knowledge. We come to know the spatio-temporal aspects of objects intuitively (that is, sensibly); and we know their relations and causal interactions intellectually.30 It is semantically that the faculties are separated. Each has its own appropriate stock of concepts and judgments: Intellectual judgments consist of abstract “metaphysical” concepts, while the concepts which compose sensible judgments are restricted by our finite grasp of spacial and temporal information. And each set of judgments has its own semantics: Intellectual judgments have a sort of Leibnizian conceptualist semantics. While sensible judgments get a version of assertibilism which respects our perceptual limitation. In 11 of the Dissertation (Akad. II, 397), Kant insists that this latter assertibilism is a fully semantic theory. It is a legitimate theory of empirical truth: Truth in judging consists in the agreement of a predicate with a given subject. But the concept of a subject, in as much as it is a phenomenon, would only be given through its relation to the sensitive faculty of cognising, and it is in accordance with the same relation that predicates would be given which were sensitively observable. Accordingly it is clear that representations of a subject and a predicate arise according to common laws, and so provide a handle for a fully true cognition.31

This is indeed a semantic humanism. But he equally strongly insists that reference — contact with actual objects — belongs only to the intellectual semantics. He makes this point ontologically in that same 11: The phenomena presented by our senses are, he says, specious things; they do not “express the internal and absolute quality of objects.” And earlier (in 8, Akad. II, 395) he tells us that they have no metaphysical weight, and thus cannot perform the duty of establishing existence. He tells us quite outrightly positively

68

CARL J. POSY

there, that only the intellect can do that referential job.32 And this difference in referential role will make a difference in the respective powers of these faculties to give the logical structure of the world. In particular, when Kant turns to a discussion of “totality” (an issue which today we designate by the use of quantifiers), he tells us that the sensible faculty is unable to grapple with this notion in an abstract logical sense that goes beyond its everyday encounters with empirical objects. It is tripped up in fact by the question of the sensible world itself as a totality. (The sensible world be its parts taken successively, in time, or simultaneously, in space can be neither a finite nor an infinite totality.) But no such difficulty attends the intellect: [N]either the successive nor the simultaneous co-ordination of several things...pertains to the intellectual concept of a whole, but only to the conditions of sensitive intuition. And so even if these co-ordinations were not to be sensitively conceivable, they do not thereby cease to be intellectual. It is sufficient for this concept that co-ordinates should be given, no matter how, and that they should be thought of as pertaining to one thing. (§2, Akad. II, 392)

The real object itself is not affected by our sensible limitations, and so neither is the logic which characterize it. In modern terminology, he is saying here that the logic of quantifiers is belongs to the intellect; and it is ultimately a classical logic, as we have seen. It is important to note that Kant does not claim any conflict between the assertabilist and referential semantics. Rather, he believed at this point that there is a conflict within the assertabilist side, and he thought that it could by finessed by separating the referential and hence the logical side of semantics. Eventually came to realize that this double-sided but logically imbalanced semantics cannot hold. That is precisely the force of his “Antinomy” argument. The realist of the “Antinomy” is in fact the author of Kant’s Dissertation. He applies a classical logic to get the structure of a world that is known by way of sensibility. Thus, in the Critique of Pure Reason, Kant came to abandon this bifurcated empirical semantics and logic. The logic which underlies the validity of our reasoning and which provides the global structure of the world is, in the Critique, a unified humanized logic. Its assertibilist side is sensibly conditioned, but it nevertheless has full referential power. This combination allowed Kant to avoid the antinomy (by a move, which I showed you in §3 above, amounts to abandoning classical logic). And, keeping in mind the metaphysical weight of reference, he came to say that the things presented by our sensible intuition, though phenomenal, are “empirically real”.

KANT’S LOGIC AND PSYCHOLOGISM

69

5. Mill’s Psychologism Returning now to Mill, let me remind you that this combination is a move that he does not make. He is a referentialist; and that, he insists (wrongly, to be sure, but adamantly) is inconsistent with a cognition-based semantics. Mill will not accept the Kantian assertibilism as a humanizing aspect of his semantics. So, if he is to humanize the foundations of logic, he must take a different tactic. Indeed, it is now fair to ask where in his views about logic lies Mill’s “humanism”? Where indeed is his signature psychologism? The Hamiltonian doctrines that he rejected seem more psychologistic that those advocated by Mill himself. It is Hamilton, after all, who relates the truth of judgment and the validity of reasoning to their success as cognitive processes; and it is Mill who inserts a non-mental element of reference to reality into the definitions of truth and validity. Mill’s psychologism in the philosophy of logic lies not in his definition of truth or of validity, but in his belief about the continuity of logic and psychology as parts of a single scientific enterprise. He chides Hamilton (mercilessly, I might add) for separating these disciplines. Here is Hamilton, quoted by Mill: The phænomena of the formal, or subjective phases of thought, are of two kinds. They are either such as are contingent, that is, such as may or may not appear; or they are such as are necessary, that is, such as cannot but appear. These two classes of phænomena are, however, only manifested in conjunction; they are not discriminated in the actual operations of thought; and it requires a speculative analysis to separate them into their several classes. In so far as these phænomena are considered merely as phænomena, that is, in so far as philosophy is merely observant of them as manifestations in general, they belong to the science of Empirical or Historical Psychology. But when philosophy by a reflective abstraction, analyses the necessary from the contingent forms of thought, there results a science, which is distinguished from all others by taking for its object-matter the former of these classes; and this science is Logic. Logic, therefore, is at last fully and finally defined as the science of the necessary forms of thought. (Lectures on Logic, V. III, p.24; quoted in Mill 1872,458.)

And here are two errors that Mill finds in this Hamiltonian proposal: (1) Hamilton’s conception of logic is altogether too narrow. Logic, for Mill, is a much broader subject. It must be a science of truth not only a science of consistency: If any general theory of the sufficiency of Evidence and the legitimacy of Generalization be possible, this must be Logic and anything else called by the name can only be ancillary to it...It is only as a means to material truth, that the formal, or to speak more clearly, the conditional, validity of an operation of thought is of any value; ...But to call this alone Logic, or this alone Pure Logic, as if all the rest of the Philosophy of Thought and Evidence were merely an adaptation of this to something else, is to ignore the end to

70

CARL J. POSY which all rules laid down for our thinking operations are meant to be subservient. The purpose of them all, is to enable us to decide whether anything, and what, is proved true. (Mill, 1872, 475)

For Mill, logic (both in the small, as a theory of consistency, and in the large, as a general theory of evidence) is an “art”, a system of “precepts” or rules for valid thinking.33 It is the theory of correct inference, including, he insists, inductive inference.34 (2) Hamilton totally misses the fact that psychology lies at the foundation of logic (again, both larger and smaller logic): [Logic] is not a Science distinct from, and coordinate with, Psychology. So far as it is a science at all, it is a part, or branch, of Psychology; differing from it, on the once hand as a part differs from the wholes, and on the other, as an Art differs from a Science. Its theoretic grounds are wholly borrowed from Psychology, and include as much of that science as is required to justify the rules of the art. (Mill 1872, 461)

However, it is this last point — the claim that psychology provides the “justification” for logic’s normative force — that strains Mill’s philosophy here. “Justification” and “theoretic grounds” are epistemological notions. To say that some logical rule or system of is “grounded” is to say that we who use this rule are justified in our beliefs that these rules will lead from true premises to true conclusions and in our trust in the normative force of the rule or rules. And to say that psychology provides the “theoretic grounds” is to say that this justification comes from psychological research. And how, according to Mill, does psychology provide those “theoretic grounds” for logic? How does it justify those rules of art? The answer is that the acts of conceiving, judging and reasoning are, after all, simply mental acts. As such they are the proper subject matter of psychology. Mill says this quite explicitly about the process of concept formation. It is psychology, he says, that gives us the laws of association and habit which govern concept formation. And it is psychology which then describes the laws of attention and imagination by means of which we focus on the concepts thus formed. (See Mill 1872, 466-68.) These laws of course, like the laws of any descriptive science, are established by processes of observation (here including introspection) and induction. Thus, on the one hand, the justification of logic (including deductive logic) is an inductive generalization from the successes that we have had in applying logical rules. Mill is quite explicit about this when he attacks Hamilton’s formulation of logic as giving the laws of the necessary forms of thought. The laws of thought, Mill insists, have the same “contingency” as the laws of any empirical science. Yet, on the other hand, Mill does admit a notion of global, logical necessity that is far stronger than this. It is apparent in his use of the term “validity” in describing the narrow Hamiltonian logic as an “art”. This logic,

KANT’S LOGIC AND PSYCHOLOGISM

71

he admits, “lays down the laws or precepts indispensable to Valid Thought; the conditions to which thought is bound to conform, under the penalty of being invalid, ineffectual, not accomplishing its end” (Mill 1872, 469-70). It is present in his view of contradiction: “Or again, the concept, the judgment, or the reasoning may involve a contradiction, and so cannot possibly correspond to any real state of facts” (1872, 471, emphasis added). And it is quite evident in his belief that there is a special force and a guarantee in the use of deductive inference. “We may abstract from actual truth, but the validity of reasoning is always a question of conditional truth — whether one proposition can be true if others are true” (1872, 472). To be sure, Mill is a bit cagey about this: On the one hand his discussion of psychology and logic lists not only conception but also judgment and inference as the central “logical” activities of the mind. Yet when it come to actually providing psychological description he limits himself to the psychology of concept formation. Nonetheless, a thorough reading shows that the tension is undeniable: Our use of inferential rules (even narrow deductive ones, like modus ponens) is justified in the same way as our use of any concepts. That is to say, modus ponens is justified by showing its origin in earlier repeated successful associations of judgments of the form and A, with judgments of the form B. The psychological story, of course, renders the continued success of this practice as contingent as any other empirical generalization. However, the notion of validity and normative force presumes the necessity of narrow logical inference. This is a real, internal tension in Mill’s treatment of the psychological roots of logic. Kant too is interested in the psychological roots of logic, but he nevertheless manages to avoid this internal tension.

6. Psychology and Epistemology in Kant Like Mill, Kant is concerned with the process of concept formation, and with its role in justification. Concept formation, for Kant, is the basic instance of what he calls “reflective judgment.” “Refection,” he says in the Lectures on Logic (1800), is “the consideration of how distinct representations can be grasped in one conscious thought” (Akad. IX, 94). In reflective judgment “only the particular is given, and the universal has to be found for it” (Critique of Judgment, Akad. VIII, 179). The point is that in reflection one finds the common aspects of distinct perceptual experiences, one abstracts from the factors that make those experiences different, and one then unites these aspects in to a single empirical concept. We have already seen that this “unification” into a single concept involves imagining yet other perceptual experiences in which those “common aspects” will again coincide, and in which the concept will then be applied.

72

CARL J. POSY

It is this process that Kant describes in the first paragraph of the second edition of the Critique of Pure Reason: There can be no doubt that all our knowledge begins with experience. For how should our faculty of knowledge be awakened into action did not objects affecting our senses partly of themselves produce representations, party arouse the activity of our understanding to compare these representations, and, by combining or separating them, work up the raw material of the sensible impressions into that knowledge of objects which is entitled experience? (Bl)

Structurally speaking, this is very much the sort of “associationism” that Mill envisions. Indeed, Kant’s notion of reflection seems to mirror Mill’s views about concept formation in two very important respects: 1) For Kant, just as for Mill, the basic procedure involved in concept formation — the procedure of unifying data under a covering notion — is a small scale version of the broader scientific activity of discovering more and more general laws that govern the regularity of nature.35 2) Kant — again, like Mill — views this process as “justifying” the subsequent uses of the empirical concepts thus formed. His point in the “Transcendental Deduction” is that empirical concepts can look the origin in individual experiences that he described above in order to establish their quid juris, their proof of “objective validity.” But in fact the analogy is superficial. A deeper look reveals that Kant systematically differs from Mill on each of the points. Regarding Kant’s notion of reflective concept formation, insofar as his description of the reflective process relates to “justification” it is not at all part of empirical psychology. Indeed, Kant’s whole machinery of “faculty psychology” is not an introspective or empirical discipline at all. Consider, for instance, the power of imagination, which underwrites the formation and applications of all our concepts. Clearly the same process of schematic imaginative projection is at work in our application of empirical concepts, in mathematics and in the universalization test imposed by the Faculty of Reason. Yet, in the latter case, Kant quite explicitly denies that this process can be called “imagination.” “The moral law,” he says, “has no other cognitive faculty to mediate its application to objects of nature than the understanding (not the imagination)” (Critique of Practical Reason, Akad. V, 69). His point in this passage is not that when we introspect in cases of ethical deliberation we will find some new, non-imaginative projective process. His point, rather is an epistemological one. When one abstracts from some aspect of an imagined situation (and thus views that situation schematically) one simply decides not to consider that aspect in drawing conclusions about that situation. Thus, for instance, the geometer refrains from considering any aspects of a perceptual manifold which are

KANT’S LOGIC AND PSYCHOLOGISM

73

independent of the general properties of space and time. Similarly, in ethical deliberation one refrains from considering causality (and with it any objective aspects of temporality) in drawing any conclusions from out imagined scenarios. Kant characterizes this as abstracting from time. And he is saying that when we refrain in that way, the process of imaginative projection can no longer technically be called “imagination.” Thus the distinction between this process and the mathematician’s projection is not introspective or even phenomenological. The difference is entirely epistemological; it is simply a question of which aspects of our imagined scenarios we can consider as evidence for our general conclusions. Indeed, the corresponding distinction between the faculties of Reason and Understanding is similarly entirely epistemological and not introspective. One might fault Kant — as Mill faults Hamilton — for using psychologistic language to describe these strictly epistemological considerations. In fact, however, it is better to say that at least from the time of the Inaugural Dissertation, he systematically understood these matters in an epistemological way. Thus, in the Dissertation itself, he takes pains to insist that the process of abstraction — the central vehicle for distinguishing one faculty from another — must be understood in a purely epistemic fashion: In so far as intellectual things strictly as such are concerned, where the use of the intellect is real, such concepts whether of objects or relations are given by the very nature of the intellect and they have not been abstracted from any use of the senses nor do they contain any form of sensitive cognition as such. But it is necessary here to notice the extreme ambiguity of the word abstracted, and I think this ambiguity must preferably be wiped clean away beforehand lest it mar our investigation into things intellectual. I mean that it would be proper to say to abstract from some things, but not: to abstract something. The first expression indicates that in a certain concept we should not attend to other things bound to it no matter how, while the second expression indicates that it would only be given concretely and in such a way that it is separated from the things joined to it. Hence an intellectual concept abstracts from everything sensitive, but is not abstracted from things which are sensitive, and perhaps it would more rightly be called abstracting rather than abstract. (1770, 6, Akad. II, 394)

Kant’s point is that we cannot introspect and observe some intellectual or abstract state of mind involved in grasping certain concepts. It is a matter of the epistemological uses rather than any special state of mind which determines that we are engaging the intellectual faculty. As for the apparent similarity between Kant and Mill regarding justification: When we understand Kant’s epistemological turn, we can also understand his distinction between those concepts on the one hand whose “application to objects of nature” is justified by tracing their origins in the process of reflection and those concepts, on the other hand, whose “origin” may be the same, but whose “justification” is purely epistemic. Ordinary empirical concepts — the concept tree, for instance — fall into

74

CARL J. POSY

the first group. The projections involved in applying these concepts must include an element of contingency. In particular, the concept tree includes the notion of a root system. And so, in applying that concept to organize a perceptual manifold I am justified in predicting that further exploration around the base will reveal a root system. But those prediction are part of the concept tree only by virtue of the original experiences (mine or humankind’s) from which the concept was formed. Therefore, in thus projecting I must consider the possibility that those experiences have misled. I must, as Kant insists, allow for revisions of my concept. But the justifications for concepts in the second group — a group that includes the concepts of mathematics and the categories — comes from transcendental arguments, arguments in the end which show how these concepts are presupposed by the very possibility of objective judgment. Or so Kant has claimed. There are no empirical objects, he has argued, without the our ability to locate them in space and time, an ability whose nature is explored by mathematics. And there would be no justified application of empirical concepts without the principles supplied by the categories. Moreover, quite in the same vein, there would be no thought without the forms of judgment that give rise to logic. In these cases — in applying the concepts of mathematics, the categories and the precepts of logic — the associated projections and practices will allow no room for revision. To be sure, there is even in these concepts, a certain element of metaphysical contingency: We can entertain the possibility of finite beings whose forms of intuition differ from our own. Such beings would find our mathematics invalid, and would indeed have quite a different categorial framework. There is even a brute-factness about the functions of finite judgment which lie at the base of logic. Kant himself raises this possibility at B145-146 in the Critique of Pure Reason: This peculiarity of our understanding, that it can produce a priori unity of apperception solely by means of the categories, and only by such and so many, is as little capable of further explanation as why we have just these and no other functions of judgment, or why space and time are the only forms of our possible intuition.

We do discover, even by trial and error, the a priori facts of mathematics, the structure of experience, and even the general laws of logic. But having uncovered these facts, structures and laws, we must also recognize their normative force for us. Once again, this is because for Kant, just as for Leibniz, there is no gap between the necessary forms of knowledge, the conditions of truth and the nature of objects. And this is what Kant means when he says: But though all our knowledge begins with experience, it does not follow that it all arises out of experience. For it may well be that even our empirical

KANT’S LOGIC AND PSYCHOLOGISM

75

knowledge is made up of what we receive through impressions and of what our own faculty of knowledge (sensible impressions serving merely as the occasion) supplies from itself. If our faculty of knowledge makes any such addition, it may be that we are not in a position to distinguish it from the raw material, until with long practice of attention we have become skilled in separating it. (B1-2)

That is where he differs most stringently from Mill on the foundations of logic. Kant may well admit that our forms of reasoning are specialized to the empirical objects of experience, and though he may well agree that our recognition of these forms derives from reflecting upon that experience. Nevertheless, he has provided a philosophical system which will still afford them with a normative validity. So we can say, in the end, that both of the Millian rifts that we have explored — the rift that he imposes, for his own rhetorical ends, on theoretical semantics and the one that exists within his own foundations of logic — both of these are defeated by the Leibniz-Kant combination. From Leibniz we learn the confluence of epistemology, semantics and ontology. And from Kant we learn how humanize these disciplines while maintaining that unity. Because of this confluence, reference and assertibility can coexist in a philosophical system. And, because of this unity, that system can balance both the contingency of the forms of thought with the fact that they are thoroughly normative for us.36 Department of Philosophy The Hebrew University of Jerusalem, ISRAEL

NOTES 1 The second edition was brought out just three months later in June of 1865. The third, much expanded, edition appeared in 1867; the French translation in 1869; and the fourth English edition in 1871. 2 That this was Mill’s intention is apparent from a letter he wrote to Bain while he was composing the work. Mill said that he expected the book would “make it very difficult to hold [Hamilton] up as an authority on philosophy” (22 December 1863). 3 “There is scarcely a point of importance,” says Mill in his letter to Bain, “on which [Hamilton] does not hold conflicting theories, or profess doctrines which suppose one theory while he himself holds another” (22 December 1863). 4 The most prominent of these challenges is Hilary Putnam’s in Reason, Truth and History (1981) and in his subsequent writings. Space does not permit me to compare Putnam’s “internal realism” with the philosophical views I explore in this essay. I hope to do so elsewhere. 5 Actually, Hamilton’s first wrong view, according to Mill, is to think that cognitive states like conception, judgment, and inference are foundations of logic in the first place. For Mill has taken a long step toward what we now call the “linguistic turn”. According to Mill, names (general and singular terms, in our modern parlance), propositions, and arguments are the proper components of logic and not mental processes. But, Mill avers, the fact that Hamilton’s logical theory is “pre-linguistic” does not by itself condemn that theory: “Many writers have given good and valuable expositions of the principles and rules of Logic, from

76

CARL J. POSY

the Conceptualist point of view. The doctrines which they have laid down respecting Conception, Judgment, and Reasoning, have been capable of being rendered into equivalent statements respecting Terms, Propositions, and Arguments.” (1872, 414). 6 See Mill’s A System of Logic (1841) Book I, Chapter VII, §4. In particular: “...a hundred generations have not exhausted the common properties of animals or of plants, of sulphur or of phosphorus; nor do we suppose them to be exhaustible, but proceed to new observations and experiments, in the full confidence of distinguishing new principles which were by no means implied in those we previously knew.” 7 “But if the mind is incapable of thinking anything respecting Noumena except the Phænomenal which it considers as proceeding from them, and to which it can appeal to test its thoughts; and if we are under no necessity of thinking these otherwise than in conformity to what they really are; we may refuse to believe that our generalizations for the Phænomenal attributes of Noumena can be applied to Noumena in any other aspect, without in the least invalidating the operation of thought in regard to anything to which thought is applicable” (1872,495). 8 ‘“If I,’ says our author [Appendix to Lectures, I, 402], ‘have done anything meritorious in philosophy, it is in the attempt to explain the phænomena of these contradictions in showing that they arise only when intelligence transcends the limits to which its legitimate exercise is restricted.’ ‘In generating its antinomies, Kant’s Reason transcended its limits, violated its laws...Reason is only self-contradictory when driven beyond its legitimate bounds...’” (1872, 494n.). 9 The latticework image is adopted from Kauppi’s Über die Leibnizsche Logik (1969). 10 A concept according to Leibniz is a “disposition to think of the object when the occasion arises.” It is thus a cognitive entity that takes mental occasions into thoughts (1686, XXVI). Leibniz is clearly “pre-linguistic” in the sense I discussed above. Indeed, self-consciously so. That is part of his critique of Locke. 11 “For if some man were able to carry out the complete demonstration by virtue of which he could prove this connection between the subject, who is Caesar, and the predicate, which is his successful undertaking, he would actually show that the future dictatorship of Caesar is based in his concept or nature and that there is a reason in that concept why he has resolved to cross the Rubicon rather than stop there, and why he has won rather than lost the day at Pharsalus, and why it was reasonable and consequently assured that this should happen” (1686, XIII). 12 I hasten to add that it is the unity and not the existence that is at issue here. The notions of “real” versus “phenomenal” are adjectives describing varieties of unity, not of existence. Leibniz is not a Berkeleyan phenomenalist, and aggregates are not mind-dependent or imaginary objects. See Mugnai’s Leibniz’s Theory of Relations (1992) for an interesting discussion of the nominalistic aspect of Leibniz’s metaphysics. 13 “The predicate or consequent therefore always in heres in the subject or an ent. And, as Aristotle, too, observed, the nature of truth in general or the connection between the terms of a proposition consists in this fact” (Leibniz 1680, 267-70). 14 A central issue in Leibniz’s philosophy is, of course, the problem of how to reconcile this analyticity with the apparent contingency of most empirical judgments, and, indeed, how to reconcile this conception of truth with the reality of free will. For a recent treatment of this issue, see Adams, Leibniz: Idealist, Determinist, Theist (1991). 15 “When my mind understands at once and distinctly all the primitive ingredients of a conception, then we have intuitive knowledge. This is extremely rare as most human knowledge is only confused or indeed assumed” (1686, XXIV). 16 Anglophone philosophers tend to emphasize the challenge of Hume’s skepticism as the entry point into Kant’s Critical Philosophy. Chief among these was Hamilton himself. (See, for instance, Hamilton 1878, 642: “To Hume we owe the philosophy of Kant, and therefore also, in general, the latter philosophy of Germany.”) But it is important to note that throughout his career Kant worked within a Leibnizian framework. 17 This is the point of his claim (at CPR A571-2/B599-600) that “every thing, as regards its possibility, is likewise subject to the principle of complete determination, according to which

KANT’S LOGIC AND PSYCHOLOGISM

77

if all the possible predicates of things taken together with their contradictory opposites, then one of each pair of contradictory opposites must belong to it.” The complete determination of “things” emphasized here is contrasted to the merely logical principle of contradiction which is described in Kant’s preceding paragraph and which governs concepts. Kant is saying here that objects, but not concepts, are completely determined. 18 If the faculty of understanding could generate an intuition, it would be a Leibnizian intuition, the grasp of an infinitely complex individual concept. Kant admits that in the abstract this is a coherent notion. He calls it “intellectual intuition.” But Kant also says that for us humans this is but an empty notion, not part of our human abilities. 19 To claim that no judgment can be singular simply in virtue of its grammatical structure is in modern terms to claim that there are no singular terms. To be sure, Kant does admit, in an abstract sense, the Leibnizian notion of a singular concept, a fully complete presentation of an object. He mentions the concept of the “perfect man” and the concept of God Himself as examples of such singular “intelligible” concepts (CPR A568/B596, A580/B608). But this presumed completeness is (as Leibniz said) beyond our human grasp, and these complete concepts do not enter into the determination of our human empirical truth. 20 This point is most clearly made in the A-Deduction, where Kant ties perception, conception, and imagination. It is here that he says in a famous footnote “that imagination is a necessary ingredient of perception itself (CPR A120n). 21 “Every concept, as partial concept, is contained in the representation of things; as ground of cognition, i.e., as a mark, [Merkmal], of those things contained under it” (Kant 1800, Akad. IX, 95). 22 See my “Imagination and Judgment in the Critical Philosophy” (1991) for a more thorough discussion of these points. 23 “What is truth? The nominal definition of truth, that it is the agreement of knowledge with its object, is assumed as granted...” (CPR A58/B82) See also A191/ B236. 24 Only by virtue of its connection to an intuition can we say that the nominal subject of a judgment exists. This is stated explicitly in the “Postulates of Empirical Thought”: “In the mere concept of a thing no mark of its existence is to be found. For though it may be so complete that nothing which is required for thinking the thing with all its inner determinations is lacking to it, yet existence has nothing to do with all this, but only with the question whether such a thing be so given us that the perception of it can, if need be, precede the concept (CPR A225/B272-3). This is, of course, also the main point of the “Ideal” chapter. 25 When we add the axioms for simple linear order. 26 “If we regard the two propositions, that the world is infinite in magnitude and that it is finite in magnitude, as contradictory opposites, we are assuming that the world, the complete series of appearances, is a thing in itself that remains even if I suspend the infinite or the finite regress in the series of its appearance. If, however, I reject this assumption, or rather this accompanying transcendental illusion, and deny that the world is a thing in itself, the contradictory opposition of the two assertions is concerted into a merely dialectical opposition” (CPR A504-5/B532-3). 27 Or, to be more precise, in the face of such an epistemic optimism, the only excuse for refusing to deny the Law of the Excluded Middle would be to endorse a niggardly here-andnowism that equates truth with knowledge right now. In that case, even optimism would not overcome the fact that some judgment is currently undecided. However, it is quite clear that Kant does not embrace so narrow an assertibilism: “We can also, however, know the existence of the thing prior to its perception and, consequently, comparatively speaking, in an a priori manner, if only it be bound up with certain perceptions, in accordance with the principles of their empirical connection (the analogies)” (CPR A225/B283). 28 This is quite different from the considerations of complexity and infinite information that often underlie our modern epistemic pessimism, 29 Indeed, in the immediately preceding passage, Kant tells us that mathematics — a science whose objects are not “outside” us — is “in a position to demand and expect none but assured answers to all the questions within its domain (quaestiones domesticae), although up to the present they have perhaps not been found” (CPR A480/B508). And in mathematics, Kant

78

CARL J. POSY

does allow definitions and classical logic. 30 Kant has already formulated his doctrine that sensible intuition is a singular representation. Since causality is a relational notion, he assigns it to the intellect. The distinction between intellectual and sensible faculties that Kant is making here becomes, in the Critique of Pure Reason, his distinction between understanding and reason. 31 My translation here differs slightly from that of Kerford and Walford. 32 “Since, then, empirical principles are not found in metaphysics, the concepts met with in metaphysics are not to be sought in the senses, ...To this genus belong, possibility, existence, necessity, substance, cause, etc., together with their opposites or correlates.” (Akad. II, 395) 33 Actually, Mill complains that once again Hamilton has given inconsistent presentations. Sometimes he recognizes this practical face of logic and sometimes he obscures it. 34 The narrow formal logic is, at its heart, the theory of syllogisms. And in Book II, Chapter iii, of his System of Logic (1841), Mill spells out his belief that deductive inference (in the form of the Syllogism) is simply a convenient shortcut in the business of inductive reasoning. 35 See, for instance, §VI of the “Introduction” to the Critique of Judgment (1790). 36 This paper is a revised version of a paper with the same title that appeared in Philosophy & Rhetoric, 30, 1997. I did the research for parts of the original version while I was a Winston Fellow at the Institute for Advanced Studies of the Hebrew University of Jerusalem. I am grateful to the Winston Foundation for its support and to the Institute for its wonderful hospitality. Research for the revision of the essay was supported by a grant from the Israel Academy of Sciences and Humanities.

REFERENCES Adams, R.M. Leibniz: Idealist, Determinist, Theist. New Haven: Yale University Press, 1991. Brouwer, L.E.J. “Über Definitionsbereiche von Funktionen”. In Mathematische Annalen 97: 60-75. Rpt. in vol. 1 of L.E.J. Brouwer, Collected Works, ed. A. Heyting, 390-405. North Holland, 1975. Dewey, John. Logic: The Theory of Inquiry. New York: Holt, 1938. Dummett, Michael. “The Philosophical Basis of Intuitionistic Logic”. In Truth and Other Enigmas. Cambridge: Harvard University Press, 1978. Kant, Immanuel. On the Form and Principles of the Sensible and Intelligible World. [1770]. Akad. II 385-419. Trans. from Selected Pre-Critical Writings and Correspondence with Beck. Trans. G. B. Kerferd and D. E. Walford. Manchester: Manchester University Press, 1968. Kant, Immanuel. Critique of Pure Reason. 1st ed. 1781. 2d ed. 1787. Trans. N. K. Smith. New York: St. Martin’s Press, 1961. Kant, Immanuel. Critique of Practical Reason. 1788. Akad. V 3-163. Trans. L. W. Beck. Indianapolis: Bobbs-Merrill, 1956. Kant, Immanuel. Critique of Judgment. 1790. Akad. V 167-485. Trans. J. C. Meridith. Oxford: Ox ford University Press, 1952. Kant, Immanuel. Lectures on Logic. Ed. G. B. Jäsche. 1800. Akad. IX 1150. Kant, Immanuel. Gesammelte Schriften. [Akad.]. Vols. 1-22. Herausgegeben von der Preussischen. Akadamie der Wissenschaften zu Berlin, 1902 -. Kauppi, R. “Über die Leibnizsche Logik”. Acta Philosophica Fennica, Fasc. 12. Helsinki, 1969.

KANT’S LOGIC AND PSYCHOLOGISM

79

Leibniz, G. W. “First Truths.” 1680. Trans. in Loemker 1956, 267-71. Leibniz, G. W. “Meditations on Knowledge, Truth and Ideas.” Acta Eruditorum (Nov. 1684). Trans. in Loemker 1956, 291-95. Leibniz, G. W. Discourse on Metaphysics. 1686. Trans. in Loemker 1956, 303-30. Leibniz, G. W. New Essays on Human Understanding. 1765. Trans. A. G. Langley. LaSalle, Ill.: Open Court, 1949. Loemker, L. E., ed. Philosophical Papers and Letters. By G. W. Leibniz. 2d ed. Dordrecht: Reidel, 1956. Mill, John Stuart. A System of Logic. London: Longmans, Green, 1841. [New Impression, 1930, based on eighth ed.] Mill, John Stuart. Letter to Bain. 22 November 1863. In Collected Works, The Later Letters, ed. F. E. Mineka and D. N. Lindley, vol. 15, pp. 9012. Toronto: University of Toronto Press, 1972. Mill, John Stuart. An Examination of Sir William Hamilton’s Philosophy and of the Principal Philosophical Questions Discussed in his Writings. 4th ed. London: Longman’s, Green, Reader, and Dyer, 1872. Mill, John Stuart. Autobiography. Ed. J. Stillinger. Boston: Houghton Mifflin, 1969. Mugnai, M. Leibniz’s Theory of Relations. Stuttgart: F. Steiner Verlag, 1992. Posy, Carl J. “Imagination and Judgment in the Critical Philosophy”. In Kant’s Aesthetics, ed. R. Merebote, 27-48. North American Kant Society Studies in Philosophy 1. Atascadero: Ridgewood, 1991. Putnam, Hilary. Reason, Truth and History. Cambridge: Cambridge University Press, 1981.

This page intentionally left blank

WERNER STELZNER

PSYCHOLOGISM AND NON-CLASSICAL APPROACHES IN TRADITIONAL LOGIC

INTRODUCTION

In the transition period from traditional logic to modern logic, which saw its culmination in the establishment of the Fregean paradigm for an objective foundation for logic, psychologism in logic was treated by most of the protagonists of the new logic as the main enemy for a well founded formal logic. The word “Psychologism passed for an ideological label for something not worthy of serious consideration from the standpoint of modern logic. This merely philosophical neglect of psychologistic logic is in accordance with the general neglect of the positive logical content of traditional logic.1 In fact, traditional logic vanished from the map of logical science at the turn of the last century almost without leaving any substantial traces in the modern logical theories behind. Reasons for this traceless disappearance can be seen in the largely non-symbolic form of the presentation in traditional logic and in the fact that the traditional logic, contrary to the developing modern logic, was oriented closer to philosophy, linguistics and common sense argumentation than to mathematics. Another important gap between traditional logic and modern logic in the Frege-paradigm was caused by the already mentioned widespread psychologistic treatment of logic by many traditional logicians. So, as a fact, the genetic role of traditional logic in the establishment of the new modern logic is very poor. Consequently, if the treatment of traditional 19th century logic is reduced to its genetic role in the development of modern logic, it seems to be justified to leave out traditional logic if one intends to write a history of the progress of logical thinking. No wonder, in treatises about the history of modern logic the ideas of scholars of the traditional logic of the second half the 19th century are hardly mentioned. This picture changed not so much since the days of 81 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 81-111. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

82

WERNER STELZNER

Kneale/Kneale [1962], where names of prominent traditional logicians of the 19th century, like Herbart, Lotze, Sigwart or Wundt, do not occur at all. And Bochen,’ ski [1956, 192] explicitly refuses to include traditional logic in the history of formal logic, labeling traditional logic as a decadent period in the history of logic. In this prevalent attitude, traditional logic comes out as one of the dead ends in the history of logic, not even being a detour to modern logic. If the logical importance of traditional logic is considered at all, then this importance is reduced to a negative one: Traditional logic finds its place in the history of logic, because it is this kind of formal logic, the overcoming of which is one of the main merits of the developing modern logic. However, this is a quite incomplete picture about traditional logic. This picture especially does not reflect the important analytic work of traditional (and especially psychologistically-minded) logicians in the field of logical analysis of language, the results of which were formulated, however, mostly in an informal manner. Even in view of the obovious technical shortcomings of psychological logicians and the opposition from the side of antipsychologistic attitudes, during the time of the transition from traditional to modern logic the paradigm of psychologism did not only allow, but even promoted the development of interesting pragmatically influenced non-classical logical ideas. So, if we look a little closer at the works of psychologistically influenced traditional logicians, then we find a manifold of interesting ideas with special relevance to modern non-classical logics. As examples one can point at entailment theory, the analysis of negation and the so-called logical basic laws of contradiction and excluded middle, the logical treatment of presuppositions, causality, the theory of judgment and propositional acts and attitudes and last but not least at the logic of modalities. All this is just one point in the picture, I wish to give in this paper: Psychologism should not be condemned and extinguished from the map of the development of logic, because several of the many kinds the philosophy of logic and of the attitude, how to do logic, which can be named uniformly as psychologism, are not only compatible with sensible ideas and developments in logic, but in some sense could even stronger promote the development of such ideas then a purist objectivistic attitude would do. Just in order to give a slight impression about this, in the second part of this paper, I will focus on non-classical approaches, developed by Christoph Sigwart (1830-1904), the Russian logician Nikolai Aleksandrovich Vasil’ev (1880-1940) and by the founder of experimental psychology, Wilhelm Wundt (1832-1920). Sigwart and Wundt were not only highly prominent inside the psychologistic movement, but found their place in the modern history of logic as main targets of Husserl’s tart antipsychologistic critique.2 Before coming to this, it seems to be appropriate to sketch out some fundamental attitudes towards the relation between logic and psychology among dedicated anti-psychologistic and psychologistic German logicians of the 19th century.

PSYCHOLOGISM IN TRADITIONAL LOGIC

83

1. The Diversity of Psychologism in the Traditional Logic of 19th Century 1.1 The “Psychologism” of Anti-Psychologism

When we look at the main differences between psychologism and antipsychologism in logic,3 the main points defended by objectivistic antipsychologism are the objectivity of truth and the independence of truth from any kinds of epistemic attitudes. In accordance with this, logic is directed to the laws concerning such objective truths, and the objectivity of the soundness of logical laws, the objective validity of logical rules is defended against psychologistic treatments. All this is connected with a clear distinction between soundness (Geltung) and genesis (Genese) of the truths, the grasping of truth (Erfassen) and the acknowledgement (Anerkenneri) of a truth. Psychologism and antipsychologism in logic are divided not so much by the use of psychologistic or epistemic terms in their logic or in the philosophy of logic. Even objectivistic logicians like Herbart, Bolzano, Lotze and Frege use terms, common in psychology. Such terms are, e.g., imagination [Vorstellung], thought [Gedanke], judgment [Urteil], thinking [Denken], etc. If treated inside logic by these logicians, such terms is given an objective, non-psychologistic content. And if treated in the sense of epistemic attitudes or acts, like in the speech-act-interpretation of judgment [Urteil] in Frege, logic is not developed between judgments, but between the (objective) content of judgments. So, to accuse these logicians to be influenced by psychologism, misses exactly the main point of the difference between psychologism and antipsychologism. This main point concerns their attitudes towards soundness, validity, genesis, grasping and acknowledgment of logical laws. And here they defend the objectivity of soundness [Geltung] against psychological treatments or they argue against the (hidden or explicit) replacement of this notion by epistemic notions like grasping of the truth or common acknowledgment of truth. As the father of 19th century objectivism and anti-psychologism one can easily identify Johann Friedrich Herbart (1776-1841). This seems to be justified because of his work as such and on the other side because he is the main target of anti-objectivistic attacks from the side of psychologistically inclined logicians of the 19th century.4 But nevertheless, if one examines the terminology of Herbart, as Rolf George did in his interesting paper, one can find many places of an “unwanted psychologism”, as George pointed out.5 However, already in Herbart, like in Bolzano6 or in Frege7, such a use of psychological terms is mostly reduced to the treatment in their Kantian form as pure [reine] notions. What they stand for is the objective content of thinking, believing, judging etc. In this context the use of such

84

WERNER STELZNER

psychologistic terms is just a way to explain more clearly the objectivistic attitude, like in the following statement of Herbart: It is true that logic treats imaginations. But it treats not the actus of imagining: also neither the way and the manner, in which we come to imaginations, neither it treats the mental state to which we are moved by this imagination. (Herbart 1808, 467)

Following Herbart, logic is concerned “only with this, what is imagined, i.e., with the object of imagination.” (Herbart 1808, 467) Another point, where psychologism can creep into logic is the applied logic. This is one reason, why objectivistically inclined logicians trend to draw a clear division between pure and applied logic. For Herbart, this seems to be an unavoidable precondition for the holding up of the objectivity for pure logic, while applied logic is open for psychological intrusion.8 Consequently, Herbart demands a strong distinction from applied logic, which together with all psychologistic elements has to be thrown out of logic. Following his anti-psychologistic confession, “according to which the logical has to be separated from any intrusion of the psychological” (Herbart 1813/1993, 96), Herbart insists on the “unavoidable sorting out of all psychological and consequently of the applied logic, which is tangled up in psychological matters” (Herbart 1813/1993, 81). The third place where psychologism could threaten the objectivist antipsychologistic logic is a possible normative claim of logic which in some sense of course leads from pure logic to applied logic. Normatively treated, the logic is not the science about how people think, but (a) how they should think according to the logical legislation,9 (a kind of unconditional normative psychologism) or (b) how people should think in order to fulfil their aims in a rational way (a kind of conditional normative psychologism). Herbart combines the disapproval of empirical psychologism with the inclination to an unconditional normative psychologism, according to which “the whole logic is a moral for the thinking, not a natural history of the intellect.” [die ganze Logik eine Moral für das Denken ist, nicht aber eine Naturgeschichte des Verstandes] (Herbart 1829/1851, 127).10 Frege also accepts a normative function of logic concerning the thinking. But Frege does not follow Herbart’s unconditional normative psychologism, but holds a form of conditional normative psychologism. So, for Frege it is not a question of some (inner?) moral of thinking, but just a question of relative rationality to be oriented on the soundness of logical laws, when planning your acts or organizing your thinking under certain aims or purposes. In this kind of conditional normative psychologism the laws of logic have no other psychological impact than other objective sound laws, like mathematical laws or laws of natural science, linguistics or even “genuine” psychological laws: When you wish to fulfill your aims it is better to be oriented at objectively sound laws. To fulfill this psychological

PSYCHOLOGISM IN TRADITIONAL LOGIC

85

function of logical laws, no foundational impact of psychology for logic is presupposed, and it is not presupposed that “these laws would reign over the thinking in the same way and manner, like the laws of nature govern the process in the outer world“ (Frege 1893, XV). So, it is not presupposed that logic is a description of the thought process. In fact, if ones wishes to scale psychologism, conditional normative psychologism seems to be one of the weakest, less pretentious, forms of psychologism. Stronger forms of psychologism11 are divided from objectivism by its readiness to give the other kinds of epistemic-pragmatic relations to sentences, judgments, soundness, truth a place inside logic. Just in order to sketch out some key directions in the diversity of such genuine forms of psychologism, one can look at main representatives of psychologism from four scientific generations of traditional logicians of the 19th century, namely Jakob Friedrich Fries (1773-1843), Friedrich Eduard Beneke (1798-1854), Christoph Sigwart (1830-1904), Wilhelm Wundt (1832-1920), Benno Erdmann (1851-1921) and Theodor Lipps (1851-1914). 1.2 Different Forms of Strong Psychologism and the Foundation of Logic 1.2.1 Fries: Anthropological Logic

The first important anti-objectivistic program in 19th century traditional logic is carried out by the often underestimated founder of logical anthropologism Jakob Friedrich Fries. Fries followed Kant in the neglect of the ontologization of logic. However, he did not follow the transcendentalism of Kant in the foundation of logic, but proposed an anthropological-psychological foundation of logic. Anthropologism in Fries is a program for philosophy and logic, which should be carried out in order to communicate philosophy and logic in an understandable manner. (This is already uttered in the first edition of his System der Logic from 1811 and repeated in the following editions, see Fries 1837 (3), IV) Somehow remaining in the Kantian schema of dividing logic into formal and transcendental logic, Fries discerns between demonstrative (formal, or philosophical) logic and anthropological logic. However, there is no absolute independence between formal and anthropological logic. According to Fries, this division is needed in order to enable anthropological logic to develop a firm foundation for formal logic: By the division a justificational circle is avoided: If anthropological logic should deliver a foundation of formal logic, then it should be independent from formal (demonstrative) logic. In the third edition of his System of Logic Fries can already answer the reproaches of Twesten, Reinhold and Herbert against his attack on objectivism in former editions. Fries insists that his anthropological logic

86

WERNER STELZNER

surmounts the main defect of Herbart’s logic, namely that Herbart’s confinement to pure logic leads to an extremely poor result: “With the omission of all psychological, he can teach nothing more about the concepts, then to say that they are combinations of attributes (Merkmale).” (Fries 1837 (3), 3) The basis for the development of Fries’s anthropological logic is empirically given inner experience. According to this, anthropological logic has the same empirical basis as psychology of those days. Nevertheless, Fries does not reduce the foundation of philosophical formal logic) to empirical generalization: “However, it would be absurd the basic principles of philosophical (demonstrative) logic, the necessary basic laws of the thinkability of things to prove by empirical psychology, i.e., by experience.” (Fries, 1837 (3), 5) In his own understanding, Fries is not an empirical foundational psychologist. Similar to the later Christoph Sigwart or Wilhelm Wundt,12 he considers psychological experience to be unavoidable for the grasping of logical laws. But these laws are not sound because of this grasping. A merely direct reduction of logic to empirical psychology not only on the side of the grasping of such laws we find immediately after Fries in Beneke’s attempt to give a foundation of logic as an empirical (psychological) science. 1.2.2 Beneke: The genetic-lively (genetisch-lebendige) method in logic

Friedrich Eduard Beneke ususally counts as the first traditional logician with a clearly empiricistic psychological conception of logic. He developes this attitude concerning logic in his 1832 Lehrbuch der Logik oder Kunstlehre des Denkens [Textbook of Logic and Art of Thinking]. By using his geneticlively method [genetisch-lebendige Methode] he wishes to overcome the “abstract attitude” which leads to an unfruitful logic. This method he criticizes especially in connection with Herbert’s anti-psychologism by which any psychological element is thrown out from logic. (See Beneke 1842, IV) Contrary to Herbart Beneke thinks that the union between psychology and logic is the unavoidable condition for the development for a contentful logic. Like for other psychologistic logicians of 19th century the connection between psychology and logic is the way to give logic a scientific basis, i.e. a basis which is connected with the progress of empirical sciences. So this is the way in which logic should not be further divided from empirical sciences but should have a gain from the use of methods which proved their fruitfulness in empirical science. Only together with psychology one can “develop logic truly as the Kunstlehre, the art of thinking” (Beneke 1842, V). For Beneke the connection between logic and psychology is a condition for the defeat of subjectivism in logic. Only psychological considerations can give a truly scientific foundation for logical rules. Such a foundation

PSYCHOLOGISM IN TRADITIONAL LOGIC

87

does not depend on abstract constructions of rules which are based on authority. (See Beneke 1842, 16). With the help of scientific psychology Beneke wishes to built logic with his genetic-lively method, in order to present, “all forms and relations of the thinking not in abstract form and as completed given in the accomplished soul, but in genetic-lively (genetisch-lebendig) form” (Beneke 1842, VI) And this means, contrary to the Herbartian objectivistic attitude, that logic should include the analysis of things in real connection with our imagination. These things should not be things, which exist independently from our imagination, but “things, as they occur in our imagination and thinking, how they are included and processed by the thinking” (Beneke 1842, 7). Notwithstanding this empirical aspect, following Beneke, logic should be the science of the real and of the ideal thinking. “Ideal” here refers to the normative, prescriptive function in contrast to the descriptive task. So, logic is not reduced to descriptive empirical psychology. Logic has to fulfill a normative task, it is the “code of thinking”, the “lawbook of thinking” [Gesetzbuch des Denkens]. But again, this lawbook of thinking cannot be developed on the abstract formal basis. The ideals of thinking have to be natural truths, which are founded in the reality of thinking, not in abstract division from thinking. (See Beneke 1842, 9). 1.2.3 Erdmann: Implicit foundational psychologism

A merely implicit foundational psychologism is advocated by Benno Erdmann, who was much criticized for his psychologism by Frege.13 Erdmann did not explicitly develop a kind of foundational psychologism. So, he does not explicitly derive the soundness of logical laws from the psychologically determined empirical processes, like the grasping of such laws by epistemic subjects or the genesis of the acknowledgement of such laws. Following his own words in his Logik from 189214 he seems to be just a defender of a normative psychologism: The judgments, conclusions, definitions, partitions etc. are processes of the consciousness, which stand in lawful connection to each other and to other processes of the imagination and of the sensation and the will. From this, however, does not follow, that the object of logic would be an object of psychology [...] The logic does investigate these processes not related to the actual conditions of their origin, their development and connections, but remains with the question, how they should be in order to become universally valid propositions about the imagined. (Erdmann 1892, 18).

This seems to be in accordance with the moderate conditional normative psychologism of the kind Frege was at least admitting. Maybe, this seemingly closeness with his own views was one reason for Frege to reveal the hidden foundational psychologistic character of Erdmann’s basic attitude concerning the status of logical laws.

88

WERNER STELZNER

The crucial point in Erdmann’s statement is the relation of logical laws to universal soundness [Allgemeingültigkeit]. Today the notion of universal soundness in logic seems to be a clearly formal or objectivist notion for sentences or logical formulae, being true in every possible interpretation. Not so during the times we are speaking about. In the traditional psychologistic treatment, this universal soundness, which is used by Erdmann as definition and criteria for the soundness of logical laws, is taken as the expression for common agreement or common agreeability about the soundness of a judgment or logical principle. For the objectivistic logician, like Frege, the thread for logic from foundational psychologism consists in the abundance of the firmness and eternity of logical principles, because logic now seems to be dependent on the psychological organization of the epistemic subjects. Erdmann does not treat this psychological organization in a transcendental sense, but as a fact of the empirical world. With this world, the psychological organization of its inhabitants can change. Thought necessity [Denknotwendigkeit] as basis for universal soundness of logical principles is then is relativized to the given empirical situation, in which the thinking takes place. And Erdmann is well aware about this, when he writes as a kind of upshot of his logic: Also in view of this, our logical principles remain their thought necessity (Denknotwendigkeit), only that this necessity is seen not as absolute, but as hypothetical. We cannot act in another way than to agree to them — according to the nature of our imagining and thinking. They hold generally, presupposed, that our thinking remains the same. (Erdmann 1892, 378)

Here Erdmann has reached the soil of foundational psychologism in a clearly empirical manner. In some sense he is outlining a conclusion from the enterprise of anthropological logic, if this is taken in an empirical sense: There can be different logics, there is not only one (classical) logic, there can be deviant (or non-classical) logics too, even if not at the same time. It is just this program, which led the Russian logician-psychologist Nicolai A. Valsil’ev to his attempt to found a new, non-Aristotelian logic without the principle of contradiction, a paraconsistent logic, which proves itself to be one of the many possible empirical logics.15 1.2.4 Lipps: Explicit foundational psychologism: logic as physics of thinking

When other advocates of psychologism16 underline their moderateness by stating that logic is not a physics but an ethics of thinking, then Theodor Lipps in his Logik from 1893 pronounces his unlimited foundational logical psychologism without any hesitation. And this foundational psychologism in Lipps is even connected with an explicit denial of the normative function of logic, because for such a normative function there is no place in Lipps conception of logic, which rests on a partial identity of the psychology of

PSYCHOLOGISM IN TRADITIONAL LOGIC

89

thinking and the logic of thinking. To characterize logic as the science of normative laws of thinking, in Lipps attitude is a plain and misleading tautology, “because we are thinking correct in every case if we are thinking”. (Lipps 1893, 1). Lipps has no problem, to pronounce logic as the “physics of thinking” instead of an “ethics of thinking”. And because of this: ...logic is a psychological discipline, as certain as the cognition occurs only in the psyche, and the thinking, which completes itself in the cognition, is a psychical event. (Lipps 1893, 1 f.).

Nevertheless, logic has a special place inside psychology, the object of logic is the investigation of the relation between knowledge and error. However, this radical psychologistic attitude is without any consequence for the formal logic, presented by Lipps in his Logik: This logic is a very conservative kind of traditional logic, without any danger to fall out of classical principles usually contained in traditional logic. I especially mention this, in order not to create the wrong impression, as if logical psychologism would automatically be connected with non-classical attitudes concerning the positive logic, developed on the psychological soil. However, psychologism in several respects is more open to such nonclassical attempts. In the following this will be demonstrated with regard to two of the most important psychologistic logicians of the late 19th century: Christoph Sigwart and Wilhelm Wundt. 2. Sigwart’s Treatment of Modalities

Sigwart’s occupation with logic was mainly directed at the application of logic in colloquial and scientific contexts, especially to the methodology of science. In these fields he saw the main tasks for logic and an unavoidable precondition in order to overcome the abstractness and meagerness of traditional formal logic. His intention is to develop logic “under the point of view of the theory of methods, and to set logic thereby in alive relationship to the scientific tasks of the present” (Sigwart 1873, I), as he pointed out already on the first page of his Logik from 1873. And Sigwart is following his idea17 not in a strong division between pure and applied logic, and because of this — as a confirmation of Herbart’s remark about the connections between psychologism and applied logic — psychological moments play a role on many occasions in Sigwart’s logic. In Sigwart we find a conditional normative psychologism connected with a psychologism, which is directed to promote the grasping of logical laws. However, the psychological involvement of Sigwart is much deeper than the conditional normative psychologism (e.g., in Frege). This is connected with his opinion that truth should be connected only with completed or proposed judgments

90

WERNER STELZNER

and the truth is not a possible feature of the content (which is given by a sentence) of a judgment in separation from any judgment. So, in Frege’s terms, in Sigwart one finds the basic fault of psychologism, the confusion between truth and the holding something for a truth [Fürwahrhalten]. However, in Sigwart this is merely a question of terminology because for Frege’s use of truth Sigwart uses the term soundness [Gültigkeit] and nothing can be true in Sigwart’s sense without being sound. Other consequences of Sigwart’s psychologism can be seen in his opinion that there are no logical laws outside epistemic contexts. The logic is undivisably connected with epistemic attitudes or acts. So, for Sigwart a logical reason, which is not known, is impossible, is a contradiction in itself.18 However, the soundness of such a reason even in Sigwart depends in no way on the fact if there are confirming or denying epistemic attitudes concerning this reason. So there is no subjectivism connected with Sigwart’s psychologism. And this may be the main reason why Frege who knew Sigwart’s Logik19 very well did not chose Sigwart as the target of his anti-psychologistic critique, but the Logik of Erdmann. In fact, if one looks a little under the surface of Sigwart’s words, then his psychologism shows itself as a quite weak form of psychologism, which at least does not lead to subjectivism. Maybe, Frege at this point had the deeper insight than Husserl later had, who took Sigwart as one of his main targets of his anti-psychologism. In Sigwart’s logic the orientation at the theory of methods and at applied logic is connected with the openness concerning the analysis of different logical forms in colloquial language and in the language of science, which are often associated with psychological involvement. In this orientation at applied logic in a wide sense, Sigwart differs essentially from the Fregean approach to the new logic. Contrary to Sigwart, for Frege logic becomes the object of his interest under the point of view of the development of a firm foundation for arithmetic, based on the development of pure logic in form of suitable formalized languages. Even if one finds in connection with arithmetic the use of the new Fregean logic as applied logic, it is not the application of logic to fields outside of logic, because in the logicist perspective the arithmetic is of course not outside logic, but a genuine part of it. So at least the Sigwartian interest in applied logic is much broader then Frege’s initial interest in logic. In Frege’s new logic the modal logic plays only a role, when its logical importance is denied. Already at the beginning of his Begriffsschrift from 1979 Frege explicitly excludes the logical analysis of modalities from this logic, because modalities in his opinion are logically not significant. The traditional modal distinctions between apodictic and assertoric judgments according to Frege are speaking only about the reasons for a judgment. In view of this, for Frege, modalities have no influence on the conceptual content of such a judgment. However, Frege’s logic is concerned exclusively with the conceptual content of judgments, and so, there is no

PSYCHOLOGISM IN TRADITIONAL LOGIC

91

place for the traditional apodictic-assertoric distinction in Frege’s logic. (See Frege l879, p. 4 f.) If nevertheless, necessity or possibility is syntactically treated inside Frege’s logic, then in many cases general propositions are identified with necessary propositions and particular propositions are identified with propositions about possibility:20 If a sentence is characterized as possible, then either the speaker abstains from a judgment, indicating that he does not know any laws, from which would follow the denial; or he says that the denial of the sentence in its generality would be false. According to the usual expression, in the last case we have a particularly confirming judgment. “It is possible that the earth will collide with another heavenly body” is an example for the first, and “A cold can result in the death” is an example for the second case. (Frege 1979, 5).

Disregarding the logical treatment of modalities, Frege has not only allies from the side of objectivistically-minded logicians, but is even in accordance with Benno Erdmann, who was criticized so sharply by Frege for his implicit foundational logical psychologism. In the object of Frege’s antipsychological critique, in the Logik of Erdmann, one can read the following statement: Special remarks about modal conclusions don’t have principle logical importance, how slippery, according to the so far mentioned, the soil is, on which they move [...] The important logical Problems, which are established for the logic by the modal judgments, don’t belong to the domain of these conclusions, but belong to the border questions between logic and the theory of knowledge. [Erdmann 1892, p. 638 f.]

Erdmann seems not to be entirely sure about the convincing power of his warning, and so, for people, nevertheless interested in the logic of modalities, Erdmann gives the following advice: “Because of this we mention here just the thorough discussion, which was given by Sigwart concerning the (understood by him in a different way) modal qualifications.” (Erdmann 1892, 638). On the next pages this advice will be followed. The treatment of modalities in Sigwart’s logic is quite multilayered and differentiated. This treatment is connected with basic attitudes of Sigwart concerning the notion of judgment, especially with the (illocutionary) truthclaims, epistemic certainty-claims and intersubjective necessity-claims, which in Sigwart’s treatment are connected with judgments. According to Sigwart, if one comes to a judgment, then this judgment is already the expression of a necessity. Because of this, Sigwart does not accept different kinds of judgments, there are not three different kinds of judgments in the Kantian sense: apodictic, assertoric and problematic judgments, but just one kind of judgment, which is connected with the mentioned illocutionary

92

WERNER STELZNER

claims. Modalities come into the play of logic not as characterizing different kinds of judgments, but as belonging to different kinds of the content of judgments. In Sigwart’s treatment the Kantian distinction of judgments according to their modality does not refer to three distinct kinds of judgments. First, the so-called problematic judgment (A can be B in the sense of A is possibly B) is no judgment at all:21 ...because it lacks the awareness of objective soundness, i.e., it is no judgment about what is referred to by the subject of the sentence. It is a judgment only insofar, as it says that the speaker is undecided concerning the question, if A would be B. (Sigwart 1911, 239 f.)

Sigwart cannot acknowledge the so-called problematic judgment as a kind of judgment, because of the truth-claim and the certainty-claim, which are connected with every judgment. Because it belongs to the essence of the judgment that it establishes an assertion, which claims to be true and to be believed: so a proposition, which does not assert and lets it free, if the opposite is true, cannot be a kind of judgment. (Sigwart 1911, 242 f.)

While the so-called problematic judgment is no judgment at all, for (because of the necessity-claim) the apodictic judgment holds that it clashes with the assertoric judgment: The so-called assertoric judgment (the simple assertion that A is B) is not essentially different from the apodictic (it is necessary to assert, that A is B), insofar as together with every fully consciously pronounced judgment the necessity to pronounce it is asserted too. (Sigwart 1911, 240)

If there should be a difference between apodictic forms of asserting a judgment and assertoric forms, then Sigwart, anticipating the words of Frege writes: “It only remains that partly the reason, on which the necessity rests is different, partly that they come in different ways to the consciousness.” (Sigwart 1873, 194; 1911, 247) However, for both Sigwart and Frege, there are no different “modalized” kinds of truth. Following this, there cannot be connected with the judgment different kinds of truth-claims. However, with the exclusion of different kinds of judgments according to their modality, Sigwart — contrary to Frege — does not deny the logical significance of modal distinctions. According to Sigwart one has to discern the possibility or necessity to judge (whereupon the Kantian distinction of the judgments according to their modality in problematic, assertoric and apodictic is directed to) and the possibility or necessity, which is expressed in the content of the judgment (corresponding to the theory of modalities): The assertion, that a judgment [an assertion] be possible or necessary, is different from the assertion, that it is possible or necessary, that a subject has a

PSYCHOLOGISM IN TRADITIONAL LOGIC

93

predicate. The first affects the subjective possibility or necessity of the judging; the second affects the objective possibility or necessity of that, what is expressed (said) in the judgment. [Sigwart 1911, 239]

Sigwart accepts the fact that modalities play an important role in the sciences and in colloquial contexts. So logic, as directed to the language and methods of science has to find the logical significant place of modalities. And this place Sigwart identifies not in different kinds of judgments, but inside the judgment, in the content of the judgment. Here modalities like possibility or necessity fulfill the function of special predicates, which like other predicates are directed to a subject. Modalities are predicated to propositions, which appear as a special sort of subjects for modal predicates. According to this, not things, properties or relations in separation are possible or necessary, but sentences, in which is referred to things, properties or relations can have modal characterizations. A judgment like “God is necessary” is then treated as an abbreviation, where the predicate necessary is predicated to a sentence, expressing the existence of god. So, in similar cases, modalities are saying something about the existence of a thing, his having a feature, or an activity of this thing (See Sigwart 1911, 267). In the same sense have even judgments about modal features of judgments to be treated: “The necessity of the judgment constitutes no exception; it is necessary that I and every thinker judges this.” (Sigwart 1911, 267) With this a new class of propositions arises, in which the subject is this, what is said in a judgment (not the judgment itself, like in the predicates true, false, believable, logically necessary). (Sigwart 1911, p. 267)

This treatment of the modalities as predicates inside the judgment gives Sigwart the possibility to develop logically significant distinctions between different kinds of modalities occurring in colloquial or scientific language. And in this context for Sigwart, contrary to Frege, the conceptual content of sentences with modalities does not clash with the conceptual content of sentences without modalities. 2.1 Kinds of Modalities As a fundamental distinction between modalities Sigwart mentions the distinction between real and logical modalities. Real necessity is founded on causal relations, while logical necessity does not depend on the causality in the given world, but on the terminological connections in a given language. Logical necessity rests alone on the meaning of the words occurring in a sentence. Contrary to this, for real necessity the reference of the words in the world has to be considered; real modalities say something

94

WERNER STELZNER

about the world, while logical modalities are directed to a given terminology. Because of this, sentences about real modalities can be verified or falsified, sentences with logical modalities neither can be empirically verified nor falsified. Special forms of real modalities are the internal and external modalities. This distinction Sigwart somehow introduces as a substitute for the traditional distinction between analytic and synthetic judgments: If the subject alone is sufficient for the necessity of its qualification by the predicate, one has a traditional analytic judgment and Sigwart then speaks about an internal necessity, and if other facts have to be added in order to receive the necessity of the predication, like in the case of a synthetic judgment, the necessity is named an external necessity.22 The internal necessity is an essential (absolute) necessity, which has no other condition for its existence then the existence of the subject of the considered predication.23 In contrast, external necessity depends on the existence of special relations between the subject and other subjects and as a kind of relative necessity it is mediated by the given circumstances: Each individual is so, because an other is so; each change of an individual thing happens, because a certain change of an other thing has occurred; the things have the power to prescribe their behavior mutually to each other; the coherence of the world exists in this necessity, which is passing over from one thing to the other thing. (Sigwart 1911, p. 270)

This external necessity is a kind of relative necessity, mediated by relations, which can be expressed by strict (necessary) conditionals or hypothetical judgments in the strong Sigwartian sense (see Sigwart 1871). The cognition of this external necessity is described by Sigwart according to the deductive-nomological scheme of explanation which has not only an important place in Sigwart’s theory of methods, but later was fundamental also for Hempel’s theory of explanation:24 The knowledge, that something is the way it is due to external necessity, that it necessarily happens, the way it happens, is composed always from two elements: the general law, and the certain data, to which this law is applicable. (Sigwart 1911, p. 270 f.)

With the external necessity Sigwart introduces a notion of conditional necessity which is not reducible to absolute necessity: The necessity of every individual is in every case a conditional necessity. [...] If something is declared to be necessary, not its cause is declared to be necessary, but its emergence from the given cause is declared to be necessary. (Sigwart 1911, 271)

PSYCHOLOGISM IN TRADITIONAL LOGIC

95

Relative necessities are an important basis for the foundation of hyphothetical judgments which do not assert the necessity of a predication but which do assert a necessary connection between different predications.25 2.2 Teleological Necessity: Voluntative Logic As a special form of external or relative necessity Sigwart treats the teleological necessity. According to Sigwart, the teleological necessity, like “Man must eat”, is reducible to the hypothetical necessity, which is explained by Sigwart with the above-mentioned deductive-nomological scheme of explanation. Teleological necessity in this sense is not an independent kind of necessity. In the foundation of the teleological necessity the purpose is taken like a real cause, which can constitute a real necessity, this purpose is the thought of a real, thinking and wanting creature. The purpose: ...is thought and wanted as something in the future, the realization of which should come. In accordance with the causal order of nature, which connects every certain success with certain causes, the purpose can be realized only by the agency of certain causes; therefore, who wants the purpose, must also want the means. The presupposed wanting of a certain purpose makes the wanting of certain means necessary. (Sigwart 1911, p. 271 f.)

However, it is not so clear in every case what the relevant means for a purpose are. Contrary to the definitness of Sigwart’s words, there are different possible means, which are sufficient for attaining the purpose. Even if the best purpose can by carried out by the worst means, from this follows no demand to carry out these bad means. Because of this, the purpose does not sanctify every kind of methods, sufficient for attaining the purpose. However, if one (unconditioned) wants the purpose, he/she should wish to have some sufficient mean in order to attain the purpose, but there is no obligation to want one set of means actually known to be sufficient for this purpose, because after wanting a purpose there should be just a wanting for finding appropriate means for attaining the purpose. Not in every case this search will be successful. The following example shows that Sigwart connects the teleological necessity, as special form of relative necessity, with necessary conditions for attaining the purpose: Man must breathe, in order to live — expresses, however, finally nothing else then the insight, that the order of nature has connected to the standstill of the breathing irrevocably the death, and that the breathing cannot be replaced by any available arrangement. So, if life would be wanted as purpose, this would require to want the breathing as the means, too. (Sigwart 1911, 272)

96

WERNER STELZNER

So, what is clear about the wanting of the means for a given purpose is that if one wants the purpose, one should want the unavoidable means to this purpose. This shows that the application of the scheme of explanation to the reduction of the teleological necessity to relative necessity is different from the usual explanation of relative necessities: Usually hypothetical necessities are introduced with reference to sufficient conditions (causes) for their realization. Teleological necessities, however, are explained as necessary conditions for the realization of the wanted result. Following this, the realization of the aspired result (formulated in the purpose) is the sufficient condition (and gives the reason) for the means. Therefore the means appear (possibly against the temporal order) as a consequence of the wanted result. While Sigwart recognizes the purpose as the cause of the teleological necessity, the temporal order between cause (which is in the case of teleological necessity a reason) and the necessity of the effect based on this cause is reestablished: First we have the purpose and from this the necessitiy of the needed means for attaining the purpose. As a special form of teleological necessity, the moral necessity of the obligation is treated analogously under the point of view of a creature, which is following moral norms or wishes to attain a moral purpose. Replacing the (positive) purpose by a negative purpose, i.e., the avoidance of a sanction, in a similar way 80 years after Sigwart has explicated the notion of obligation according to the Anderson-Scheme:26 H is obligatory (OH), if and only if from the non-performance of H follows a sanction (S).27 In terms of teleological necessity: H is teleological necessary, if and only if from the non-performance of H follows a sanction. While Sigwart would have used for the explication of the entailment between non-performance of H and the entering of the sanction his hyphothetical judgment,28 Anderson uses a necessary implication for his explication. Therefore, from the non-performance of H follows a sanction S if and only if ~H implies necessarily that a sanction is imposed, i.e.: or shorter, ~H strictly implies S: where is the symbol of strict implication. In Sigwart’s original explication the entering of the sanction is replaced by the lapse of the realization of a certain aim or purpose (P). Consequently, in the sense of Sigwart with the means used in modern logic can be defined:

PSYCHOLOGISM IN TRADITIONAL LOGIC

97

The given definitions are, however, are connected with one uneasiness: They are formally (and as regards the content) correct only in case that S and P are constants. To overcome this confinement, the notions of Ideological necessity should be introduced as notions, which are relativized to the relevant sanction or the purpose. This relatedness could be expressed by an explicit parameter, where S and P are now treated as variables: and with reference to a special purpose:

This parameter-relatedness of the teleological necessity and obligation can be used to introduce different modified notions of absolute teleological necessity and obligation. One possibility was realized already by the initial explication according to the Anderson-scheme: It consists in the assumption of fixed constant sanctions (and aims) for all teleological necessities in a considered situation. Without the reference to constant purpose or sanction, it is possible to define H as teleological necessary or obligatory, if there is at least one purpose P or aim, for the fulfillment of which the realization of H is unavoidable:

Using the above given definition for

this results in:

In this case the teleological necessity is related to a set of possible aims and purposes, and not every purpose out of this set should be compatible with all the other possible purposes. The possibility for conflicting aims and purposes should be considered as well. And consequently, the situations can arise, where different teleological necessities contradict each other. In view of the possibility of conflicting teleological necessities, it does not hold generally:

And from this we get that it does not hold:

98

WERNER STELZNER

So, in the sense of O’ it is possible that the performance and the nonperformance of the same act is Ideological necessary. Logically such situations are excluded only then, if the set of all possible purposes referred to by the quantification does not contain conflicting purposes. With this, we stop the treatment of Sigwart’s theory of necessity, even if there would be much more to say about this (e.g., about the relation between generality and necessity or between necessity and hypothetical judgment). 2.3 Possibility and Necessity Concerning teleological possibility in contrast to teleological necessity, there seem to be interesting questions to raise, which could lead to different explications for this kind of possibility. If one would like to discern alethic possibility of the form “It is possible that man breathes” from genuine teleological possibilities, this should direct to teleological can-possibilities. These can-possibilities seem to be the right counterparts to teleological must-necessities: “Man must breath” as a teleological necessity then has “Man can breath” (not meant just as a physical ability) as its counterpart in the form of a teleological possibility. Nevertheless, all this does not say anything about a definite treatment of this teleological can-possibility. Just to mention two meaning-variations: Should the teleological “can do H” be not more then the negation (or the absence) of the “must do not-H”? This would in a relevant sense be somehow awkward because then teleological irrelevancies would result in teleological possibilities, merely in the sense of not forbidden. From a teleologic can one could expect a more positive character, like in answers to the question “Can I attain the purpose by doing H?”. The answer “You can do H” then says much more in the teleological sense then in the sense of a permission. In the teleological sense possibility in this sense would mean that the realization of H is a sufficient condition for attaining the purpose, while the teleological necessity would express a necessary condition for attaining the purpose. Sigwart did not treat questions about teleological possibility. This is a little surprising, because on the other side he treated possibility to a large extent under the viewpoint of can-modalities, however mostly with a physical interpretation. Generally speaking, Sigwart treated, depending from the context of application, different kinds of possibilities which all are in different ways related to necessities. Even if Sigwart for some kinds of possibilities admits the interdefinability between necessity and possibility, in many cases possibilities are treated quite independent from necessity. Along with forms of possibility under the conditions of freedom and indetermination, Sigwart discusses questions of possibility under the condition of lack of freedom and of complete determinism, which lead in one direction to the discussion of possibility from ignorance.

PSYCHOLOGISM IN TRADITIONAL LOGIC

99

Under the conditions of freedom and indetermination, for an agent his own decision is deciding, which act will be performed by this agent. Here the agent is in a situation of free choice, which is not determined by any necessity.29 This again is a form of can-possibility. According to Sigwart, this free choice is contrary to physical and pragmatic limitations: It is opposed to the obligation not to do something and it is opposed to the (physical) impotence to perform the chosen act. At first glance, possibility under the conditions of lack of freedom and complete determination seems to be excluded (if not derived as weakened necessity). However, with his multidimensional treatment of modalities Sigwart comes up with several variants of sensible speech about possibilities under such hostile conditions for strong possibility. Sigwart points to the fact that one often speaks about possibility when the changing developmental stages of a thing, an animal etc. or just different appearances of the same entity are under consideration: Also the “not free” thing is active in different ways, in so far as it is changeable and the necessity does not prescribe for the not free thing for all times to be the same and to do the same [...] Our thinking, which anticipates the future, fluctuates between different predicates. Certainly, which of these predicates will at a certain point of time enter really, does not depend on the autarkic decision of the thing. It is determined by the necessity, [...] and it is only the difference of the time, which forces, to mark the future not as a being, but merely as a possibility. (Sigwart 1911, p. 277 f.)

An important class of sentences expressing such changing features of subjects are sentences with dispositional predicates. Sigwart considers examples of such sentences as “water can freeze and evaporate”, “iron can melt”, “table salt is soluble in water” etc. These are sentences about an ability or capability, directed to features of the general subjects “water”, “iron”, “table salt”. The dispositions express that such subjects can have different features under different conditions. Nevertheless, in dispositions there is a persistent internal reason of the changing subject which causes the different features in different situations. [See Sigwart 1911, 280]. If taken as sentences about a relation between subject and the changing conditions, disposition sentences, nevertheless have the character of necessary general sentences expressing a causal connection between subject and its changing features, depending on the given situation. Other cases of the use of possibility under strong determination rest on the truth of a particular judgment like “Some man are bold”, and “A man can be bold” or “It is possible that a man is bold”, say not more than “It is not excluded that a man is bold”, even if it could be that for every single man his being bold or not being bold is strongly determined. By the way, this use of possible is the one which Frege mentions too.

100

WERNER STELZNER

The second form of sentences about possibilities, mentioned by Frege, is one which Sigwart treats as possibility from ignorance which again is compatible with complete determination. Such speech about possibility can occur in the case, that no law is known which would determine a future development, or they can be connected again if a predicate does not belong to all individuals belonging to a general subject. E.g., if I do not know enough about a certain horse, I can only say that it is possibly black, that it is possibly brown or that it is possibly white. A general feature of the possibility from ignorance can be expressed in the following way: The more ignorance one has, the more will be possible for him; the possibility from ignorance grows with the extent of the ignorance. If we take complete determination as the expression for the fact that all what occurs in a world w, comes about necessarily in w, then all forms of possibility treated so far could be introduced because the treated kinds of possibility were not excluding the kind of necessity used in the introduction of complete determination. Treated by Sigwart kinds of possibility could apply even in the case that change and development would be impossible. The given sketch of the treatment of modalities in Sigwart is only one illustration for the development of non-classical elements in the Sigwart’s logic. A manifold of such elements is contained in his treatment of different kinds of negations, in his theory of hypothetical judgments and entailment.

3. Wundt’s Logic of Concepts In the logical work of Sigwart’s contemporary Wilhelm Wundt a manifold of different tendencies in the development of logic in the transition from traditional to modern logic at the turn from 19th to 20th century is present. In this work by Wundt in a unique manner such treatments of logic are joined together, which are not only in this time usually opposed. Like with Sigwart, dominating as starting point and as aim of Wundt’s logical efforts is his interest for applications of logic to sciences, which underlines the specific shapes of applied logic in different branches of science. Connected with a strong content-logical attitude, Wundt pursues this traditional turn to the logical analysis of non-logical branches of knowledge and of the logical analysis of scientific terminology, methods of proof and procedures of argumentation. Based on this, Wundt reveals distinctive nonclassical elements in the treatment of the determination of concepts in the form of a specific logical multiplication and of the summation of concepts in form of a logical addition. Developing and underlining these non-classical elements, Wundt follows his maxim that logic does not treat the knowledge and science from outside, like an independent stranger, but logic itself depends on the specifics of the sciences, to which logic is to be applied:

PSYCHOLOGISM IN TRADITIONAL LOGIC

101

If the logic is willing, to be submitted to those conditions, to which the science is submitted everywhere, then logic cannot proceed under the condition that the forms of thinking would be indifferent concerning the content of knowledge. [Wundt 1919, 8]

Starting from scientific practice, Wundt aims at an adequate analysis of the logical relations inhering in the sciences. With this endeavor, Wundt uses the technical means of logical algebra in order to sharpen the theoretical aspects perceived of in content logical relations. As means of presentation and systematization, Wundt eventually develops a specific nonclassical identity calculus on this basis. These efforts are carried out by a logician who is to be considered as a typical exponent of normative psychologism. Wundt is not following a program of substituting logic by empirical psychology and he does not try to derive the laws of logic from the laws of empirical psychology. Hinting at this, Wundt repulses the reproach of psychologism, raised by Husserl [1900] in the first volume of Logical Investigations. The difference between logic and psychology Wundt explains with the normative aspect of logic: While psychology is teaching us, how the course of our thought really proceeds, logic aims to determine, how one should proceed in order to arrive at scientific knowledge. [Wundt 1919, 1],

However, logic fulfills Its normative task as an empirical science: This work [Wundt’s Logik] aims to be something entirely different from a psychology of thinking [...]. It aims to be an empirical science sui generis what psychology — according to its real content and according to its suppositions — never can be. [Wundt 1919, p. VIII]

Despite of his differentiation between logic and psychology, for Wundt logic is not independent from psychology. He, nevertheless, believes in the “inseparable boundness of the logical laws to the psychological forms of the development of thinking.” [Wundt 1919, p. 90] 3.1 Content-Logical Foundation When Wundt is approaching the logical content of the general logic, which constitutes a common core of the different applied logics, he acts as a traditional concept-logician, who is tempting to explain the different forms of judgment and conclusion on the basic of the relations between concepts: Wundt developed the content-logical foundation of his system as an explication of the logical content of relations between notions. These content-logical ideas Wundt transferred into the extensional algebraiclogical formalism of Boole/Schröder. This way, Wundt tried to give his content-logical ideas distinctive symbolic expression, even if he considered

102

WERNER STELZNER

this formalism to have only a small autonomous general logical value. In order to adapt the system of Schröder to his content-logical ideas, Wundt constituted a variation of this system, so that in the revised system commutativity of logical multiplication did not hold. However, even in these variations the resulting system cannot claim to be an adequate formalization of Wundt’s logical ideas, which differ from the classical logic to a considerably higher extent than the given by Wundt formalization does. In the following, basic features of Wundt’s system of logic will be discussed in connection with his content-logical ideas. The suggested modifications of Wundt’s own trials to formalize his ideas aim to explicate in the spirit of Wundt the logical content of Wundt’s logical ideas more adequate than Wundt’s own formalizations did. The purpose pursued hereby is not primarily to deliver a critique of Wundt’s ideas or a complete description of these ideas, but to contribute to the grasping of the logical relevant and explicable non-classical content of Wundt’s System of Logic. 3.2 Content and Extension Following the tradition, Wundt discerns the content and extension of concepts. The content is defined by the sum of its features, while the extension is defined by the sum of those concepts, which are subsumed under the given concept. But Wundt feels the limitations connected with this treatment of concepts. Especially his interest in the sciences and their methodology convinced Wundt that the logical content of a concept is only partially determined by subsumtion of one concept under another. Wundt refers to relations of dependence, which cannot be expressed just by speaking about the subsumtion of concepts. Wundt sees another weak point of the traditional treatment in the unsatisfactory attention for relational concepts. However, Wundt does not criticize the traditional treatment of concepts in a general way. He just stresses the incompleteness of this treatment and tries to supplement this treatment. 3.3 Categories The addiction to the traditional treatment of concepts is already characterized, for instance, by the importance of different logical categories of notions in Wundt’s Logik. By the help of grammatical categories he introduces four different logical categories: 1. the category of objects is introduced by a hint at nouns, 2. the category of properties is given by a hint at numerals and adjectives, 3. verbs lead to the category of the states of affairs,

PSYCHOLOGISM IN TRADITIONAL LOGIC

103

4. the category of relations is given by adverbs of time and tense, prepositions, case endings, tempora and modi of verbs. The attention to categorical differences has special importance for Wundt’s treatment of the operations with concepts and for the comparisons between concepts, which are expressed in judgments: Exclusively concepts of the same category are comparable. In order to allow the comparison of formerly incomparable concepts, it is possible to perform categorical shifts, especially into the category of objects. So one can shift from “good” to “goodness” or to “good being thing”. Only in the category of objects “green” (in the sense of green thing) can be subsumed under the concept color, and only as a feature colored is more general as green. 3.4 Determination The determination of concepts compared to abstraction has the converse function: From a more general concept (determinand) by determination with the determinator a more specific new concept is built. So in “white sheep” “sheep” is the determinand and “white” is the determinator. In symbolic form a determination is expressed by b.A, where b is the determinator and A is the determinand. Wundt mentions the analogy between determination and logical multiplication, but stresses the difference between determination and multiplication: Never in the field of logic the determinator can be converted into the determinand or vice versa, without getting an entirely different composed concept. If I would, e.g., in the concept “white sheeps” (bA) convert the attribute into the subject and vice versa, there would arise the entirely different concept “the white of the sheeps”. (Wundt, Logik (4) I, 238)

Following Wundt, this difference between determination and multiplication is expressed most clearly by the fact that algebraic multiplication is a commutative operation, while logical determination is not a commutative operation. And here Wundt sees the originality and the advantage of his treatment of determination compared with the treatment by Leibniz until Boole and Schröder, (see Wundt, Logik (4) I, 239) Given a specific row of concepts (e.g., of color concepts), one can determine a concept from this row by concepts out of other concept rows, while the result of the determination nevertheless belongs to the row of concepts of colors. By determination with concepts out of the row of levels of brightness, we get colors again, determined by their level of brightness. Even if Wundt did not explicitly explain this, so an essential reason for Wundt’s deviation from the traditional treatment of the determination of concepts is the following: The logical features of concepts depend not only

104

WERNER STELZNER

on the content and the extension of the concepts, but on the row of concepts from which the given concept is a member. Precisely this relatedness to a given row of concepts or a given whole (totality) of concepts to which the concept belongs could be changed, if determinator and determinand would change their logical role. Accordingly, the full reduction of the logical determination to classical multiplication of concepts does not agree with an essential aspect of Wundt’s treatment of determination, because the classical treatment reduces all relations between concepts to relations between concepts of the category of concepts for objects: Furthermore, the view that any concept refers to a class of objects has contributed to the quoted meaning [that determination should be a commutative relation]. If x is the class of the white objects and y is the class of such objects, which are sheeps, then, however, the composed by x and y products xy and yx are commutative: their reference is the class of such objects, which are both white objects and sheeps. [...] A symbolism which relies on this view, however, does not represent the real relations of thinking. [Wundt, Logik (4) I, 239]

Relatively to the determination of the determinand Wundt discerns two logically significant cases: 1. Determinators can be connected directly with the determined concept, like in “good grateful people”. Such determinators are coordinated, i.e., they belong to the same main concept and for such determinators commutativity holds: “good grateful people” is equivalent with “grateful good people”. 2. There are cases, in which from two determinators only one is directly connected with the determinand. And this determinator again is determined by the other determinator. The second determinator then is a determinator of second order, or a mediated determinator. E.g., in “a nicely furnished house” the second order determinator “nicely” is included, which determines the determinator “furnished”. This determination (and of higher order too) cannot be reduced to the intersection of nice houses and furnished houses and is in fact unanalyzable in terms of categories (or classes) of objects. In a similar way, related to different categories and sets of concepts, Wundt introduces not only concept summation, which is non-associative, but different kinds of non-classical negations.

3.5 Negation 3.5.1 Contrary Negation

For Wundt’s treatment of concept-negation again the relatedness of concepts to a given row of concepts or the whole of concepts to which the concept belongs is of crucial importance. Consequently, every negation contains as

PSYCHOLOGISM IN TRADITIONAL LOGIC

105

logical parts, “1) the notion, which has to be negated and 2) the whole of concepts, to which this concept belongs”. If one designates the negated concept by P, the row of concepts to which the negated concept belongs by X, so the area inside which the negated concept is moved can be expressed as [...] X-P, using the symbolism introduced for algebraic operations. [Wundt (4) I, 252]

The negation of the concept P which belongs to the row of concepts X is a concept belonging to X, but not being P or, to put it in other words, the negated concept is a concept belonging to X-P, i.e., to the complement P relatively to the row of concepts X, which comprises P. Consequently, is sound, while is not sound. The concept-negation P is not the complement of P relatively to the P comprising set of concepts X, but it is an undetermined indicated part of this whole of concepts. Consequently the negation-judgment “S is non-P” stands for “S is a X, which is not P”. This treatment of concept-negation would be unambiguous (definite) only if there was just one whole of concepts X for a given concept. However, even under this condition the negation of a concept-negation leads back to the initial unnegated concept, only if the supposed whole of concepts contains just two concepts. Suppose, A, B and C belong to the same whole (totality) of concepts. Then the negation of concept A is one of the concepts B or C, but not the sum of B and C. Because of this, Wundt treats concept-negation as (undetermined) disjunction. Accordingly, the negation of the negation of A is either the negation of B or the negation of C. In the first case the negation of the negation of A is A or C and in the second case it is A or B. This way, the negation of the negation of A is an entirely undetermined one of the concepts A, B, or C. So the law of double negation is not sound in this treatment.30 Using the this can be expressed as:

Let be G = {A,B,C,}, then

Consequently:

and for double-negation

106

WERNER STELZNER

Contrary to this, Wundt in his calculus of identity nevertheless postulates the soundness of the law of double negation, indicating that the formalism forced him to change the original treatment of concept negation. The other way of Wundt, in order to get the law of double negation for his calculus of identity — against his original contrary treatment of conceptnegation — consists in following the symbolism of algebraic logic and in designating the totality of concepts connected with a concept P by the symbol “1” and expressing the negation of P by 1-P. But here the negation of P is treated as the complement of P relatively to the totality of connected concepts 1. The double negation then is 1-(1-P), and this (at least algebraically) equals P. 3.5.2 Relativized Totalities

Nevertheless, there remains a serious problem: Wundt uses the symbol “1” for all concepts for the symbolization of the totality of concepts, to which a concept belongs. Accordingly, “1” in connection with the concept A and “1” in connection with the concept B can express different totalities of concepts. Even if, following Wundt, for any concepts A and B we have both (a) and (b) in the case that A and B belong to different totalities of concepts, it is impossible to infer from this that holds. This indicates that 1 is not an independent symbol, but its meaning depends on the context in which it occurs. So this symbol is contextually determined by the concepts occurring together with 1 in an identity. If one wishes to disambiguate the symbol 1, one could suggest to use 1 as symbol for the general totality of concepts, but then all of Wundt’s talk about the connection of concepts with the totality of concepts belonging to them would be superfluous. A way, more adequate in Wundt’s intuition, would be to introduce indices in order to disambiguate the occurring in formulae (a) and (b) symbol 1: and is the symbol for a relativized totality, namely for the totality of which A is a part. If then we have and More general disambiguation of the symbol 1 we receive, if we indicate with the index not just the concepts which belong to the totality, allowing for a concept to belong to different totalities, but let designate just different totalities of concepts. Then we can also express different negations of a concept relatively to different totalities to which the concept belongs. E.g. it could be that and what gives different relativized negations for A in each case. Then, consequently, negations could be introduced, which are explicitly relativized to certain totalities of concept.

PSYCHOLOGISM IN TRADITIONAL LOGIC

107

3.5.3 Relativized Empty Classes

In analogy with the specific totalities for an adequate treatment of concept negation, the empty class too should be introduced relatively to different concepts. This could be expressed again by a specific index of the zero class sign “0”. E.g., for the empty class belonging to the concept totality of A, relatively to B, etc. Purely extensional in relation to the objects belonging to a class, the empty classes and do not differ: Neither nor contain an object. However, and are nevertheless not identical with each other. The identity presupposes that they are relativized to the same totality of concepts, what is especially important for (contrary and contradictory) negations of or Only for complementary negations logically holds and However, are neither for contrary nor for complementary negation sound, because there is no logical law stating that A and B belong to the same totality of concepts. The introduction of empty classes, which contain different non-real elements, is used by Wundt’ s contemporary Hugh MacColl in order to avoid the so-called paradoxes of existential import,31 which are based on the fact that arbitrary empty classes are included into any class (or: empty concepts have any concept as generic term). In MacColl’s treatment, accordingly, does not hold that all round squares are triangles, but no round square is a triangle. However, no round square is a square neither. If we, following Wundt, discern empty classes according to the totalities of concepts to which they belong, then the class of round squares remains included in the class of squares, but it is not included in the class of triangles. Round squares are not triangles, but at least squares. MacColl could have reached an analogous result if he would allow for different empty classes with nonreal elements, which are related to definite non-empty classes.

4. Conclusion Not only with their non-classical treatment of modality, conceptdetermination and of different negations Sigwart and Wundt deliver further examples that psychologism in the transition period between traditional and modern logic was able to promote the development of interesting nonclassical logical ideas. On the psychologistic soil, e.g., the important nonAristotelian ideas of the Russian logician Nicolai Vasil’ev were developed. On the other hand, Wundt with his non-classical calculus of identity has demonstrated that a psychologistic attitude can harmonize with mathematical-algebraic presentations of logic, a symbioses, which was alive already in the works of George Boole [1847, 1854]. If looking at the history of traditional logic and psychologism in logic, it is clear that psychologism

108

WERNER STELZNER

establishes no guarantee for the development of interesting non-classical logical approaches. But because of the connection with applications of logic to the analysis of language, terminology and argumentation patterns and to special epistemic and pragmatic components in science and colloquial contexts, psychologism can constitute a fruitful soil for the development of non-classical ideas. In fact, it was not psychologism, what prevented psychologists like Sigwart, Wundt, or Vasil’ev to go the way to the new logic, but their inability to leave the traditional paradigms of formal logic, their inability to come to precise formal explications of their non-classical logical ideas.32 Fachbereich 9/SG Philosophie Universität Bremen, GERMANY NOTES 1

By traditional logic in this paper I mean the German traditional logic, which developed in the 19th century under the influence of the Kantian formal logic. 2 Husserl (1900). 3 In its most influential forms in 19th century logic, anti-psychologism is mostly connected with objectivism, while in epistemology transcendentalism is or phenomenalism are the most influential foundations for anti-psychologism. However, in any case it would be misleading to strongly connect psychologism with subjectivism. 4 For the justification of this assertion, one could point e.g. to many places in Sigwart’s logic. However these critiques are much less aggressive than the later anti-psychologistic critique, in early 20th century, when the quarrel between anti-psychologism and psychologism was influenced not only by philosophical or logical positions, but became a kind of ideological war around fundamental (philosophical) beliefs and was connected even with social components. See Rath 1994, Kusch 1995, Peckhaus 1997. 5 George (1997, 229-231). 6 Bolzano (1837). 7 Frege (1879). 8 This holds even in Husserl’s (1910) understanding, where applied logic is not free from psychological involvement. 9 This seems to be in good accordance with the transcendental interpretation of logic and logical laws. 10 Later we will see that in the strongest form of empirical psychologism by Lipps both attitudes of Herbart are turned to the opposite. 11 If in the following “psychologism” is used without further qualification, then a kind of strong psychologism is referred to. 12 The basic philosophical attitudes of Sigwart and Wundt will be treated more extensively in connection with their non-classical logical ideas in the second part of the paper. 13 Frege (1893), Einleitung. 14 And in the following editions. 15 See Vasil’ev (1910, 1912), Bazhanov (1988, 1995), Stelzner (1998, 2001). 16 Like (already before Erdmann) among others, Sigwart or Wundt. 17 He does this against Herbart’s demand for withholding from treating applied logic together with pure logic, which was mentioned above. 18 Sigwart (1911).

PSYCHOLOGISM IN TRADITIONAL LOGIC

109

19 Against his usual habits, Frege even wrote an extract of around 30 pages about Sigwart’s Logik. 20 Haaparanta (1988). 21 Sigwart’s exclusion of the possibility of problematic judgments later is taken up by Vasil’ev, who uses it in his attempt to establish a non-Aristotelian logic: With reference to Sigwart, Vasil’ev excludes from the area of logic the weak particular judgments of the form “Some or all S are B” (which Vasil’ev understands — in contrast to Sigwart — as a kind of traditional problematic judgment). See Vasil’ev (1910), Raspa/Vergauwen (1997). 22 Sigwart (1911), 268 f. 23 A paradigmatic kind of inner necessity is mathematical necessity. A special feature of such a necessity is the constancy of the causes for such a necessity. Because of this, with the soundness of a general sentence in mathematics a mathematical law is connected. This seems to be the reason why even Frege was sometimes inclined to ignore the difference between the accidental generality and the necessary generality (a difference, which is only given outside mathematics). 24 See Hempel (1956). 25 Sigwart (1871). 26 Anderson (1956, 1958). 27 Having in mind Sigwart’s remarks about the necessity of things, it is of course not the thing sanction, which follows, but the entering of a sanction follows. So S is just an abbreviation for the proposition that a sanction S will enter (or will be imposed). 28 Sigwart (1871). The hypothetical judgment in Sigwart’s treatment is a necessary connection between antecedent and consequent with elements of relevance between them. See Stelzner(2001). 29 We have here a close similarity to the analysis of “Could have done otherwise” in the context of Nuel Belnap’s “Seeing to it That” (see Belnap/Perloff (1988), Perioff 1991). 30 However it is sound for special cases, e.g., if negation is treated as strong contrary negation, which picks out as equivalent to the negation of the concept a concept, which is in an ordered raw in the same distance from the center like the negated concept. 31 See MacColl (1906), 42 f. and S. 76 f. 32 The author is grateful to the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG — Bonn) and the University of Bremen for their invaluable support, which made it possible to realize this research.

REFERENCES Anderson, Alan Ross. “The Formal Analysis of Normative Systems”, New Haven (1956). Rpt. in: Rescher, Nicholas (ed.), The Logic of Decision and Action. Pitsburgh: University of Pittsburgh Press, 1967, 147-213. Anderson, Alan Ross. “A Reduction of Deontic Logic to Alethic Modal Logic”, Mind, 67, 1958, 100-3. Bazhanov, Valentine A. Nikolaj Aleksandrovich Vasil’ev: 1880-1940. Moscow, 1988. Bazhanov, Valentine A. The Interrupted Flight. The History of “University” Philosophy and Logic in Russia. Moscow, 1995. Belnap, Nuel. D./Perloff, Michael. “Seeing to it That: A Cannonical Form for Agentives”, Theoria, 54, 1988, 175-99. Beneke, Friedrich Eduard. Lehrbuch der Logik als Kunstlehre des Denkens. Berlin/Posen/Bromberg, 1832.

110

WERNER STELZNER

Beneke, Friedrich Eduard. System der Logik als Kunstlehre des Denkens. Erster Theil. Berlin: Ferdinand Dümmler, 1842. Bochen,’ ski, Innocent Maria. Formale Logik. Feiburg i. B. und München: Alber, 1956. Bolzano 1837. Wissenschaftslehre. Boole, George. The Mathematical Analysis of Logic. Being an Essay Towards a Calculus of Deductive Reasoning. Cambridge/London, 1847. Boole, George. An Investigation of the Laws of Thought, on which are Founded the Mathematical Theories of Logic and Probabilities. London, 1854. Erdmann, Benno. Logik, 1. Band: Logische Elementarlehre. Max Niemeyer, Halle, 1892. 2., erw. Aufl., Berlin 1907. Frege, Gottlob. Begriffsschrift, eine der mathematischen nachgebildete Formelsprache des reinen Denkens. Halle a. S.: L. Nebert, 1879. Frege, Gottlob. Grundlagen der Arithmetik. Breslau: W. Koebner, 1884. Frege, Gottlob. Die Grundgesetze der Arithmetik. Vol. I, Jena: H. Pohle, 1893, Vol. II, Jena: H. Pohle, 1903. Fries, Jakob Friedrich, System der Logik (1837). 3. Aufl., Heidelberg: Christian Friedrich Winter, 1811 (1). George, R. “Psychologism in Logic: Bacon to Bolzano”, Philosophy and Rhetoric, 30, 1997, 213-42. Haaparanta, Leila. “Frege and His German Contemporaries on Alethic Modalities”, in: Knuuttila, S. (ed.), Modern Modalities. Dordrecht/Boston/London: Kluwer, 1988, 239-74. Hempel, Carl Gustav. Aspects of Scientific Explanation and Other Essays in the Philosophy of Science. New York/London: The Free Press, 1956. Herbart, Johann Friedrich. Lehrbuch der Psychologic, in: Johann Friedrich Herbarts Sämmtliche Werke, hrsg. von G. Hartenstein, 5. Band., Leipzig: Leopold Voss( 1829) 1851, 1-190. Herbart, Johann Friedrich. Lehrbuch zur Einleitung in die Philosophie. A. W. Unzer, Königsberg, 1813; reprinted: accord. to the 4th ed. Hamburg: Felix Meiner, 1993. Herbart, Johann Friedrich. Sämmtliche Werke, 12. Bände; Herausgegeben von G. Hartenstein. Leipzig: Leopold Voss, 1850-1852. Husserl, Edmund. Logische Untersuchungen, Bd. 1, C. E. Pfeffer (Robert Stricker), Halle a. S., 1900. Kneale, William, and Kneale, Martha. The Development of Logic. Oxford: Oxford University Press, 1962. Kusch, M. Psychologism. London/New York: Routledge, 1995. Lipps, Theodor. Grundzüge der Logik. Hamburg: Voss, 1893. Lotze, Rudolf Herrmann, 1843, Logik. Weidmann’sche Buchhandlung, Leipzig, 1843. Lotze, Rudolph Hermann, 1874, Logik. Drei Bücher vom Denken, vom Untersuchen und vom Erkennen, S. Hirzel, Leipzig (= Lotze, System der

PSYCHOLOGISM IN TRADITIONAL LOGIC

111

Philosophie, T. 1). New edition: Lotze, Rudolph Hermann, 1989, Logik. Erstes Buch. Vom Denken (Reine Logik). Philosophische Bibliothek Band 421, Meiner, Hamburg. Lotze, Rudolph Hermann, 1989, Logik. Drittes Buch. Vom Erkennen (Methodologie). Philosophische Bibliothek Band 408, Meiner, Hamburg. MacColl, Hugh. Symbolic Logic and Its Applications. London/New York, Bombay: Longmans, 1906. Peckhaus, Volker. Logik, Mathesis universalis und allgemeine Wissenschaft, Berlin: Akademie-Verlag, 1997. Raspa, Venanzio, and Vergauwen, Roger. “Possible Worlds with Impossible Objects: The Imaginary Logic of N.A. Vasil’ev”, Logique et Analyse, 159, 1997, 225-48. Rath, M. Der Psychologismusstreit in der deutschen Philosophie. Freiburg/ München: Alber, 1994. Schröder, Ernst. Abri? der Algebra der Logik. Bearb. von E. Müller. Teil I: Elementarlehre, Leipzig/Berlin 1909, Teil II: Aussagentheorie, Funktionen, Gleichungen und Ungleichungen, Leipzig, 1910. Schröder, Ernst. Der Operationskreis der Logik. Leipzig, 1877. Sigwart, Christoph. Beiträge zur Lehre vom hypothetischen Urteil. Tübingen, 1871. Sigwart, Christoph. Logik, Vol. I: Die Lehre vom Urteil, vom Begriff und vom Schluss, 1873/1878; Vol. II: Die Methodenlehre, Heinrich Laupp, Tübingen, 2. Aufl. Freiburg 1889-1893, 4., durchgesehene Auflage, J.C.B. Mohr (Paul Siebeck), Tübingen, 1911. Stelzner, Werner. “Vasil’ev, Psychologism, and another Foundation for Tautological Entailments”, in: P. Weingartner/G. Schurz/G. Dorn (eds.), The Role of Pragmatics in Contemporary Philosophy, Kirchberg am Wechsel 1997, 920-5. Stelzner, Werner. “Zur Behandlung von Widerspruch und Relevanz in der russischen traditionellen Logik und bei C. Sigwart”, in: W. Stelzner/M. Stöckler (eds.), Zwischen traditioneller und moderner Logik. Nichtklassische Ansätze. Mentis Verlag: Paderborn 2001, 239-96. Vasil’ev, Nicolai A. “On Particular Judgments, the Triangle of Opposites, and the Law of Excluded Fourth”, Uchennye zap. Kazan. un-ta 77, kn. 10, 1910, pp. 1-47. Vasil’ev, Nicolai A. 1912 “Imaginäre (nichtaristotelische) Logik” (in Russian). In: Zhurnal m-va nar. prosveshcheniya Nov. ser. (Ch. S. 1912, Ch. 40), 207-46. Wundt, Wilhelm. Logik. Eine Untersuchung der Prinzipien der Erkenntnis und der Methoden wissenschaftlicher Forschung, Band I: Allgemeine Logik und Erkenntnistheorie, Stuttgart 1919 (4), 1880 (1), 1893 (2), 1906 (3). Band H: Logik der exakten Wissenschaften, Stuttgart 1920 (4), 1883 (1), 1894 (2), 1907 (3).

This page intentionally left blank

J.N. MOHANTY

THE CONCEPT OF ‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

Both Frege and Husserl are well known to have campaigned against a philosophical position known as psychologism. I will begin this essay by focusing upon the nature of psychologism in general and then upon the specific sort or sorts of psychologism Frege and Husserl were up against. Next, I will look at the arguments they pressed against that position. Finally, I will ask if a radical antipsychologistic position is a defensible one.

1. Psychologism: An Ally of Idealism or of Naturalism? It is time to recognize that Frege’s attack on psychologism could not have been motivated by the idea of attacking (Hegelian) idealism. Hegel’s idealism had ceased to be a dominant philosophical theory by the later half of the nineteenth century, and some philosophers even regarded psychologism as their way of correcting the alleged excesses of Hegelian idealism. For example, as Frank Kirkland (1993) notes, Jakob Fries’s psychologism was advanced as an anti-Hegelian position. In addition, Kirkland cites two passages from Hegel that clearly show Hegel’s extreme hostility and negative attitude toward any intrusion of psychological considerations into logic. Not unlike Frege and Husserl, Hegel insisted upon keeping Vorstellung and Gedanke apart. Vorstellung, for Hegel, is pictorial thinking, which should be kept apart from pure thought. Pictorial thinking is an ally of naturalism, depending for its categories on things outside of logic. Pure thought derives its categories from its own internal development. Kirkland is therefore led to as sert that Hegel’s critique of psychologism is an instance of his critique of pictorial thinking, which is indirectly a critique of naturalism. It would seem, then, to be more in order of things that Frege’s critique of psychologism, instead of being a critique of Hegelian idealism, was rather intended to be a part of his overall rejection of naturalism. 113 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 113-130. © 2003 All rights reserved. Printed by Kluwer Academic Publishers, the Netherlands.

114

J.N. MOHANTY

Confusion about this link between psychologism and naturalism is muddied by Frege’s action of Husserl, in his review of the latter’s Philosophie der Arithmetik (1894), of reducing everything objective to the subjective, so that psychologism is maintained as leading to a sort of phenomenalism. Michael Dummett’s linkage of psychologism to idealism — which he now thinks was a mistake (1991, 80) — was possibly encouraged, not alone by Frege’s misleading review of Husserl’s work, but also by various other statements by Frege. One such statement is this: “Psychological treatments of logic...lead then necessarily to epistemological idealism. Since all knowledge is judgmental, every bridge to the objective is now broken off. This flowing into idealism is most remarkable in physiological psychology, because it is in such sharp contrast to its realistic starting point” (1969, 115). What Frege asserts here is that psychologism leads not to phenomenalism or to subjective idealism, but to “epistemological idealism,” namely, the thesis that knowledge is confined to ideas and cannot reach the real objects outside of the mind, and so to a sort of skepticism. More important still for Frege is that psychologism cannot preserve the objectivity of logical and mathematical cognitions for, as he wrote, “neither logic nor mathematics has the task of investigating minds and the contents of consciousness whose bearer is an individual person” (1967, 359). I will not, on this occasion, attempt to decide whether Frege was a realist, as Dummett takes him to be, or an objective idealist, as Hans Sluga takes him. But one thing is clear: Frege’s antipsychologism is motivated by the goal of preserving the objectivity of logical and mathematical truths and of setting aside the possibility of skepticism with regard to any cognition. The sort of psychologism that threatens to have these ruinous consequences is, for him, based on a naturalistic philosophy. Physiological psychology and empiricistic psychology of the British empiricists are two such theories. Neither the Kantian nor the Hegelian idealism has those consequences — in particular, not the Hegelian, which sought vehemently to keep psychological contents entirely separate from the categories of logic.

2. Types of Psychologism Many authors have attempted a taxonomy of the various types of psychologism (see esp. DeBoer 1978, esp. 116-17; Seebohm 1991; and Cussins 1987). Without recalling those attempts, I want to begin by distinguishing ‘naturalistic psychologism’ from ‘transcendental psychologism’. While Husserl generally used ‘Psychologism’ for the former, he often, especially in the Formal and Transcendental Logic (1969; see esp. 254-55), used the latter designation to stand for a variety of psychologism that his refutation in the Prolegomena (1928, §§28, 38) had left unaffected. It seems that by ‘transcendental psychologism’ he has in

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

115

mind the Kantian sort of thesis that the pure forms of logic have their origin in the a priori structure of the human mind (so that pure transcendental logic has its foundation in a tran tal psychology). I will return to this variety of psychologism later in this essay, but for the present I will focus upon the former, that is, naturalistic psychologism. The “naturalism” that underlies naturalistic psychologism is, according to Husserl, a metaphysical thesis that regards the world as con ing only of individual things, cognizable by outer and/or inner perception, and the psychology based upon such a theory as one that undertakes the task of inquiring into causal relations among such individual mental contents and other mental contents and physical entities. This naturalism can be held by realists and idealists alike (see Levinas 1973, xxxvi, 97, 14-15). Naturalistic psychologism may then be upheld by either phenomenalists or by psychophysical researchers or by many other asserted varieties of psychological theories that are founded upon a naturalistic world view (see Margolis 1995, esp. 88-89). Both Frege and Husserl reject them all, not psychological theories per se, but insofar as these psychological theories claim to provide the foundation for logic and mathematics and — only in Husserl’s case — insofar as they advance a theory of the nature of consciousness. Under the general rubric of ‘naturalistic psychologism’ can be distinguished various types, which depend upon the fundamental concept to which the naturalistic psychology is being applied. Thus we have: (A) logical psychologism, i.e., psychologism with regard to logic; (B) psychologism in theory of meaning; (C) psychologism in theory of truth; (D) epistemological psychologism; and (E) psychologism applied to metaphysics.

In each case, there is a corresponding antipsychologism, and possibly a strong antipsychologism and a weak one. Frege was opposed to (A) through (C), and he possibly thought he was opposed to (D), but as a matter of fact he was not. Husserl was opposed to (A) through (D). Neither had anything to do with (E). In this essay I will not be concerned with (E), which includes any attempt to use a psychological category for understanding/interpreting the nature of reality. A good example of this is Schopenhauer’s characterizing Reality, the Kantian Ding an sich, as Will. Setting aside such attempts to ground metaphysics upon a psychological theory, let me briefly explain (A) through (D). Logical psychologism is the theory that logic is based upon psychology. As has been noticed by many, there is both a weak form and a strong form of logical psychologism. Of the various accounts avail able of this distinction, I will mention here the one given by Hanna (1993): according to strong logical psychologism, all pure logical propositions (truths, principles,

116

J.N. MOHANTY

axioms) are necessarily dependent on the human mind, are necessarily contained in the human mind, so that logical objectivities are mental entities. For the formulation proposed by Hanna, the thesis may be stated perspicuously as: (1) Necessarily y thinks x))

(x is a logical proposition

(y is a human being &

However, note that ‘y thinks x’ (in the above) itself allows of a psychologistic reading, meaning that if y thinks x, then x is a mental entity. If one rejects this reading and allows that ‘y thinks x’ does not necessarily entail that x is a mental entity, then (1) does not represent strong logical psychologism. For even if (1) is true, it may be that x is not a mental entity. So I would revise (1) into: (1 *) Necessarily (x is a logical proposition & x is a content of y’s mind))

(y is a human being

Weak logical psychologism, in Hanna’s formulation, amounts to (2) (x is a logical proposition thinks x))

Possibly

(y is a rational thinker & y

Replacing ‘y is a human being’ in (1*) by ‘y is a rational thinker’ in (2) has the advantage that, while (1*) implies anthropologism and relativism with regard to a natural kind, (2) does not. Rather, it uses a “structural” concept of a “rational thinker.” Of Frege, we can say that he rejects both (1) and (1*) as well as (2). Husserl certainly rejects (1), perhaps also (1*), but it may be argued that he does not reject (2). We have noticed that ‘y thinks x’ allows of two readings: one psychologistic, the other not. For one reading, it means ‘x is a content of y’s mind’; for the other, it means ‘x is an object of y’s thinking’. One may want to point out that this concerns ‘meaning-theoretic psychologism’. To hold that ‘y thinks x’ equals ‘x is a content of y’s mind’ is to opt for a psychologistic theory with regard to meanings or thoughts, whereas to hold that ‘y thinks x’ equals ‘x is an object of y’s thinking’ is to opt for a nonpsychologistic theory of meaning. Both Frege and Husserl rejected identification of meanings with images or any other mental entities, for both Sinne are objective entities that transcend the mental experiences (or acts) of any thinker. From psychologism with regard to meanings, one can distinguish psychologism with regard to ‘truth’. Both Frege and Husserl want to keep the idea of truth secure from intrusions of psychology: Frege because logic

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

117

for him is an elaboration of the idea of truth, Husserl because for him psychologism leads to relativism regarding ‘truth’ and so is incoherent. Psychologism with regard to ‘truth’ leads, according to Frege, to identifying ‘true’ with ‘taken to be true’, which he rejects but is not sure that he can refute. Husserl, in contrast, takes enormous pains to refute the position, but concedes that he would not be able to persuade the psychologistic and relativistic thinkers (see Føllesdal 1994). The ultimate constituents of ‘truth’, for Husserl, are ‘meaning’ and ‘object’. Meanings are, for his (and Frege’s) theory, ideal entities transcending the subjective life of persons. Objects, by definition, are ontological. The adequacy of meanings to objects is truth. Truth by itself is an ideal entity, not an individual being, thing, event, or fact. Even if there are no intelligent creatures to be aware of truth, truth remains in itself what it is, it retains its ideal being. It is “a unity of validity” in the timeless realm of ideas. As Frege would have said, truth is independent of human judgment. To relativize truth is to relativize being, for ‘truth’ and ‘being’ are correlative categories (Husserl 1928, 129-32). Can there be an epistemological psychologism as distinguished from all the above three? If, as Kitcher (1979) has maintained, Frege’s epistemological theory is psychologistic, and if, as would be uncontroversial, his theses on logic, meaning, and truth are antipsychologistic, then it would seem that one can distinguish between epistemologiccal psychologism and the three others. Kitcher recognizes that Frege’s antipsychologism is restricted to preserving the objectivity of truth; guarding against identifying meanings with subjective ideas, as well as against taking mathematical objects as ideas; and also rejecting the view that logic is concerned with the actual process of thinking. However, according to Kitcher, all these antipsychologistic doctrines are compatible with an epistemological theory whose basic framework was taken over from Kant’s inventory of sources of knowledge and types of cognition (analyticsynthetic, a priori, a posteriori). To the extent that Kantian epistemology does not presuppose a naturalistic world view, the resulting psychologism of Frege’s epistemology, I would add, is not “naturalistic” in the sense about which we are now talking. Husserl also wanted to avoid a naturalistic epistemology. Levinas expresses the basic idea of a naturalistic epistemology by saying that it “places subject and object in the same world, which it calls nature, and studies their relation as a relation of causality,” (1973, 15). Husserl had to reformulate the problem of knowledge radically — in this sense, Frege’s was closer to Kant’s while Husserl, though Kantian in a very general sense, rejected the Kantian idea of “faculties” and “sources” of cognition and the Kantian taxonomy of judgments. The ideas of intentionality, meaning fulfillment, evidence, and constitution shape both his epistemological problem and his solution — still leaving a flavor of psychologism, though of a

118

J.N. MOHANTY

nonnaturalistic sort. Both Frege and Husserl were aware of the epistemological discussion within the neo-Kantian circle, especially of Marburg. Hermann Cohen had rejected psychologism in favor of an objective theory of knowledge. Cohen’s program was to ground the fact of science objectively. Natorp, in a well-known essay, “Über objektive und subjektive Bergründung der Erkenntnis” (1887), showed the superiority of objective grounding of the sciences over subjective grounding, the object being what is valid independently of the subjective. The object, for Natorp, equals the law, which again refers to the process of scientific objectification. The scientific object is constituted inasmuch as appearances are governed by laws; the ultimate foundation of science lies in the Gesetzlichkeit itself. The so-called subjective foundation lacks “fundamental scientificity,” lacks “insight into principles.” While Natorp had welcomed Husserl’s purely objective account of logic in the Prolegomena, Husserl’s further attempts to give a phenomenological foundation for logic were criticized by him as amounting to a reversal to psychologism. It is important to ask in what way these forms of psychologism or antipsychologism are mutually interconnected or interdependent. For Husserl they are all of a piece. Logic deals with ideal meanings of words and sentences, and, insofar as logic is concerned with truth, truth, too, must be objective. Rejecting psychologism in these three areas of problems, one can possibly still subscribe to a sort of psychologism in epistemology as one does when one formulates the epistemological problem in terms of justification of beliefs.

3. The Antipsychologistic Arguments of Husserl: A Résumé Husserl gives the most wide-ranging and painstaking critique of psychologism that is to be found in the philosophical literature. The bulk of this critique is concentrated in chapters 4 through 8 of the Prolegomena, but there are also criticisms elsewhere in the Logische Untersuchungen. For a criticism of a psychologistic theory of meaning, one should turn to investigation 1 of the Untersuchungen, where Husserl argues for the distinction between psychological contents (images, accompanying intuitions, and/or feelings) of experience of understanding or of meaningfully using expressions or of thinking, and their logical contents; also between acts of meaning and the meaning intended, between judging and the proposition asserted or denied, between acts of inferring and the syllogism, between having an idea and the concept entertained. Terms such as ‘idea’, ‘judgment’, ‘affirmation’, ‘denial’, and ‘inference’ are said to be psychological or “psychologically slanted,” playing no role in any pure theory. The scientist asserts propositions, but says the propositions are true

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

119

or false: “his premises are propositions, and so are his conclusions” (1928a, Bd. 2, Inv. 1, §29). To sum up: “All theoretical science consists, in its objective content, of one homogeneous stuff: it is an ideal fabric of meanings”, and meanings are “ideal unities of validity” like the the theories they constitute (§29). We can then say that if psychology itself is to be a theory, it must consist of such ideal meanings (concepts and propositions) and contain no “psychological stuff.” For Husserl’s criticism of a psychologistic theory of truth, one should refer to the Prolegomena‚ sections 24, 36, 39, and 50. At all these places, while refuting relativism, individual or specific (which Husserl takes psychologism to imply) and while refuting what he calls “anthropologism” (of Sigwart and Erdmann), he makes the same point: relativism cannot be coherent. I will return to this charge of incoherence later in this essay. For the present, I would stress the claim that “no truth is a fact, i.e. something determined as to time” (1928b, §24). Husserl is not denying that there are changes, so he adds, “A truth itself is, however, raised above time, i.e. it makes no sense to attribute temporal being to it, nor to say that it arises or perishes” (§24). If it is true that it is raining in Ambler at time this truth itself does not occur in time; the truth is timeless. The timelessness of truth would render, Husserl thinks, any relativistic, and so psychologistic, theory of truth absurd. But why this should be so we are not told by him. Once meanings and truth are shown to be ideal entities, and so not real things (individuated by time and by the mental life of a person), it becomes, for Husserl, easier to reject psychologism with regard to logic. The criticism shows that already in the Prolegomena Husserl was searching for a middle ground between psychologism and antipsychologism while conceding that the truth lies more on the antipsychologistic side (1928b, §20). Chapter 3, which reviews a series of arguments and counterarguments by the opposed parties, brings out that the antipsychologistic logicians were misled when they thought that they could refute psychologism by placing emphasis on the normative conception of logic (which, by the way, is one of Frege’s arguments against psychologism) (Frege 1964, xvi). As a matter of fact, it might seem as though Husserl’s sympathy lies with the psychologistic logicians’ response that every ought rests on is. But the psychologistic logician, taking logic as a technic, a Kunst, argues that no technology can ignore causal connections. Husserl, however, uses “ought presupposes is” to maintain that every normative dis pline presupposes a theoretical discipline. The normative sentence “An A ought to be B” presupposes the theoretical sentence “An A that is not B is a bad A.” In this sense, normative logic presupposes “pure logic.” From Frege’s point of view, it is not necessary to distinguish between the ideal and the normative. Husserl, however, not only distinguishes between them, but, by grounding the normative upon the ideal, more successfully overcomes psychologism.

120

J.N. MOHANTY

Chapter 4 of the Prolegomena draws out the consequences of psychologism. First, as psychology is an empirical science, its laws cannot be exact, cannot help being vitiated by vagueness. Logical laws, in contrast, are exact. Psychologism in theory of logic would have the absurd consequence of rendering the laws of logic vague and inexact. Second, no natural law can be known a priori; they all are established by induction. Logical laws, on the contrary, are a priori established by “apodictic evidence” (an idea to which I will return later in this essay). This distinction would be obliterated by psychologism. Third, psychologism construes logical laws such as the well-known principle of noncontradiction as natural laws governing human thinking in a rather causal manner. However, one cannot establish a causal law by apodictic evidence. Moreover, in order to exercise a causal influence on thinking, the law must be a judgment or a cognition, itself a mental entity, but such an assumption already confuses the distinction between the law and one’s thought, or cognition of the law. Still more pertinent, one would be confusing the distinction between the law as a term in causation with the law as the rule of causation. To bring out this latter confusion, Husserl imagines an ideal person whose thinking entirely proceeds “as logical laws require,” or even a computer whose procedure would be causally ex plained by appealing to the initial arrangement regulated by natural laws that agree with those of arithmetic. But neither the ideal thinker nor the machine thinks in accordance with the laws. Finally, no logical law presupposes the existence of persons with mental experiences such as judging. The laws are about truths, but truths are not facts (Prolegomena §23). In chapters 5 and 6, Husserl considers some specific attempts (especially by Mill and Sigwart) to give psychologistic renderings of principles of logic — of the principle of noncontradiction and the principles of syllogism in particular — and argues that these renderings should be rejected. For example, Husserl wants to know if the principle of noncontradiction is taken to be the empirical, psychological law that it is impossible for us to believe in an explicit contradiction, what is meant by impossible in this statement and what are the circumstances under which opposed beliefs cannot coexist? One perhaps would say something like “in a single consciousness,” or “in a normal conscious in a reasonable person’s mind,” each of which is too vague to enter into the formulations of a law. Basically what the psychologistic rendering does is replace “not-both-being-true of two contradictory propositions” by “incapacity to entertain contradictory beliefs.” Chapter 7 interprets psychologism as leading to relativism; distinguishes between two kinds of relativism, individual and specific; and undertakes to refute each. Specific relativism becomes anthropologism if one restricts the species to the human kind. Re ing individual relativism, Husserl recognizes that one cannot refute it, for “refutation presupposes the leverage of certain

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

121

self-evident, universally valid convictions,” and yet proceeds to argue that the content of the relativist’s assertion (that ‘true’ is always ‘true for me’) contradicts “what is part of the sense or content of every assertion” (namely, that if true, it is true for every one) (1928b, 35). Is this then intended to be a refutation? What if the individual relativist does not accept the absolute validity of the principle of contradiction, or the claim that every assertion carries the claim that if true it is true for every one? What if the relativist of this brand does not recognize the concept of a proposition at all? As against specific relativism, Husserl argues — as before — that the thesis is in conflict with its own sense, meaning thereby that ‘truth for this, e.g., the human species’ is an absurd idea, and the relativist gives the words true and false new meanings. But this argument takes for granted precisely what the relativist denies, namely, that ‘truth’ means ‘absolute truth’. The anthropologist’s thesis ‘If there were no such constitution (viz. the human), there would be no truth’ is regarded by Husserl to be absurd insomuch as the consequence ‘There is no truth’ is equivalent to ‘There is a truth that there is no truth’ while the antecedent is the negation of a fact and so is false, not absurd. Thus, the thesis deduces an absurdity from a meaningful antecedent. Pierre Adler (1993) has distinguished between two kinds of criticism Husserl advances against psychologism-cum-relativism. One of them is what Adler calls ‘the ontological critique’ according to which psychologism confuses between the real laws and ideal laws. The other he calls ‘linguistic critique’, which may also be called ‘pragmatic critique’, for the incoherence Husserl brings out is not logical, but linguistic-pragmatic. I think Husserl has two varieties of criticisms under this heading, one of which is logical (‘Logic being a theory of all theories cannot presuppose another theory viz. psychology’) and the other pragmatic (‘The relativist cannot assert his thesis’). One thing seems to be clear from the above exposition. Husserl’s critique of logical psychologism, even if its strength is undeniable, does make use of an ontological dualism between the real and the ideal that Husserl has not yet, in the Prolegomena, established. That is why to a careful reader he gives the impression of begging the issue, that is, of making use in the premises the very distinction between psychology and logic that he wants to prove. In the first logical investigation, he shows why meanings have to be ideal entities; later, in the third and the fourth investigations, he shows why logical laws have to be analytic (and in what precise sense). I think Husserl’s strongest arguments are those that point out an incoherence in the psychologistic and relativistic positions. The arguments are still valid against a psychologism that does not make use of the empiricistic psychology of the sort Husserl (and Frege) had in mind. Husserl’s point that if mathematics is fundamentally based upon logic, any psychology that, like the modern cognitive science, makes use of mathematics cannot provide the foundation of logic would still be valid.

122

J.N. MOHANTY

Nevertheless, it must be conceded that in the Logical Investigations Husserl has refuted not all forms of relativism, but only those that are consequences of the sort of psychologism with which he was concerned.

4. Responses to Husserl’s Antipsychologism Before briefly recalling the responses to Husserl’s antipsychologism,1 I will note that the important distinction between logic and psychology was first made not by Frege or Husserl, but by Kant: “Die reine Logik...keine empirische Prinzipien” and “nichts...aus der Psychologic schöpft” (Kant 1787, B78). Kant also writes, “Nähmen wir die Prinzipien aus der Psychologic, d.h. aus der Beobachtungen über unseren Verstand...dies wurde also zur Erkenntnis bloss ‘zufälliger’ Gesetze führen. In der Logik ist aber die Frage nicht nach zufälligen, sondern nach notwendigen Regeln; nicht, wie wir denken, sondern, wie wir denken sollen” (Kant 1800, 334). Here Kant makes recourse to a normative conception of logic. Earlier, I referred to Hegel’s rejection of psychologism. Herbert and Lotze are the two major philosophers after Hegel who also rejected psychologism. Thus writes Herbert: “In der Logik ist es nothwendig, alles Psychologisches zu ignorieren, weil hier lediglich diejenigen Formen der möglichen Verknüpfung des Gedachten sollen nachgewiesen werden, welche das Gedachte selbst nach seiner Beschaffenheit zulässt” (1891, 68). As regards Lotze, it is well known that Husserl acknowledges Lotze’s influence on him insofar as his overcoming of psychologism is concerned (Husserl 1938, 128f.). Among Husserl’s contemporaries — leaving aside those who greeted the Prolegomena with unqualified enthusiasm — the antipsychologism was received with mixed feelings and several kinds of criticisms. Some found his conception of psychology defective, and others critiqued his conception of logic. Some greeted his antipsychologism, but wished that, after logic and psychology had been separated, steps should have been taken to bring them into a satisfying connection. Others found his arguments against psychologism defective. Still others lamented Husserl’s own relapse into psychologism. There were also philosophers who were criticized by Husserl as being psychologistic, but who rejected this characterization. To this last group belonged such men as Brentano and Schuppe. Husserl, of course, resisted offending his teacher by characterizing his position as psychologistic in the pejorative sense, but the fact that Brentano had to defend himself against such a charge shows that Brentano suspected Husserl’s secret intention in this matter.2 Schuppe had already said in 1879 that the task of logic was to separate the pure Gedankenelemente not only from their linguistic garbs, but also from acts of thinking (2d ed., 1910, 1-3). So he, too, rejected Husserl’s characterization of his position as being

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

123

psychologistic. Benno Erdmann, referring to Husserl’s criticism of his view, conceded that some of the criticisms were justified inasmuch as he himself did not correctly state his views and had distinguished logic from psychology in a rather unsatisfactory manner. Nevertheless, he added, Husserl had totally misunderstood him, so that without himself deciding who was right in this case, he would rather leave it to the “ Unbeteiligten der jungen Generation” (1907, Bd. 1, 532-33 fn.). The most negative response came from Wundt (1910). Wundt begins by conceding that he does not know who coined the name “Psychologism,” but regards the name “logicism” as being still more recent. Wundt then traces psychologism to British empiricism, while regarding logicism as being as old as speculative metaphysics. Husserl’s program, he continues, is to expel psychologism, not only from logic, but also from psychology (569), for what Husserl eventually aims at is a rationalistic psychology. He goes on to criticize Husserl’s concept of mental phenomena as acts. He finds this conception of act to be completely formal and empty of contents and accuses Husserl of using “scientifically unexamined concepts of vulgar psychology” (579). Like his “descriptive psychology,” Husserl’s pure logic is scholastic and concerned with word meanings. Husserl’s so-called ideal meanings are his creations. There is no such invariant meaning. The only compliment he pays to Husserl, his erstwhile student, is that he is an excellent critic, such that “Im Zerstören sucht er seinesgleichen” (608). But he hardly substantiates his positive thesis, except by appealing to evidence; but evidence itself is an Erlebnis and so a psychological phenomenon. Sigwart (1911), whom Husserl criticized as being psychologistic, defends himself against Husserl’s criticisms. Against Husserl’s criticism that according to Sigwart ‘truth’ requires an intelligence that thinks it true, Sigwart contends that true or false, in the original senses of those words, can be ascribed only to an assertion or an opinion and that an assertion or an opinion necessarily presupposes a thinking mind. To hypostatize propositions into autonomous essences is to set up a mythology (24 fn.). If a judgment has not been made (as, e.g., the Law of Gravitation before Newton), it is then neither true nor false. Sigwart asks rhetorically, “ Kann man von einem ungeborenen Menschen sagen er gesund ist, well er gar nicht existiert, ist er dann krank?” (24-25). Sigwart also asks a question (which I will return to later in this essay), namely, if the process of psychological thinking does not lead to logical cognition, how then do we acquire knowledge of logic? Husserl’s position leads to two consciousnesses — one empirical-actual (which must be completely independent of logical laws) and one ideal (which apprehends the nontemporal truths). How then could the one reach the other? Husserl, he points out, has a concept of psychology that studies mental life under causal laws and so reaches only vague generalities. It does not analyze selfconsciousness and discover in it consciousness of logical necessity. He sets

124

J.N. MOHANTY

up a “ wind gegen die er kampflustig seine Lanze einlegt” (Sigwart 1911, 25). Yet Husserl also says that the certainty of logical laws is an Erlebnis. Is not an Erlebnis an empirical psychological fact? Is not an evidence a state of the mind that we experience in time? Natorp (1977), in his review of Husserl’s book, both recognizes the value of Husserl’s antipsychologistic arguments and questions whether philosophically one can rest with Husserl’s absolute separation between the real and the ideal, the psychological and the logical. While agreeing with Husserl’s conception of pure logic, Natorp questions if pure logic has to be formal. In his view, pure logic also has to be material, as Kant’s transcendental logic shows. For this issue, one needs to recall that Husserl’s pure formal apophantic logic is based on pure logical grammar, which is not merely a theory of pure syntax, but also has to rule out material countersenses such as “Virtue is green”; for this last purpose, pure logical grammar has to take into account which material concepts go together and which not. This last must appeal to the way the world is structured, as Husserl comes to recognize in the Formal and Transcendental Logic. Natorp also takes Husserl to task for finding psychologism in explicitly antipsychologistic authors, remarking, “One who wants to find the psychological, will find it everywhere, also in Husserl.” Just recall Husserl’s use of insight and evidence. Natorp, of course, rejects Husserl’s psychologistic reading of Kant, as though Kant’s so-called “transcendental psychology (of faculties of the soul)” is also a psychology. But Natorp’s main criticism of Husserl is this: Husserl leaves the problematic in an unsatisfactory state. After having separated logic from psychology, Husserl still needs to relate the two. Otherwise, the material, empirical, psychological, that is, the real remains “an uncomprehended, irrational surd” (Natorp 1977, 65). Husserl has to explain “why [it is] that the deepest investigation into the constitution of objectivity cannot avoid taking subjectivity into consideration” (66).

5. Some Recent Appraisals Among more recent commentators, I will mention only three. On the one side, there are Dallas Willard and Barry Smith; both use a most interesting reading of Husserl to show why and how, in spite of his use of an act-based theory of meaning, Husserl still succeeds in avoiding psychologism. On the other side, there is Michael Dummett; Dummett finds Husserl’s act-based theory of meaning as not only leading to psychologism (which he, in agreement with Frege, rejects), but also as giving rise to several serious problems. Willard (1977) first formulates what he calls “the paradox of logical psychologism” that arises out of two conflicting claims: (i) that the logical

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

125

truths are about and apply to particular conceivings, inferrings, and so on, of particular persons; and (ii) that they do not derive their evidence from examination of these mental events of particular persons, Husserl’s solution consists in maintaining that such entities as propositions (in general, ideal meanings) relate to particular acts (of judging, for example, in which the same proposition is entertained) as the species Redness relates to particular instances of red color. Propositions, then, are not objects of acts of thinking, but what Willard calls “complex referential characters or qualities of such acts.” The solution of the paradox then is that the logical truths are primarily about, not acts themselves, but certain universal characters of those acts. The truths, therefore, apply to the acts and are derived from examination of those acts — while still being strictly universal. In a later essay, Willard writes, “For Husserl, the signification is never the object in the act...Rather, it is the intentional character or property of the act or expression” (1994, 256). The act is not directed towards the sense, but exemplifies it. Barry Smith (1994) develops a similar interpretation of Husserl. (Note that both Willard and Smith prefer the theory of meaning that Husserl had in the first edition of the Logische Untersuchungen and that he abandoned very soon afterwards in favor of his famous concept of the noema.) Comparing the competing strategies of Frege and Husserl for overcoming psychologism, Smith raises two questions for both: First, how do the mind-independent senses relate to the psychological acts of the mind? Second, how are we able to grasp the senses? His answer to these two questions lies in the way Husserl conceives of the relation of senses to mental acts. Mental acts are complex events that can be sliced into dependent parts in many ways. Apart from such parts as sensory contents and Auffassungsmomente, there is also an act-part, which is none other than the act’s object-directedness. A sense is the universal species that is instantiated in such act-parts (of many different acts directed toward the same object in the same way). The mental acts therefore are necessarily subject to the laws obtaining among such species, just as geometrical laws govern real spatial figures. Dummett (1991) is not convinced that Husserl’s act-based theory of meaning can explain how logical truths are grasped and how logic applies to mental acts. Recalling Frege’s well-known exasperation regarding the “mystery” surrounding the idea of “grasping” a thought, Dummett concedes that “it is in taking thoughts as objects of mental acts that Frege goes astray” (253). Frege should have said that they are rather ways in which things are given to us. This is certainly an important improvement upon Dummett’s earlier view, namely, that Frege avoids psychologism while Husserl gets caught in it because of Frege’s attention to language (and Dummett’s Wittgensteinian view of linguistic practice) and Husserl’s concern with acts. Dummett should have recognized that making meanings dependent upon linguistic practice would have led to what Husserl calls “anthropological” or “specific” relativism. But Dummett now recognizes that “where both

126

J.N. MOHANTY

Husserl and Frege failed was in drawing an absolute line of separation between the psychological and the logical.” They thereby “deprived themselves of the means to explain what it is to grasp a thought” (256). As regards the claim by Willard and Smith that the theory of meaning developed by Husserl in the first edition of the Logische Untersuchungen solves this problem about “grasping,” we should ask, if the act instantiates a sense, how can it grasp the universal instantiated? Husserl’s obvious answer is “by reflection.” The prereflective act is object-oriented; it grasps the sense only by reflecting upon its object-directedness. It does not initially grasp the sense. Here Husserl’s formulation was more cautious than Frege’s.

6. Problem of Logical Cognition The problem of what is called ‘logical cognition’ is not merely the problem of how it is possible to grasp such ideal (or abstract) entities as propositions (with which Frege was concerned and to which Smith and Willard find a cogent reply in Husserl), but also the problem of how logical truths are apprehended. Husserl was concerned more directly with the latter problem than with the former. It is, of course, obvious that the latter presupposes the former. The mind must be able to grasp propositions in reflection before it can apprehend logical truths, for logical truths are either propositional truths or involve propositions. With regard to the second problem of logical cognition, perhaps the best account and defense of Husserl’s theory is to be found in Hanna’s paper “Logical Cognition: Husserl’s Prolegomena and the Truth in Psychologism” (1993). Hanna takes Husserl to have rejected a strong logical psychologism, but not weak logical psychologism. According to weak logical psychologism (which, in Hanna’s formulation, it is correct to say Husserl held), (2*) (x is a logical proposition Possibly (y is a finite rational thinker & x is an object of y’s thinking)) (allowing for the changes I introduced in [2] earlier). Note that a ‘logical proposition’ here is not just any proposition or any thought the mind can grasp, but a logical truth. (But one may generalize the thesis to any proposition or thought and hold that, if x is a thought, then it is necessarily possible that there is a finite rational thinker who thinks it.) According to Hanna, Husserl’s view is that only nonempirical or rational insight yields the sort of evidence that is rationally conclusive for belief in a logical truth. Experience of self-evidence, as Husserl states in the Prolegomena (1928b, §51), does not guarantee the truth of a proposition. But if a proposition is true, then it necessarily must be possible for some rational thinker to have the experience of self-evidence about it. If the

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

127

proposition is a contingent factual truth, the experience of self-evidence will be connected with empirical perception; if the proposition is a logical truth, then the experience of its self-evidence must be connected with a rational, nonempirical insight. That is, all logical truths are thinkable through rational insight — but there does not necessarily have to be someone who has rational insight into it. Husserl clearly will reject the empiricist claim that all knowledge is justified by appeal to inductive generaliztion on the basis of empirical evidence. This latter thesis is inadequate to account for modal knowledge. What is important is that the appeal to insight is not falling back to psychologism; it does not amount to grounding logical truth on a psychological experience. For the same reasons, grounding empirical truths on empirical perception does not amount to psychologism. First, it is taken to be not a theory of truth, but rather an epistemological theory of experience of truth, so there need be no worry that truth itself is being psychologized. Second, the thesis 2 is intended to be a conceptual truth and not a contingent truth about human minds (Husserl 1928b, §50). It is not intended to be an inductive generalization.

7. Husserl’s Problems Husserl needs to: (1) have a satisfying way of connecting the logical and the psychological after the ideality of the logical has been secured; (2) have an account of our cognition of the logical; (3) have an account of how the logical truths apply to mental actsof judging, inferring, etc.; (4) develop a nonpsychologistic theory of evidence (and so of cognition and truth); and (5) be able to overcome naturalism in a more radical way than he does in the Prolegomena. (6) He should also take Kant’s transcendental psychologism more seriously and be able to show why and how it is that objectivity is constiuted in subjectivity just as there is no subjectivity without objective directedness. For this purpose, he needs (7) a satisfactory theory about the relation between the transcendental and the empirical. He needs also to argue (8) that the subjective is not eo ipso psychological (or even neurophysiological), that the psychological is only a certain interpretation of the subjective life of consciousness. Psychologism can be overcome only when the origin of this interpretation is identified.

128

J.N. MOHANTY

It is no wonder that in his writings after the Prolegomena, Husserl searches for a deeper way of both understanding and overcoming psychologism. Such an overcoming would be refuting it not merely by showing its logically (or pragmatically) incoherent consequences, but also by showing the historical origin of such a naturalistic interpretation of the mental life. In the fifth logical investigation (1928a, Inv. 5, §27), he speaks of “psychological apperception,” which has to be overcome through a phenomenological intuition of the essence of the sort of experience that is being investigated. In the foreword to the second edition of the sixth investigation, he writes that the “grotesque reproach” that he had fallen back into psychologism shows that the critics had not carefully read the chapter “The a priori laws of Authentic and Inauthentic Thinking” (1928a, Vorwort, Bd. II, Teil II). At this place, Husserl maintains that the laws of logic are fulfilled by corresponding categorical intuition of the states of affairs intended and serve as norms for inauthentic thinking. These laws have nothing to do with ‘our’ mental organization. They are grounded in the essences of the act-species and act-forms. “What is seen to be incompatible in specie, cannot be brought together and be rendered com ble in empirical instances” (1928a, Inv. 6,§64). Again in the Formal and Transcendental Logic, Husserl refers to the “bogy of Psychologism.” The Logical Investigations, he tells us, “had established the objectivity of ideal structures as against psychologically reinterpreting” them (1969, §56). He makes it clear that in the Prolegomena psychologism pure and simple was never “thematized.” Rather the discussion concerned a psychologism with a quite particular sense, namely the psychologizing of the irreal significations that are the theme of logic. He concedes that at that time he had not entirely overcome “a universal epistemological psychologism.” Now he is in a position to radically overcome “the problem of transcendental psychologism.” I will not in this paper present Husserl’s solution of these problems. But one may venture the suggestion that the subsequent development of his thought into a transcendental phenomenology — with the method of epoché, the idea of noesis-noema correlation, and the idea of constitution — was occasioned by various attempts to solve the problems raised in connection with the problem of psychologism and to estalish a satisfying connection between real mental life and ideal meanings. Department of Philosophy Temple University, Philadelphia, USA NOTES 1 Kusch gives a wide-ranging survey of responses to Husserl on this issue in his “The Criticism of Husserl’s Arguments against Psychologism in German Philosophy 1901-1920”

‘PSYCHOLOGISM’ IN FREGE AND HUSSERL

129

(1994). Brentano’s alleged psychologism, see Chisholm’s “Brentano’s Descriptive Psychology” (1977, esp. 98).

2 For

REFERENCES

Adler, P. “Prolegomena to Phenomenology: Intuition or Argument?”, Graduate Faculty Philosophy Journal, 16, 1993, 3-76. Chisholm, R. “Brentano’s Descriptive Psychology.” In The Philosophy of Brentano, ed. Linda McAlister, 91-100. Atlantic Heights, N.J.: Humanities Press, 1977. Cussins, Adrian. “Varieties of Psychologism”, Synthese, 70, 1987. DeBoer, Th. The Development of Husserl’s Thought. Trans. Th. Plantinga. The Hague: Nijhoff, 1978. Dummett, M. Frege and Other Philosophers. New York: Oxford University Press, 1991. Erdmann, B. Logik. 2d ed. Halle: Max Niemeyer, 1907. FØllesdal, D. “Husserl and Frege: A Contribution to Elucidating the Origins of Phe cal Philosophy.” Trans. Claire Hill. In Mind, Meaning, and Mathematics: Essays on the Philosophical Views of Husserl and Frege. See Haaparanta 1994. Frege, G. “Rezension von E. Husserl, Philosophie der Arithmetik”, Zeitschrift für Philosophie und philosophische Kritik, 103, 1894, 31332. Frege, G. The Basic Laws of Arithmetic. Berkeley: University of California Press, 1964. Frege, G. Kleine Schriften. Ed. I. Angeleli. Hildescheim: G. Olms, 1967. Frege, G. Nachgelassene Schriften. Ed. H. Hermes, F. Kambartel, F. Kaulbach. Ham burg: Felix Meiner, 1969. Haaparanta, Leila, ed. Mind, Meaning, and Mathematics: Essays on the Philophical Views of Husserl and Frege. Dordrecht: Kluwer, 1994. Hanna, R. “Logical Cognition: Husserl’s Prolegomena and the Truth in Psychologism”, Philosophy and Phenomenological Research, 53, 1993, 251-76. Herbert, J. Sämtliche Werke. Langensalza: Hermann Beyer & Söhne, 1891. Husserl, E. Logische Untersuchungen. 2d ed. Halle: Max Niemayer, 1928a. Husserl, E. Prolegomena zur reinen Logik. Bd. 1 of Logische Untersuchungen. 1928b. See Husserl 1928a. Husserl, E. “Entwurf einer Vorrede zu den ‘Log. Unt.’, 1913,” Tijdschrift voor Filosofie, 1, 1939, 106-33, 319-39. Husserl, E. Formal and Transcendental Logic. Trans. D. Cairns. The Hague: Nijhoff, 1969. Kant, I. Kritik der reinen Vernunft. Zweite Auflage. Riga: Hartkroch, 1787.

130

J.N. MOHANTY

Königsberg: Nicolovius, 1800. Kant, I. Logik. Ein Handbuch zu Vorlesungen, herausg. von Jäsche. Immanuel Kants Werke, herausg. E. Cassirer, Bd. VIII. Kirkland, Frank. “Hegel’s Critique of Psychologism.” In Phenomenology East and West: Essays in Honor of J. N. Mohanty, ed. D. P. Chattopadhyaya and Frank Kirkland, 219-44. The Hague: Kluwer, 1993. Kitcher, P. “Frege’s Epistemology”, The Philosophical Review, 88, 1979, 235-62. Kusch, Martin. “The Criticism of Husserl’s Arguments against Psychologism in German Philosophy 1901-1920.” In Mind, Meaning, and Mathematics: Essays on the Philosophical Views of Husserl and Frege, 51-83. See Haaparanta 1994. Levinas, E. The Theory of Intuition in Husserl’s Phenomenology. Trans. Andre Orianne. Evanston: Northwestern University Press, 1973. Margolis, J. Historied Thought, Constructed World. Berkeley: University of California Press, 1995. Mohanty, J. N., ed. Readings on Edmund Husserl’s Logical Investigations. The Hague: Nijhoff, 1977. Natorp, P. “Über objektive und subjektive Begründung der Erkenntnis”, Philosophische Monatshefte, 23, 1887. Natorp, P. “On the Question of Logical Method in relation to Edmund Husserl’s Prolegomena to Pure Logic.” Trans. J. N. Mohanty. In Readings on Edmund Husserl’s Logical Investigations, 55-66. See Mohanty 1977. Schuppe, W. Grundriss der Erkenntnistheorie und Logik. 1st ed. Berlin: Weidenmannsche Buchhandlung, 1878. 2d ed., 1910. Seebohm, Th. “Psychologism Revisited.” In Phenomenology and the Formal Sciences, ed. Seebohm, Føllesdal, and Mohanty, 149-82. Dordrecht: Kluwer, 1991. Sigwart, C. Logik. Bd. 1. 4th ed. Tübingen, 1911. Smith, B. “Husserl’s Theory of Meaning and Reference.” In Mind, Meaning, and Mathematics: Essays on the Philosophical Views of Husserl and Frege, 169-83. See Haaparanta 1994. Willard, D. “The Paradox of Logical Psychologism: Husserl’s Way Out.” In Readings on Husserl’s Logical Investigations, 43-54. See Mohanty 1977. Willard, D. The Integrity of the Mental Act: Husserlian Reflection on a Fregean Problem. In Mind, Meaning, and Mathematics: Essays on the Philosophical Views of Husserl and Frege, 235-62. See Haaparanta 1994. Wundt, W. “Psychologismus und Logizismus.” In Kleine Schriften. Bd. 1. Leipzig: Verlag von W. Engelmann, 1910.

MARTIN KUSCH

PSYCHOLOGISM AND SOCIOLOGISM IN EARLY TWENTIETH-CENTURY GERMAN-SPEAKING PHILOSOPHY

I The importance of the early-twentieth-century psychologism dispute for understanding the development of subsequent Western philosophy has been recognised for some time. It is not hard to see why philosophers regard a proper understanding of this episode as imperative. Undoubtedly the central reason is the emergence of cognitive science since the seventies: many philosophers feel that cognitive scientists are the true heirs of Erdmann, Lipps, Sigwart or Wundt, the very psychologist-philosophers under attack in the anti-psychologistic writings of Frege and Husserl. In this paper, I want to sketch a different link between the historical psychologism dispute and contemporary philosophy. This link involves a third player, or a second historical controversy: the controversy over ‘sociologism’ in the Germany of the 1920s and early 1930s. I too am ultimately interested in understanding, and contributing to, a contemporary issue. This issue is the status of the sociology of knowledge as a challenger to philosophical theories of knowledge. I want to identify historical predecessors, and historical origins, of arguments used in the contemporary debates over the sociology of knowledge. This involves finding out where certain types of argument first arose, how they were transmitted to us, and why we — at least some of us — find them compelling. Perhaps one might call this project a ‘genealogy of antisociologism’. This paper constitutes a very early, and modest, step in this project. My emphasis will be on historical understanding rather than on philosophical evaluation. At least in this paper, I am interested in unearthing what the arguments were, not in whether we should judge them to be good or bad arguments.

131 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 131-155. © 2003 All rights reserved. Printed by Kluwer Academic Publishers, the Netherlands.

132

MARTIN KUSCH

II From the 1880s onwards, a number of German philosophers started to attack what they eventually came to call ‘psychologism’. To opponents of psychologism, the following kinds of arguments seemed indicative of psychologistic tendencies in the writings of their contemporaries: (P1) Logic studies specific ‘laws of thought’. Psychology is the study of all kinds of laws of thought. Therefore, logic is a part of psychology. (P2) Normative disciplines must be based upon descriptive-explanatory disciplines. Logic is a normative discipline with respect to human thinking. Psychology is the descriptive-explanatory discipline with respect to human thinking. Therefore, logic must be based upon psychology. (P3) Logic is a theory about judgements, concepts, and inferences. Judgements, concepts, and inferences are human mental entities. Therefore, logic is a theory about human mental entities. (P4) The touchstone of logical truth is self-evidence. Self-evidence is a human mental experience. Therefore, logic is about a human mental experience (and thus a part of psychology). (P5) We cannot conceive of an alternative logic. The limits of conceivability are our mental limits. Therefore, logic is relative to the human species, and thus a subfield of psychology. In some cases there was considerable dispute concerning the question who actually held these views; we need not enter into these debates here. It seems not altogether unfair, however, to attribute (P1) to Theodor Lipps (1893) and Gerardus Heymans (1890, 1894, 1905), (P2) to Wilhelm Wundt (1880/83), (P3) to Wilhelm Jerusalem (1905) and Christoph Sigwart (1921), (P4) to Theodor Elsenhans (1897), and (P5) to Benno Erdmann (1892). Arguments against these and other allegedly psychologistic authors were provided by many writers, amongst them Gottlob Frege (1884; 1934, 1893) and Edmund Husserl (1900; 1975), as well as the Neo-Kantians Hermann Cohen (1902), Paul Natorp (1887, 1901), Heinrich Rickert (1904), and Wilhelm Windelband (1884, 1909). Of these and the many other critics, Husserl was undoubtedly the most successful. Here is my reconstruction of some of Husserl’s central arguments (of the Prolegomena to his Logische Untersuchungen (1900; 1975)):

PSYCHOLOGISM AND SOCIOLOGISM

133

(H1) If logical rules were based upon psychological laws, then all logical rules would have to be as vague as the underlying psychological laws. But, not all logical rules are vague. Therefore, not all logical rules are based upon psychological laws (1900; 1975 §21). (H2) If laws of logic were psychological laws, then they could not be known a priori. They would be more or less probable rather than certain, and justified only by reference to experience. But, laws of logic are known a priori; they are justified by apodictic self-evidence, and certain rather than probable. And therefore, laws of logic are not psychological (1900; 1975 §21). (H3) If logical laws were psychological laws, they would refer to psychological entities. But, logical laws do not refer to psychological entities. And therefore, logical laws are not psychological laws (1900; 1975 §23). (H4) All forms of psychologism imply species relativism: Logic is construed as being relative to the human species, that is, to human psychology. Species relativism is an absurd doctrine: If the truth of a proposition were contingent on membership in species than would be false for members of But this contradicts the meaning of truth. Truths are unchanging, and independent of the human mind (1900; 1975 §31-39). A good deal of the psychologism controversy concerned the validity and originality of Husserl’s arguments. Several commentators maintained that Husserl's arguments H1 to H3 all begged the question (Heim 1902, Heymans 1905, Lapp 1913, Schlick 1910, 1918). They also challenged Husserl’s claim that the laws of logic did not imply the existence of matters of fact. Schlick maintained that psychological acts of judgement and logical sentences were intertwined, such that the logical sentence and its truth “can never be found independently of the act of judgement” (1910, 405). Husserl’s critics also denied that truths were eternal or independent of humans. As Sigwart put it: “When no judgements have been made, then there is nothing of which “true” or “false” could be predicated” (1921, 23). The debate over psychologism was vigorous. It was not merely the relationship between logic and psychology that was at issue, but the relationship between psychology and philosophy as a whole. This was a pressing issue not least because the practitioners of the then new (experimental) psychology worked in philosophy departments, and because between 1873 and 1913 the number of full professorships held by these 'psychologists' increased from 1 to 10. Practitioners and advocates of the new psychology sometimes claimed that psychology provided the key to real

134

MARTIN KUSCH

progress in philosophy or that it was the heart of philosophy (Kusch 1995, 123-26; Frischeisen-Köhler 1913). This last-mentioned contention was opposed by a quickly growing group of ‘pure’ philosophers — led by the Neo-Kantians, and by Eucken, Dilthey and Husserl. They deemed it harmful to philosophy and its progress that experimental psychology should be regarded as part and parcel of philosophy (Kusch 1995, 160-210). The prolonged debate between the advocates and opponents of psychology as a philosophical discipline was eventually brought to an end by World War I.

III The psychologism controversy of the first two decades of the twentieth century was followed — in the German-speaking world of the late 1920s and early 1930s — by another fierce debate: the dispute over the nature and philosophical aspirations of the sociology of knowledge. Given how central the accusation of ‘sociologism’ figured in this debate, it seems adequate to call it the ‘sociologism dispute’. Much of the sociologism controversy centred around the work of Karl Mannheim but three other authors, Wilhelm Jerusalem, Max Scheler and Max Adler also deserve to be mentioned, at least briefly. (Due to limitations of space, I leave aside Oswald Spengler’s Untergang des Abendlandes (1918).) Jerusalem was influenced by Durkheim and Lévy-Bruhl, and thus he emphasised the role of ‘collective beliefs’. A belief was collective if it was shared by the members of a group, and if it was routinely invoked by group members in their dealings with one another. The more often a belief was thus invoked, the stronger it became. Jerusalem called this process of strengthening a collective belief its ‘social condensation’ (1924; 1982, 37). Although all groups had collective beliefs, such beliefs were particularly pervasive in so-called ‘primitive societies’ (1924; 1982, 39). By contrast, individuals in modern, scientific, societies, were able to distance themselves from their group and its collective beliefs. The beliefs of the modern individual “were directed at the objective world” and thus were true in a way that exceeded the ‘merely intersubjective’ truth of collective beliefs in primitive societies (1924; 1982, 39-40). Historically, the rise of individualism — and thus the ‘direction towards the objective world’ — had been a side-product of craftsmanship; it was through craftsmanship that individuals found out how to act successfully in the material world and what they found out in this way was not immediately dependent upon collective beliefs (1924; 1982,39-40). Jerusalem’s advocacy of individualism had its limits however. In particular, he was highly critical of philosophers like Husserl who:

PSYCHOLOGISM AND SOCIOLOGISM

135

...either fail to notice, or else consciously ignore, the important social factor in the development of knowledge. Stable knowledge cannot be arrived at other than by a process of co-operation; and it is only through the...process of condensation that knowledge gains the necessary permanence and practical usefulness....The subjective feeling of self-evidence...is nothing more than an effect of the socially determined general and confirmed experience (1924; 1982, 46).

The central theme in Scheler’s sociology of knowledge was the criticism of Comte’s three-stage model of progress (from religion through metaphysics to science) (Scheler 1921; 1982). Scheler claimed that religion, metaphysics and science were three different basic forms of knowledge. They were due to different fundamental biological drives, and thus part and parcel of every human condition. None of the three could be reduced to any of the others. The biological drive responsible for science was the ‘drive for domination and power’ (1924; 1982, 71). For each of the three basic forms of knowledge, Scheler asked what social conditions best furthered its growth and development. The development of science was dependent upon the interaction between two social strata: a stratum of “free contemplative human beings”, and a stratum of “humans that have accumulated work and craft experience”. These two strata had to mix in order for systematic research into the natural world to succeed (1924; 1982, 90). Scheler went to great pains to distinguish his sociology of knowledge from Marxist “economism...psychologism, sociologism, and historism”. Economism treated the different modes of production as causes of different scientific theories. Scheler allowed only for a ‘parallelism’ between the two realms, and insisted that developments in the two realms had a common cause: “the inherited structure of drives of the leaders” (1924; 1982, 93-4). Sociologism and historism allegedly wanted to ‘denigrate’ science by reducing it to social factors whereas psychologism put too much emphasis on the motivations and interests of the individual scientist. Scheler thought that his sociology of knowledge avoided both types of mistake. On the one hand, in his sociology of knowledge social factors had at most a selecting function: They explained why a given culture attended to one type of truth rather than to another — with both types of truth being in principle independent of human beings. On the other hand, psychologism could be kept at bay by recognising that the explanandum was not the thinking of the individual but the “origin of the thought apparatus of categories, the objective and comprehensive goals of and the ‘methods’ of research...as all these are effective beyond the will, wish and subjective intention of the individuals” (1924; 1982, 107). Scheler’s sociology of knowledge was discussed in some detail by the Austrian Marxist Max Adler (1925; 1982b). Adler’s main criticism was that

136

MARTIN KUSCH

Scheler’s sociology of knowledge was not really sociological at all. In particular Adler criticised Scheler’s emphasis on drives: “The drive is always something biological and there is no path from biology to sociology” (1925; 1982b, 136). Adler’s own proposal for a sociology of knowledge was influenced by Kant and Marx. Adler argued that the central question of Kant’s epistemology — ‘how is science possible?’ — was a sociological question and that Kant had intended it as such. This could be seen, Adler thought, from the fact that Kant had not posed the question as ‘how can the individual have knowledge?’ Indeed, to substitute this second question for the first was to be guilty of psychologism (1925; 1982a, 177). Adler distinguished between two projects in the sociology of knowledge. The first project was to study the ways in which specific historical forms of knowledge were shaped and determined by social conditions. The second project was concerned with the social nature of knowledge in general; it sought to demonstrate that all forms of knowledge were socially constituted. It was in the latter endeavour that Kant’s first critique was a key resource. Kant had shown how common and shared knowledge enabled humans to live in a common and shared world, and how knowledge could be common and shared in the first place (1925; 1982a, 179).

IV The polemics around Scheler’s sociological writings were only a minor skirmish compared to the discussion surrounding the sociological work of the Hungarian-born Karl Mannheim. Here is first my reconstruction of Mannheim’s sociology of knowledge in six theses: (M1) Knowledge comes in two main forms: The first form — political knowledge and knowledge of the humanities and the social sciences — is ‘bound to existence’ (seinsgebunden), that is, it is ‘situationally dependent’ or ‘relative to a situation’. Such knowledge is ‘ideological’: It is determined by its holders’ conditions of existence (1931; 1936, 243-44). (M2) We need to distinguish between two main conceptions of ideology: According to the ‘particular’ conception of ideology, some of the collectively held beliefs of a group are ideological; according to the ‘total’ conception of ideology, all collectively held beliefs are ideological. The sociologist of knowledge assumes the total conception in a ‘general’ form: She does not set her own beliefs apart from those of other groups; her beliefs too are ideological (1929; 1936, 49-51, 68-9).

PSYCHOLOGISM AND SOCIOLOGISM

137

(M3) The sociology of knowledge concerns itself with the study of ‘ideology’ and ‘utopia’. Both ideology and Utopia constitute cases where the collective knowledge of a group may — but need not — fail to correspond to conditions of existence: Ideologies may misrepresent reality in the attempt to stabilise society; Utopias may misrepresent reality in an attempt to radically change society (1936b, 36). (M4) The sociology of knowledge will become the basis of a new form of scientific politics. In recent decades a new form of intelligentsia has emerged. This intelligentsia is ‘socially unattached’. It is not bound to any particular class. This intelligentsia is able to understand the partial truth of different viewpoints and able to mediate between them. A viewpoint of a group is partially true if it ‘fits’ the conditions of existence of that group (1929; 1936, 137-8). (M5) To say that knowledge (of the humanities and the social sciences) is ‘bound to existence’ is not tantamount to downgrading it as ‘merely relative’. It is true that such knowledge does not possess the objectivity of, say, 2x2=4, but this only means that the objectivity of social-scientific, historical or political knowledge is of a different kind from mathematical knowledge. The sociology of knowledge does not advocate a ‘relativism’ according to which everything is relatively true. Instead, it advocates a ‘relationism’ according to which certain truths can only be grasped relative to a standpoint (1929; 1982a, 331). (M6) In light of the sociology of knowledge, epistemology needs to be radically reformed. On the one hand, epistemology needs to free itself from individualism and thus rebuild itself around the insight that the knower is always the member of a group. On the other hand, epistemology must overcome its current adherence to a ‘static’ conception of truth. According to the static conception of truth, truths and the meaning of ‘truth’ are eternal. Epistemology must acknowledge that truth has a history: The very meaning of ‘truth’ has changed over time. Shifting from a static to a dynamic conception of truth also leads to a new perspective on the relationship between the genesis and the content of judgements. Although the psychological genesis of a judgement might be irrelevant to its content, the socio-political context of a judgement is part of its meaning. Finally, even in the case of 2x2=4 it is misleading to say that there is such a thing as ‘truth in itself’. Speaking in this way is of heuristic value only (1931; 1936, 258-67).

138

MARTIN KUSCH

V All of Mannheim’s main ideas came repeatedly under heavy fire. The Seinsgebundenheit thesis (=M1) met with opposition from both writers who lamented the exempting of the natural sciences and writers who denied the Seinsgebundenheit of even the humanities and the social sciences. As concerns the former line of criticism, Werner Sombart simply noted that “of course all forms of knowledge, including those of the natural sciences...are bound to existence” (1928; 1982, 377). As concerns the latter form of critique, Alfred Weber insisted that there were many forms of knowledge — even forms of knowledge outside of the natural sciences — that were not determined or influenced by the social world (1928; 1982, 373) (see Curtius 1929; 1982, 424; Barth 1934; 1982, 674). As concerns M2, M3, and M4, Mannheim’s focus on ‘ideology’ and politics stimulated many contemporaries to consider the relationship between the sociology of knowledge and Marxism. Some commentators thought that Mannheim’s sociology was a faithful continuation of the Marxist tradition; others insisted that Mannheim had broken away from Marxism. Whether either move was applauded or criticised depended on the commentators own position regarding the Marxist tradition. Praise for Mannheim’s alleged proximity to Marx was rare (Alfred Meusel 1928; 1982, 387). Writers who saw Mannheim as a Marxist tended to be critical of Marxism. Alfred Weber, for example, deplored the fact that Mannheim had failed to pass beyond the old ‘materialistic view of history’ (1928; 1982, 376). Of writers who saw Mannheim as breaking away from Marxism, only Sombart commended Mannheim. According to Sombart, Mannheim had passed beyond materialism by accepting ‘the objectivity of being’ and ‘the reality of the spirit’. This, incidentally, was an interpretation that Mannheim himself endorsed (Sombart 1928; 1982, 377). This leaves those authors who criticised Mannheim for deviating from Marxism. Max Horkheimer clearly belongs here. Particularly anti-Marxist, in Horkheimer’s view, was Mannheim’s talk of ‘conditions of existence’. Mannheim wanted to measure the degree of truth of an ideology by determining how well the ideology ‘fitted’ with the conditions of existence, the ‘being’ (Sein), of those who believed in that ideology. To Horkheimer this criterion was ‘saying nothing’ as long as the concept of ‘being’ remained ‘empty of content’ (1929; 1982, 472). Herbert Marcuse’s criticism belaboured a similar point. Even if a given form of thinking fitted with social conditions, it could still be ideological in the sense of being wrong (1929; 1982, 466). Moreover, Marcuse insisted that there were social truths — for instance that socialism is better than capitalism — that transcended all ideologies (1929; 1982, 472). Otto Neurath agreed (1930; 1982). Neurath was especially hostile to Mannheim’s way of treating bourgeois ideologies and proletarian sciences symmetrically

PSYCHOLOGISM AND SOCIOLOGISM

139

and even-handedly. Neurath insisted that Mannheim’s relativism was appropriate only so far as different proletarian scientific projects were concerned; it was misplaced as a way of finding compromises between bourgeois and proletarian views (1930; 1982, 588). As concerns the specific claims of M2, M3 and M4, Paul Tillich took exception to Mannheim’s distinction between ‘total’, ‘partial’ and ‘general’ conceptions of ideology (=M2). Mannheim’s conception of ideology was not ‘total’ since Mannheim often exempted the sociology of knowledge. And it was not general either since at least one group of people, the freefloating intelligentsia, was able to liberate itself from all ‘dependence on existence’ (Tillich 1929; 1982, 456). Regarding the work of the free-floating intelligentsia (=M3, M4), Hans Jonas objected that it would be of little use to practical politics since it would only ever be able to carry out its task retrospectively (Jonas 1928; 1982, 393). Günther Stern felt that it was incoherent to exempt the free-floating intelligentsia from being seinsgebunden (1930; 1982, 510; see Speier 1930; 1982, 547). Wittfogel suggested that Mannheim’s thinking at this point was influenced by the Social democrats, and thus a ‘cowardly escape’ from the class struggle (1931; 1982, 611). Mannheim’s attempt to steer a path between relativism and absolutism (=M5) did not meet with much enthusiasm either. Horkheimer insisted that Mannheim should have followed Marx on this point. Whereas Marx had taken the view that all knowledge was relative to class, Mannheim was trying to re-introduce an absolutistic perspective. For Mannheim all particular viewpoints ultimately seemed to fit together into one big absolute viewpoint (Horkheimer 1929; 1982, 486). Ernst Grünwald chose a different line of attack. Grünwald accused Mannheim’s sociology of knowledge of being both ‘absurd’ and ‘false’. The charge of absurdity was clearly modelled on Husserl’s criticism of Erdmann and Sigwart. In Grünwald’s text, the charge is justified as follows: ‘“Relationism” claims that all thinking is valid only relative to a standpoint; but for this very sentence, that is, that all thinking is valid only relatively, relationism demands absolute validity’ (1934; 1982, 750). But Mannheim’s ‘sociologism’ was also based on an altogether false assumption. This was the assumption that thinking in general, and judgements in particular were ‘bound to existence’. According to Grünwald this had to be false: Since judgements could be studied from many different viewpoints — those of sociology, psychology, psychopathology and linguistics, for example — judgements in and by themselves could not be reduced to any of these viewpoints (1934; 1982, 707; cf. von Schelting 1934; 1982). Finally Grünwald argued that the sociology of knowledge could never replace epistemology and that research in the sociology of knowledge would always have to presuppose prior epistemological research. Epistemology studied the validity claims of judgements, and these validity claims were

140

MARTIN KUSCH

independent of, and prior to, external factors like social interests (1934; 1982, 701-7). This criticism was of course directed at M6 rather than M5. In assaulting M6, Grünwald did not stand alone. Kurt Singer (1928; 1982) suggested that the sociology of knowledge needed epistemology since only the latter could answer questions of validity (1928; 1982, 383). Robert Curtius saw the sociology of knowledge as necessary subordinate to “a philosophical theory of the essence of the human being...” (1929; 1982, 424). And to Eduard Spranger it was precisely the attempt to — at least partially — replace epistemology that justified the charge of ‘sociologism’. But epistemology could never be replaced: “The idea and meaning of knowledge as such” could not be reduced to social conditions (1930; 1982, 635). Siegfried Marck (1929; 1982) and Hannah Arendt (1930; 1982) made a similar point in metaphysical and ontological rather than epistemological terms. Marck insisted that the essence of the human being could not be derived from her existence; essence preceded existence (1929; 1982, 446). Arendt expressed a closely related distinction using Heideggerian concepts. Arendt argued that sociology could replace philosophy only if it could be shown that that there were no “formal [ontological] structure of human existence”. As Arendt saw it, this proof had not been produced by anyone (1930; 1982,519). VI

I shall not argue here that, like the psychologism dispute, the sociologism controversy was shaped by social and political interests. The social dimension of the dispute is close to the surface, and visible even to the sociologically naked eye. In the last few sections, I have presented the two disputes as separate events. This is inaccurate in more than one respect. For instance, Mannheim too was occasionally accused of psychologism. Lewalter justified the charge with the claim that Mannheim was trying to explain the thinking of the individual; and Marck on the general ground that Mannheim was putting existence before essence (Lewalter 1930; 1982, 563; Marck 1929; 1982, 446). Much more common than the charge of psychologism was of course the accusation of sociologism (e.g. Curtius 1929; 1982, 423; Grünwald 1934; 1982; Marck 1929; 1982, 446; Spranger 1930; 1982, 635). Interestingly enough, the accusers often pointed to a parallel, or a historical connection, between the illness of psychologism and the disease of sociologism. For example, Spranger remarked that “this new sociologism...seems to take over the part of...psychologism” (1930; 1982, 635). Marck saw psychologism and

PSYCHOLOGISM AND SOCIOLOGISM

141

sociologism as the two main species of what he called ‘existentialism’, and what we today would call ‘naturalism’ (1929; 1982, 438-46). Mannheim was not happy with either charge and deplored the tendency in German philosophy “to shout ‘sociologism’ or ‘psychologism’ as soon as a special science starts to tackle general problems of knowledge”. To take such invectives seriously, Mannheim wrote, would amount to 'intellectual suicide'(1929; 1982, 434). Like ‘psychologism’, ‘sociologism’ was defined differently by different authors — if they cared to define the term at all. According to Scheler, sociologism was the ‘mistake’ of thinking “that one can devalue modern science by explaining it sociologically” (1925; 1982, 106). Curtius wrote that “sociologism is the utopian interest-ideology of sociology dressed up in theoretical form” (quoted in Grünwald 1934; 1982, 747); J. SchmidtHartefeld suggested that “It is sociologistic to overestimate the influence of social reality upon cultural phenomena” (quoted from Grünwald 1934; 1982, 746); and Grünwald proposed that: Sociologism can be summarised in the following two thesis: first, all thinking and knowing is bound to existence; and second, this dependence on existence has a bearing on the validity of a judgement (Grünwald 1934; 1982, 748).

No doubt more interesting than the issues of labels, accusations and definitions are questions concerning how different authors in both debates thought about the relationship between psychology, sociology, and philosophy. A first observation here is that significant (alleged) advocates of psychologism had already put forward important ‘protosociological’ or ‘sociophilosophical’ ideas. For instance, Benno Erdmann — the main target of both Frege’s and Husserl’s antipsychologistic criticism — proposed something of a consensus theory of truth. According to Erdmann, a judgement was true if and only if it was generally valid; and a judgement was generally valid if and only if all those forming that very judgement agreed about its content (Erdmann 1892, 272-75). To this Frege famously replied that “something can be true even if all take it to be false” (1893, xvi), and Husserl retorted by insisting that the truth lay with the few not with the many (1900; 1975 §40). Another eminently social thinker among those accused of psychologism was Wilhelm Wundt. Much of Wundt’s ethics, metaphysics and collective psychology would today qualify as social theory and social epistemology. Wundt argued, inter alia, for the existence of collective mental states; for the ontological and ethical priority of the collective over the individual; or the cognitive dependence of the individual on its linguistic community (see Kusch 1999, Ch. 5). At least some of the self-proclaimed antipsychologistic philosophers, especially the Neo-Kantians and the phenomenologists, can also be read as early contributors to sociological theorising. This is obvious enough in

142

MARTIN KUSCH

Husserl’s case; after all Husserl’s writings on intersubjectivity and the its constitution became the starting point of an influential school in sociology (Husserl 1931; 1960, 1954, 1973a, b, c, Schütz 1970, Berger and Luckmann 1966; 1979). Advocates of the sociology of knowledge had different views on the importance of psychology for philosophy or sociology. Wilhelm Jerusalem was one of the very few self-proclaimed advocates of psychologism. Jerusalem supported “the so-called psychologism, that is, the line of thought which gives a psychological justification for logical truths and thus wants to turn the latter into psychological truths” (1905, 78). Jerusalem did not make clear how exactly he conceived of the integration of psychological and sociological perspectives in his theory of knowledge but there can be no doubt that he wished to ‘naturalise’ epistemology and logic from both directions. Scheler too had a long-standing involvement with the psychologism issue. Scheler’s Habilitationsschrift of 1901 opposed psychologism as the view according to which “the specifically philosophical disciplines are all parts of psychology” (1901; 1971, 320). In subsequent years Scheler joined the phenomenological movement and became one of its most visible representatives. Scheler did not abandon antipsychologism once his interests turned towards the sociology of knowledge in the early 1920s. But the meaning of the very term ‘psychologism’ now underwent a change: It now referred to the attempt to explain the content and the development of science on the basis of the social-political interests of individual scientists (1925; 1982, 106-7). As mentioned above, Adler explicitly presented the sociological approach as a way to overcome the shortcomings of psychologism. At the same time, Adler was not altogether dismissive of psychologism. Indeed, he regretted the contemporary ‘fear of psychologism’ which in his view had led to an unfortunate separation of “the problem of validity from all links to the fact of experience” (1925; 1982a, 176). Like Adler, Mannheim was willing to go along with the antipsychologists in regarding psychologism as a mistake that needed to be overcome. For example, even though he felt that talking of eternal truths was highly misleading, he granted that such talk had been of “great value in the struggle against psychologism” (1931; 1936, 263). Psychologism was wrong, Mannheim thought, since it tried to explain meaning and intentionality as arising from psychological processes like association. This was hopeless since these latter processes were neither intentional nor meaningful (1931; 1936, 264-66). We have seen how the sociologists viewed psychologism. But how did the psychologists assess the sociology of knowledge? This question is of special interest today since in our time and age psychological and sociological explanations are often in direct competition. Unfortunately, in

PSYCHOLOGISM AND SOCIOLOGISM

143

the Germany of the 1920s and early 1930s, psychologists had largely withdrawn from epistemological concerns. To be sure, if Wilhelm Wundt had still been around, he would undoubtedly have responded to Scheler and Mannheim with a little book. Alas, Wundt died in 1920. Thus there was no criticism of the sociology of knowledge starting from psychological or psychologistic premises. A final issue concerning the links between the disputes over psychologism and sociologism is the role of phenomenology in both controversies. Phenomenology was very much involved in the dispute over psychologism: Husserl’s first booklength phenomenological treatise, the Logische Untersuchungen (1900; 1975), started with a long attack on psychologism. For many contemporaries, phenomenology became inseparable from opposition to both experimental psychology and psychologism. Phenomenology was also central to the debate over sociologism. Scheler was a self-proclaimed phenomenologist and one of the best-known philosophers of the Weimar era. Therefore, anyone who discussed Scheler’s sociology of knowledge was forced to comment on phenomenology more broadly. One idea that figured centrally in both Husserl’s antipsychologism and Scheler's sociology of knowledge was the ‘independence theory of truth’; that is, the idea that truths were eternal and existed independently of human knowers. This view of truth was important in Scheler’s sociology of knowledge: Scheler thought that social and biological factors merely determined which eternal truths became accessible at which historical moments. But this was all — truths themselves were neither constituted nor shaped by human consensus. As Scheler put it, sociological explanation concerned ‘the selection’ of truths, not their ‘meaningful content’ (1925; 1982, 106). Jerusalem and Mannheim saw things very differently. For them, the sociology of knowledge was incompatible with central aspects of Husserl’s phenomenology, especially with the independence theory of truth. I have already mentioned Jerusalem’s criticism above, but Mannheim’s assault on Husserl and Scheler is worth presenting in more detail. Mannheim did not agree with the way in which Husserl and Scheler distinguished between the ‘validity’ and the ‘genesis’ of judgements. For the Husserl of the Logische Untersuchungen and for Scheler, this distinction was sharp and absolute — after all, truths were independent of the human mind. Mannheim believed not only that the genesis of a judgement was visible in its very content, but also that speaking of a realm of ‘truth as such’ was of mere heuristic value. According to Mannheim, the content of every judgement was, as it were, indexed to a social situation and thus indexed to its genesis. Moreover, since the content of every judgement was indexed in this way, so too was its truth. Thus it made no sense to speak “of a sphere of

144

MARTIN KUSCH

validity in which the criteria of truth are independent of...the origins [of the judgement]” (1931; 1936, 244, 258). Mannheim also dismissed Husserl’s and the Neo-Kantians’ view according to which epistemology was somehow prior with respect to the sciences. New sciences, Mannheim insisted, ...grow out of the conditions of collective life and do not depend for their emergence upon the prior demonstration by a theory of knowledge that they are possible...The relationship is actually quite the reverse: ...The theory of knowledge takes over from the concrete conditions of knowledge of a period (and thereby of a society) not merely its ideal of what factual knowledge should be, but also the utopian construction of a sphere of ‘truth as such’...the utopian pattern of correctness, the idea of truth, arises out of the concrete modes of obtaining knowledge prevailing at a given time (1931; 1936, 261).

Despite this harsh criticism of Husserl and Scheler, Mannheim was willing to acknowledge the influence of phenomenology upon his work. In particular Mannheim praised Scheler’s and Heidegger’s attack on ‘modern intellectualism’ and individualism. Indeed, Mannheim went so far as to credit ‘the phenomenological school’ with the insight that “certain types of knowledge are relative to existence” (Seinsrelativität bestimmter Erkenntnisse) (1929; 1936, 331). Mannheim’s leaning towards, and praise for, Heidegger was particularly pronounced. He called Heidegger’s “struggle for an ontology...one of the most decisive achievements of contemporary philosophy” (1929; 1982, 434), and used two of Heidegger’s key concepts — ‘the public interpretation of Being’ (die öffentliche Auslegung des Seins) and ‘das Man’ — to characterise the aim of his own work: The sociology of knowledge sought to describe and explain the social struggle over “the public interpretation of Being”; and it provided a social account of the genesis of das Man (1929; 1982, 335). Other contributors to the sociologism dispute also referred to Husserl and Heidegger. Paul Eppstein praised Mannheim for his “combination of phenomenological vision with dialectical method” (1928; 1982a, 253; 1928; 1982b, 393). Marck (1929; 1982) saw clear parallels between Heidegger and various approaches to the sociology of knowledge: Both were allegedly based on ‘epistemological determinism’ and ‘existentialism’ (1929; 1982, 446). And to mention a final case, Horkheimer alluded to both Husserl and Heidegger in his defence of relativism. He noted that the accusation of relativism had at first only been directed against those who had tried to derive logical principles from psychological matters of fact; and that later the accusation had been extended to anyone who refused to accept ‘eternal truths’. Horkheimer suggested that the accusation was based upon an outdated ontology according to which truths were independent of the knowing subjects. In this context Horkheimer noted that phenomenology

PSYCHOLOGISM AND SOCIOLOGISM

145

itself had shifted away from this static conception and that Heidegger was to be credited with having initiated this shift (1930; 1982, 485). It also deserves to be mentioned that phenomenology — in both its Husserlian and its Heideggerian form — could also be used as a critical resource against Mannheim. Thus Hannah Arendt used Heidegger’s conception of authenticity — authenticity as withdrawal from das Man — to challenge Mannheim’s vision of authenticity based on group membership (1930; 1982, 527). And Grünwald’s book was in large part a very close translation of Husserl’s arguments against psychologism into a series of arguments against Mannheim’s alleged sociologism (1934; 1982, 753).

VII At least with respect to the psychologism dispute, subsequent generations of philosophers have not had many doubts on who won. Of course, Frege and Husserl did. The strength of the consensus can — amongst other things — be read from the fact that key antipsychologistic treatises of the psychologism dispute have become canonical texts of twentieth-century Western philosophy in general, and analytic philosophy in particular. Despite the strong consensus on who won the psychologism dispute, the accusation of psychologism has been with us all through this century. And it has not just been directed at people like Quine who first presented his project of a naturalised epistemology under the title “Epistemology Naturalised: Or, the Case for Psychologism” (Willard 1984, 161; Quine 1969; 1985). It has also been aimed at Armstrong, Ayer, Carnap, Dummett, Geach, Goodman, Kuhn, McDowell, Neurath, Popper, Russell, Sellars, and Wittgenstein. Philosophers hunting for psychologistic tendencies in others have included Chisholm, Dummett, Katz, Kitcher, McDowell, Musgrave, Popper, Sellars, and Sober (Kusch 1995, 7). The afterlife of the sociologism dispute has been different. I shall again focus on the Anglo-American world. None of the major texts arguing for or against the sociology of knowledge has made it into the canon of twentiethcentury philosophy. Nevertheless, the sociologism debate has influenced twentieth-century English-speaking philosophy in some indirect ways. One important channel of transmission has been the link between Ludwik Fleck’s Entstehung und Entwicklung einer wissenschaftlichen Tatsache (1935; 1980) and Thomas Kuhn's The Structure of Scientific Revolutions (1962; 1996). Fleck cited several of the early sociologists of knowledge — with the notable exception of Mannheim — and he also seems to have been influenced by some of the protosociological ideas of psychologists like Wundt — whose logic Fleck quoted (1935; 1980, 42). Clearly Wundtian for example was Fleck’s suggestion according to which a group qua collective of individuals could be more stable and coherent than a

146

MARTIN KUSCH

personality qua collective of contradictory drives (Fleck 1935; 1980, 60; e.g. Wundt 1897). The same holds for Fleck’s insistence that the thought collective was more than the simple sum of thinking individuals (1935; 1980, 56). Fleck also applauded Jerusalem’s idea that the frequent invoking of an idea led to its ‘condensation’ (1935; 1980, 53). At the same time, Fleck’s sociology of knowledge went beyond Mannheim in dropping the distinction between two forms of knowledge, only one of which is situationsgebunden. For Fleck the sociologist did not stop at the gate of the natural sciences. Kuhn acknowledged in the preface to Structure that Fleck’s book had anticipated many of his own ideas, and that Fleck’s book that taught him about the crucial role of the scientific community (1962; 1996, viii-ix). Perhaps the most important channel for antisociologism has been the work of Karl Popper. Popper was a Ph.D. student in Vienna when the debate over sociologism was in full swing. Popper was familiar with Mannheim’s work, and he was very critical of it. Indeed, a whole chapter of Popper’s The Open Society and Its Enemies was devoted to ‘The Sociology of Knowledge’ (1945; 1974, Ch. 23), and Popper returned to the topic in The Poverty of Historicism (1957; 1960, 155-56). Popper’s criticism of Mannheim was similar to those presented in the original debate. For example, Popper accused Mannheim of being hostile to rationalism; of being insufficiently reflexive in not applying the Seinsgebundenheit thesis to the intelligentsia; of locating scientific objectivity in individuals rather than in the intersubjective method; of being an incoherent relativist; and of failing to see that true assertions are true eternally (1945; 1974, 213, 216, 221). At the same time, however, Popper also passed on — or invented anew — something of the anti-individualism so characteristic of Jerusalem, Scheler, Adler and Mannheim. Popper contended that scientific objectivity was the outcome of the “friendly-hostile co-operation” of many scientists, and that Robinson Crusoe on his own could not produce objective knowledge (1945; 1974, 219-20). Popper thus expressed ideas from both sides of the sociologism debate of the twenties, and in so doing, passed on a somewhat ambiguous message. Perhaps Popper felt this tension himself: His anti-individualism pushed him towards collectivism, but his antisociologism did not allow him to adopt collectivism. The outcome of this tension was Popper’s crypto-Platonist notion of the ‘third world’ (cf. Bloor 1974). Kuhn’s and Popper’s links to the sociology-of-knowledge debate of the 1920s puts their own ‘sociologism dispute’ of the 1960s in a new light. In some respects, and mutatis mutandis, the debate between Kuhn, Popper and Lakatos continued the debate over the sociology of knowledge that Hitler’s rise to power had interrupted. Again the dispute was over the question whether the sociology or social history of scientific knowledge was of

PSYCHOLOGISM AND SOCIOLOGISM

147

significance for epistemology and the philosophy of science; and again the charges of relativism, psychologism and sociologism were key features of the debate (Kuhn 1970, Popper 1970, Lakatos 1970, Watkins 1970).

VIII The debate between Kuhn and Popper brings us back to the sociology of knowledge — and this not only because recent sociology of knowledge learnt a lot from Kuhn and often defined itself in opposition to Popper. The even stronger link is that one of the first contributions of the recent ‘strong programme’ in the sociology of scientific knowledge was an interpretation of the Kuhn-Popper debate based upon Mannheim’s sociology of knowledge (Bloor 1976; 1991, 55-83). The historical story of this paper might be taken to indicate that this interpretation was not anachronistic. In this paper I have tried to give an overview over the early-twentiethcentury disputes over the relations between psychology, sociology, and philosophy. These disputes did not settle what these relations should be — albeit that analytic philosophers, for some decades at least, were able to convince themselves that psychology had nothing to contribute to epistemology, and that the sociology of knowledge provided a no serious challenge. Today we see these old certainties destroyed. Giving them up might be easier once we realise of how recent origin they are. Department of History and Philosophy of Science University of Cambridge, UK REFERENCES

Adler, M. (1925; 1982a), “Soziologie und Erkenntniskritik”, in Meja and Stehr (1982), 158-91. (First published in 1925.) Adler, M. (1925; 1982b), “Wissenschaft und soziale Struktur”, in Meja and Stehr (1982), 128-57. (First published in 1925.) Arendt, H. (1930; 1982), “Philosophie und Soziologie”, in Meja and Stehr (1982), 515-31. (First published in 1930.) Barth, H. (1934; 1982), “Geist und Gesellschaft”, in Meja and Stehr (1982), 663-78. (First published in 1934.) Beneke, Friedrich Eduard. Lehrbuch der Logik als Kunstlehre des Denkens. Berlin/ Posen/Bromberg, 1832. Beneke, Friedrich Eduard. System der Logik als Kunstlehre des Denkens. Erster Theil. Berlin: Ferdinand Dümmler, 1842.

148

MARTIN KUSCH

Berger, P. L. and T. Luckmann. The Social Construction of Reality: A Treatise in the Sociology of Knowledge, Harmondsworth: Penguin, 1966; 1979. Bloor, D. “Popper’s Mystification of Objective Knowledge”, Science Studies 4, 65-76, 1974. Bloor, D. “Wittgenstein and Mannheim on the Sociology of Mathematics”, Studies in the History and Philosophy of Science, 4, 1973, 173-91. Bloor, D. Knowledge and Social Imagery, 2nd ed., Chicago and London: Chicago University Press, 1991. (First published in 1976.) Bochen,’ ski, Innocent Maria. Formale Logik. Feiburg i. B. und München: Alber, 1956. Bolzano 1837. Wissenschaftslehre. Boole, George. An Investigation of the Laws of Thought, on which are Founded the Mathematical Theories of Logic and Probabilities. London, 1854. Boole, George. The Mathematical Analysis of Logic. Being an Essay Towards a Calculus of Deductive Reasoning. Cambridge/London, 1847. Born, Friedrich Gottlob. Versuch über die ursprünglichen Grundlagen des Menschlichen Denkens. Leipzig: Barth, 1791. Reprinted Brussels (Culture et Civilisation) 1969. Brockhaus, R. “Realism and Psychologism in 19th Century Logic”, Philosophy and Phenomenological Research, 51, 1991, 493-523. Calker, Johann Friedrich Augustran. Denklehre: oder Logik und Dialektik. Bonn: Eduard Weber, 1822. Carl, Wolfgang. Frege’s Theory of Sense and Reference. Cambridge: Cambridge University Press, 1994. Cassirer, E. Das Erkenntnisproblem in der Philosophie und Wissenschaft der neueren Zeit, vol. III, Berlin: Cassirer, 1920. Cohen, H. Logik der reinen Erkenntnis, Berlin: Cassirer, 1902. Curtius, E. R. “Soziologie — und ihre Grenzen”, in Meja and Stehr (1982), 417-26. (First published in 1929.) Dunkmann, K. “Die soziologische Begründung der Wissenschaft”, in Meja and Stehr (1982), 192-210. (First published in 1927.) Elias, N. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 388-90. (First published in 1929.) Elsenhans, T. “Das Verhältnis der Logik zur Psychologie”, Zeitschrift für Philosophie und philosophische Kritik, 1897, 109, 195-212. Eppstein, P. “Die Fragestellung nach der Wirklichkeit im historischen Materialismus”, in Meja and Stehr (1982), 251-324. (First published in 1928.) Eppstein, P. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 393-95. (First published in 1929.)

PSYCHOLOGISM AND SOCIOLOGISM

149

Erdmann, B. Logische Elementarlehre, Halle: Niemeyer, 1892. Erdmann, Benno. Logik, 1. Band: Logische Elementarlehre. Max Niemeyer, Halle, 1892. 2., erw. Aufl., Berlin 1907. Fleck, L. Entstehung und Entwicklung einer wissenschaftlichen Tatsache: Einführung in die Lehre vom Denkstil und Denkkollektiv, ed. by L. Schäfer and T. Schnelle, Frankfurt am Main: Suhrkamp. (First published in 1935.) Frege, G. Grundlagen der Arithmetik, Breslau: Marcus, 1934. (First published in 1884.) Frege, Gottlob. Begriffsschrift, eine der mathematischen nachgebildete Formelsprache des reinen Denkens. Halle a. S.: L. Nebert, 1879. Frege, Gottlob. Die Grundgesetze der Arithmetik. Vol. I, Jena: H. Pohle, 1893, Vol. II, Jena: H. Pohle, 1903. Fries, Jakob Friedrich, System der Logik (1837). 3. Aufl., Heidelberg: Christian Friedrich Winter, 1811 (1). Frisby, D. The Alienated Mind: The Sociology of Knowledge in Germany 1918-33, London: Heineman, 1983. Frischeisen-Köhler, M. “Philosophie und Psychologie”, Die Geisteswissenschaften, 1913, 1, 371-73; 400-3. George, R. “Psychologism in Logic: Bacon to Bolzano”, Philosophy and Rhetoric, 30, 1997, 213-42. Grünwald, E. “Systematische Analyse der wissenssoziologischen Theorien”, in Meja and Stehr (1982), 681-747. (First published in 1934.) Grünwald, E. “Wissenssoziologie und Erkenntniskritik”, in Meja and Stehr (1982), 748-55. (First published in 1934.) Haaparanta, Leila. “Frege and His German Contemporaries on Alethic Modalities”, in: Knuuttila, S. (ed.), Modern Modalities. Dordrecht/Boston/London: Kluwer, 1988, 239-74. Heidegger, M. Sein und Zeit, Tübingen: Niemeyer, 1927. Heim, K. Psychologismus oder Antipsychologismus? Entwurf einer erkenntnistheoretischen Fundamentierung der modernen Energetik, Berlin: Schwetschke, 1902. Hempel, Carl Gustav. Aspects of Scientific Explanation and Other Essays in the Philosophy of Science. New York/London: The Free Press, 1956. Herbart, Johann Friedrich. Lehrbuch der Psychologie, in: Johann Friedrich Herbarts Sämmtliche Werke, hrsg. von G. Hartenstein, 5. Band., Leipzig: Leopold Voss (1829) 1851, 1-190. Herbart, Johann Friedrich. Lehrbuch zur Einleitung in die Philosophie. A. W. Unzer, Königsberg, 1813; reprinted: accord, to the 4th ed. Hamburg: Felix Meiner, 1993. Herbart, Johann Friedrich. Sämmtliche Werke, 12. Bände; Herausgegeben von G. Hartenstein. Leipzig: Leopold Voss, 1850-1852.

150

MARTIN KUSCH

Heymans, G. Die Gesetze und Elemente des wissenschaftlichen Denkens. Ein Lehrbuch der Erkenntnistheorie in Grundzügen, vol. 1, Leipzig: Harrassowitz, 1890. Heymans, G. Die Gesetze und Elemente des wissenschaftlichen Denkens. Ein Lehrbuch der Erkenntnistheorie in Grundzügen, vol. 2, Leipzig: Harrassowitz, 1894. Heymans, G. Die Gesetze und Elemente des wissenschaftlichen Denkens: Ein Lehrbuch der Erkenntnistheorie in Grundzügen, vol. 1, 2nd, rev. ed., Leipzig: Barth, 1905. Horkheimer, M. “Ein neuer Ideologiebegriff”, in Meja and Stehr (1982), 474-96. (First published in 1930.) Husserl, E. “Philosophie der Arithmetik: Logische und psychologische Untersuchungen”, in E. Husserl, Philosophie der Arithmetik, ed. by L. Eley, Husserliana XII, The Hague: Nijhoff, 1970; 5-283. (First published in 1891.) Husserl, E. Cartesianische Mediationen und Pariser Vorträge, ed. by S. Strasser, Husserliana I, The Hague: Nijhoff, 1960. (First published in 1931.) Husserl, E. Die Krisis der europäischen Wissenschaften und die transzendentale Phänomenologie, Husserliana VI, ed. by W. Biemel, The Hague: Nijhoff, 1954. Husserl, E. Logische Untersuchungen. Erster Band: Prolegomena zur reinen Logik, ed. by E. Holenstein, Husserliana XVIII, The Hague: Nijhoff, 1975. (First published in 1900.) Husserl, E. Zur Phänomenologie der Intersubjektivität: Texte aus dem Nachlass, Dritter Teil: 1929-1935, ed. by I. Kern, Husserliana XIII, The Hague: Nijhoff, 1973c. Husserl, E. Zur Phänomenologie der Intersubjektivität: Texte aus dem Nachlass, Erster Teil: 1905-1920, ed. by I. Kern, Husserliana XIII, The Hague: Nijhoff, 1973a. Husserl, E. Zur Phänomenologie der Intersubjektivität: Texte aus dem Nachlass, Zweiter Teil: 1921-1928, ed. by I. Kern, Husserliana XIII, The Hague: Nijhoff, 1973b. Husserl, Edmund. Logische Untersuchungen, Bd. 1, C. E. Pfeffer (Robert Stricker), Halle a. S., 1900. Jerusalem, W. “Die soziologische Bedingtheit des Denkens und der Denkformen”, in Meja and Stehr (1982), 27-56. (First published in 1924.) Jerusalem, W. “Psychologen und Philosophen”, Die Zukunft, 88, 1914, 8597. Jerusalem, W. Der kritische Idealismus und die reine Logik: Ein Ruf im Streite, Wien und Leipzig: Braumüller, 1905.

PSYCHOLOGISM AND SOCIOLOGISM

151

Jonas, H. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 382-93. (First published in 1929.) Kettler, D. and V. Meja Karl Mannheim and the Crisis of Liberalism: The Secret of these New Times, New Brunswick and London: Transaction Publishers, 1995. Kiesewetter, J.G.C.C. Grundriß einer allgemeinen Logik nach Kantischen Grundsätzen. 2 Vols (pure and applied logic). Leipzig: Köchly, 1824/25. Reprinted Bussels (Culture et Civilisation) 1973. First ed. 1793. Kneale, William, and Kneale, Martha. The Development of Logic. Oxford: Oxford University Press, 1962. Kuhn, T. “Reflections in My Critics”, in Lakatos and Musgrave (1970), 23178. Kuhn, T. The Structure of Scientific Revolutions, 3rd ed., Chicago: Chicago University Press, 1996. (First published in 1962.) Kusch, M. Psychological Knowledge: A Social History and Philosophy, London: Routledge, 1999. Kusch, M. Psychologism. London/New York: Routledge, 1995. Kusch, M. Psychologism: A Case Study in the Sociology of Philosophical Knowledge, London: Routledge, 1995. Lakatos, I. “Falsification and the Methodology of Scientific Research Programmes”, in Lakatos and Musgrave (1970), 91-196. Lakatos, I. and A. Musgrave (eds.) Criticism and the Growth of Knowledge, Cambridge: Cambridge University Press, 1970. Lapp, A. Die Wahrheit: Ein erkenntnistheoretischer Versuch orientiert an Rickert, Husserl und an Vaihinger's 'Philosophie des Als-Ob', Stuttgart: Spemann, 1913. Leibniz, Gottfried Wilhelm. Briefwechsel zwischen Leibniz und Wolff, ed. C.J. Gerhard. Halle, 1860. Lewalter, E. “Wissenssoziologie und Marxismus. Eine Auseinandersetzung mit Karl Mannheims Ideologie und Utopie von marxistischer Position aus”, in Meja and Stehr (1982), 551-83. (First published in 1930.) Lipps, T. Grundzüge der Logik, Hamburg and Leipzig: Verlag von Leopold Voss, 1893. Lipps, Theodor. Grundzüge der Logik. Hamburg: Voss, 1893. Locke, John. An Essay Concerning Human Understanding. Ed. Alexander Campbell Fraser, 2 Vols. Oxford: Oxford University Press 1894. Reprinted New York: Dover, 1959. Lotze, Hermann. Logic in Three Books: of Thought, of Investigation and of Knowledge. Translated by several and edited by Bernard Bosanquet. Oxford: The Clarendon Press, 1888. Reprinted New York: Garlandm 1980. First German edition 1874. Lotze, Rudolf Herrmann, 1843, Logik. Weidmann’sche Buchhandlung, Leipzig, 1843.

152

MARTIN KUSCH

Lotze, Rudolph Hermann, 1874, Logik. Drei Bücher vom Denken, vom Untersuchen und vom Erkennen, S. Hirzel, Leipzig (= Lotze, System der Philosophie, T. 1). New edition: Lotze, Rudolph Hermann, 1989, Logik. Erstes Buch. Vom Denken (Reine Logik). Philosophische Bibliothek Band 421, Meiner, Hamburg. Lotze, Rudolph Hermann, 1989, Logik. Drittes Buch. Vom Erkennen (Methodologie). Philosophische Bibliothek Band 408, Meiner, Hamburg. MacColl, Hugh. Symbolic Logic and Its Applications. London/New York, Bombay: Longmans, 1906. Mannheim, K. “Die Bedeutung der Konkurrenz im Gebiete des Geistigen”, in Meja and Stehr (1982), 325-70. (First published in 1929.) Mannheim, K. “Ideologische und soziologische Interpretation der geistigen Gebilde”, in Meja and Stehr (1982), 213-31. (First published in 1922.) Mannheim, K. “Ideology and Utopia; The Prospects of Scientific Politics; The Utopian Mentality”, in Mannheim (1936a), 49-236. (First published in 1929.) Mannheim, K. “The Sociology of Knowledge”, in Mannheim (1936a), 23780. Mannheim, K. “Zur Problematik der Soziologie in Deutschland”, in Meja and Stehr (1982), 427-37. (First published in 1929.) Mannheim, K. (1936b), “Preliminary Approach to the Problem”, in Mannheim (1936a), 1-48. Mannheim, K. Ideology and Utopia, London: Routledge and Kegan Paul, 1936a. Mannheim, K. Konservatismus: Ein Beitrag zur Soziologie des Wissens, ed. by D, Kettler, V. Meja, and N. Stehr, Frankfurt am Main: Suhrkamp, 1984. (First published in 1925.) Marck, S. “Zum Problem des ‘seinsverbundenen Denkens‘”, in Meja and Stehr (1982), 438-50. (First published in 1929.) Marcuse, H. “Zur Wahrheitsproblematik der soziologischen Methode”, in Meja and Stehr (1982), 459-73. (First published in 1929.) Meier, George Friedrich. Auszug aus der Vernunftlehre. Halle: Gebauer, 1752; reprinted in Kant, Akad 16. Meja, V. and N. Stehr (eds.) Der Streit um die Wissenssoziologie, 2 vols., Frankfurt am Main: Suhrkamp, 1982. [Vol. 1, 1-413; Vol. 2, 414-975.] Meusel, A. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 387-88. (First published in 1929.) Natorp, P. “Ueber objective und subjective Begründung der Erkenntnis I”, Philosophische Monatshefte, 23, 1887, 257-86. Natorp, P. “Zur Frage der logischen Methode. Mit Beziehung auf Edm. Husserls ‘Prolegomena zur reinen Logik’”, Kantstudien, 6, 1901, 27083.

PSYCHOLOGISM AND SOCIOLOGISM

153

Neurath, O. “Bürgerlicher Marxismus”, in Meja and Stehr (1982), 584-93. (First published in 1930.) Platner, Ernst. Philosophische Aphorismen. Leipzig: Schwicker, 1793. Reprinted Brussels (Culture et Civilisation) 1970. Plessner, H. “Abwandlungen des Ideologiegedankens”, in Meja and Stehr (1982), 637-62. (First published in 1931-32.) Popper, K. “Normal Science and its Dangers”, in Lakatos and Musgrave (1970), 51-8. Popper, K. The Open Society and Its Enemies, Vol. II, The High Tide of Prophecy: Hegel, Marx, and the Aftermath, 4th rev. ed., London: Routledge and Kegan Paul, 1962. (First published in 1945.) Popper, K. The Poverty of Historicism, London: Ark, 1986. (First published 1957.) Quine, W. “Epistemology Naturalized”, in H. Kornblith, Naturalizing Epistemology, Cambridge: Bradford Books, 1985, 15-29. (First published in 1969.) Rath, M. Der Psychologismusstreit in der deutschen Philosophie. Freiburg/ München: Alber, 1994. Rehmke, J. Logik oder Philosophie als Wissenslehre, Leipzig: Quelle und Meyer, 1918. Reid, Thomas. “Analysis of Aristotle’s Logic” published as an appendix to Essays on the Intellectual Powers of Man, ed. G.N. Wright. London: Tegg, 1843, 553-600. First published as an appendix to Home, Henry, Sketches of the History of Man. Edinborough, 1774. Reimarus, H.S. (1776): Vernunftlehre. Hamburg (Bohn). Rickert, H. Der Gegenstand der Erkenntnis: Ein Beitrag zum Problem der philosophischen Transcendenz, 2nd, rev. ed., Tübingen: Mohr, 1904. Rickert, H. Philosophie des Lebens, Tübingen: Mohr, 1920. Scheler, M. “Die positivistische Geschichtsphilosophie des Wissens und die Aufgaben einer Soziologie der Erkenntnis”, in Meja and Stehr (1982), 57-67. (First published in 1921.) Scheler, M. “Die transzendentale und die psychologische Methode. Eine grundsätzliche Erörterung zur philosophischen Methodik”, in M. Scheler, Frühe Schriften, ed. by M. Scheler and M. S. Frings, Bern und München: Francke, 1901; 1971, 197-336. Scheler, M. “Wissenschaft und soziale Struktur”, in Meja and Stehr (1982), 68-127. (First published in 1925.) Schelting, A. v. “Die Grenzen der Soziologie des Wissens”, in Meja and Stehr (1982), 756-890. (First published in 1934.) Schlick, M. “Das Wesen der Wahrheit nach der modernen Logik”, Vierteljahrsschrift für wissenschaftliche Philosophie und Soziologie, 34, 1910, 386-477. Schlick, M. Allgemeine Erkenntnislehre, Berlin: Springer, 1918.

154

MARTIN KUSCH

Schröder, Ernst. Abrifl der Algebra der Logik. Bearb. von E. Müller. Teil I: Elementarlehre, Leipzig/Berlin 1909, Teil II: Aussagentheorie, Funktionen, Gleichungen und Ungleichungen, Leipzig, 1910. Schröder, Ernst. Der Operationskreis der Logik. Leipzig, 1877. Schütz, A. On Phenomenology and Social Relations: Selected Writings, ed. by H. R. Wagner, Chicago: Chicago University Press, 1970. Sigwart, C. Logik, 4th, rev. ed., edited by H. Maier, Tübingen: Mohr, 1921. Sigwart, Christoph. Beiträge zur Lehre vom hypothetischen Urteil. Tübingen, 1871. Sigwart, Christoph. Logik, Vol. I: Die Lehre vom Urteil, vom Begriff und vom Schluss, 1873/1878; Vol. II: Die Methodenlehre, Heinrich Laupp, Tübingen, 2. Aufl. Freiburg 1889-1893, 4., durchgesehene Auflage, J.C.B. Mohr (Paul Siebeck), Tübingen, 1911. Singer, K. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 382-83. (First published in 1929.) Sombart, W. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 376-80. (First published in 1929.) Speier, H. “Soziologie oder Ideologie? Bemerkungen zur Soziologie der Intelligenz”, in Meja and Stehr (1982), 532-50. (First published in 1930.) Spengler, O. Der Untergang des Abendlandes: Umrisse einer Morphologie der Weltgeschichte. Erster Band. Gestalt und Wirklichkeit, München: Beck, 1918. Spranger, E. “Ideologie und Wissenschaft”, in Meja and Stehr (1982), 63436. (First published in 1930.) Stern, G. “Über die sog. ‘Seinsverbundenheit’ des Bewusstseins”, in Meja and Stehr (1982), 497-514. (First published in 1930.) Tillich, P. “Ideologie und Utopie”, in Meja and Stehr (1982), 451-58. (First published in 1929.) Watkins, J. W. N. “Against ‘Normal Science’”, in Lakatos and Musgrave (1970), 25-38. Weber, A. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 371-76. (First published in 1929.) Wilbrandt, R. Comment on Mannheim (1982; 1929), in “Diskussion über ‘Die Konkurrenz’”, in Meja and Stehr (1982), 371-401, here 380-82. (First published in 1929.) Willard, D. Logic and the Objectivity of Knowledge: A Study in Husserl's Early Philosophy, Athens: Ohio University Press, 1984. Windelband, W. “Kritische oder genetische Methode?”, in Präludien, Freiburg i. Br.: Mohr, 1884, 247-79.

PSYCHOLOGISM AND SOCIOLOGISM

155

Windelband, W. Die Philosophie im deutschen Geistesleben des XIX. Jahrhunderts: Fünf Vorlesungen, Tübingen: Mohr, 1909. Wittfogel, K. A. “Wissenschaft und Gesellschaft”, in Meja and Stehr (1982), 594-615. (First published in 1931.) Wundt, W. Logik: Eine Untersuchung der Prinzipien der Erkenntnis und der Methoden wissenschaftlicher Forschung, 2 vols., Stuttgart: Enke, 1880; 1883. Wundt, W. System der Philosophie, 2nd rev. ed., Leipzig: Engelmann, 1897.

This page intentionally left blank

VINCENT COLAPIETRO

THE SPACE OF SIGNS: C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

1. Introduction Despite the efforts of his father, the mathematician Benjamin Peirce, to dissuade him (Brent, 78) from doing so, Charles Peirce devoted himself to logical investigations with a singular passion. Even in his later life, he identified himself most often as a logician.1 From the first moment he spoke publicly on “The Logic of Science”2 to the end of his life, he was strenuously opposed to all attempts to ground logic in psychology (W 1). His opposition to psychologism could not have been more thoroughgoing. This is true even though in looking back upon his own work he in later years discerned some of his own writings to be tainted by this doctrine. For example, in his 1903 lectures on pragmatism, he implicitly recalled the essay in which he first put forth his pragmatic maxim,3 though without naming it. Near the center of Peirce’s pragmatism, in his mature as well as youthful formulation, was a conception of belief borrowed from the Scottish psychologist Alexander Bain (5.12).4 In 1903, he noted: “ belief consists mainly in being deliberately prepared to adopt the formula believed in as the guide to action” (5.27). In 1878, he wrote, more simply: “The essence of belief is the establishment of a habit” (5.398). Doubt indicates the disruption of belief and, as such, instigates a struggle to recover the settled state of an efficacious habit. Peirce identified this struggle as the process of inquiry (see, e.g., 5.374). This doubt-belief theory of inquiry was the context in which the pragmatic maxim was first proposed and afterwards always defended. But, when Peirce after 18985 returned to the topic of pragmatism, he was especially sensitive to the transformation of his own logical doctrine into a psychological teaching.6 Hence, in the opening lecture of the 1903 series on this topic, he stressed:

157 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 157-179. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

158

VINCENT COLAPIETRO My original article carried this [point regarding belief] back to a psychological principle. The conception of truth, according to me, was developed out of an original impulse to act consistently, to have a definite intention. But in the first place, this was not very clearly made out; and in the second, I do not think it satisfactory to reduce such fundamental things to facts of psychology. For man could alter his nature, or his environment would alter it after it if he did not voluntarily do so, if the impulse were not what was advantageous or fitting. Why has evolution made man’s mind to be so constructed? That is the question we must nowadays ask, and all attempts to ground the fundamentals of logic on psychology are seen to be essentially shallow. (5.28; emphases added).

Yet, insofar as Peirce remained committed to understanding the logic of inquiry in terms of belief, doubt, and allied notions, it must seem to contemporary readers that he was inextricably ensnared in the psychologism he so emphatically denounced. For such notions of what is commonly called today “folk psychology” as those of belief and doubt are clearly psychological. As we shall see, Peirce did not himself think that such notions were psychological in a sense vitiating his critique of psychologism. He defined “logic proper” as “the critic of arguments, the pronouncing them to be good or bad” (5.108). But logic in this sense was, in his judgment, “a particular species of morality,” since it was at bottom the identification of norms to which deliberative agents are compelled to adhere, insofar as they are faithful to their defining commitments (above all, their espousal of the ideal of truth and, thereby, to the most reliable methods for discovering unknown truths or for testing epistemic claims). In general, the normative sciences (logic, ethics, and esthetics) are inseparable from a commonsensical conception of deliberative agency.7 We can mind what we say and infer. The critic of argument ultimately makes sense only in reference to our capacity to exert a critical, decisive measure of control over what we say and how we reason. Logic as the critic of argument, thus, implies a critic in the sense of a deliberative agent self-consciously assuming a critical stance toward the use of signs as instruments of inquiry (Savan, 63). Peirce confessed he was quite aware such a conception of logic appeared to be psychologistic. Yet he took this appearance to be deceiving. There is a sense in which logic, fully conceived “rests on certain facts of experience among which are facts about men” and women (5.110). But these facts are very far removed from those discovered by experimental psychology or any other specialized study of the distinctive features of the human mind or closely allied forms of mentality. “Everyday experience, such as presses in upon every man, at every hour of his life, is open to no other doubt than that it may not have been correctly formulated in general terms. This must be the main source of what little matter of fact logic has occasion to assert” (2.75). Peirce stressed “the experiential element in logic is all but nil” (2.65); but note he does not say it is completely absent. Logic

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

159

“is an observational science in some sense; every science is that. Even mathematics observes its diagrams” and the results of experiments (or alterations) of the diagrams to which it directs its critical attention (2.65).8 What our observations of diagrams exhibiting the forms of inference allow us to discern is an important species of objective relations pertaining to the evidential force of each inferential pattern. There is at the center of logic the business of identifying, classifying, and testing these patterns of inference. But, for Peirce at least, this critic of arguments is, in the end, but a part of a more inclusive undertaking, one concerned with the use of signs in their myriad forms, precisely in their role as instruments of objective inquiry. Hence, he was led to claim: “Logic, in its general sense, is...only another name for semiotic the quasi-necessary, or formal, doctrine of signs” (2.227). Peirce’s conception of logic (or one of its prinicpal parts) as Critic was rooted in his early crucial encounter with Kant. This encounter helps us appreciate important features of the context in which Peirce’s critique of psychologism took shape. Yet, in time, this critique took shape as part of Peirce’s lifelong efforts to institute a truly general theory of signs, though one focused principally upon their heuristic function.9 Closely linked to this was Peirce’s vision of nature as a space of signs in which humans live and move and have their being, precisely as human. 10 The space of signs is akin to what Wildred Sellars identified as the space of reasons, but it is by no means simply identifiable with the space of reasons. Accordingly, the remainder of this paper is divided into three sections. The first deals with the context of Peirce’s critique of psychologism, the second with this critique itself, and the third with the space of signs. The immediate context of Peirce’s critique might be identified as his lifelong endeavor to offer a normative account of objective inquiry. His own endeavor was deeply indebted to certain Immanuel Kant’s critical project. The main thrust of Peirce’s critique is inseparable from his inclusive conception of logic. Hence, my discussion of his critique will encompass a consideration of this conception. Finally, the centrality of his antipsychologistic conception of a sign is in the final section underscored. Moreover, several important implications of this conception are noted.

2. The Critic of Experimental Reason Charles Peirce came to philosophy from a careful reading of Immanuel Kant’s first Critique (CN III, 94).11 He supposed that the title of this work is better translated as Critic of Pure Reason (see, e.g., CP 2.205-206; CN III, 95), in part because critique suggests something altogether too literary12 and in part because critic more clearly evokes one of the most important metaphors structuring Kant’s critical philosophy. The image of a tribunal or

160

VINCENT COLAPIETRO

judge before which we stand, and by whom our actions are appraised, is arguably as integral to Peirce’s as to Kant’s own (A xi-xii). In any event, a distinction linked to this image and crucial for a proper understanding of Kant’s critical undertaking is a distinction central to Peirce’s own conception of logic. It is, for both thinkers, imperative to distinguish between quaestio facti and quaestio juris; moreover, it is necessary for the logician and the critic of knowledge to identify properly the kind of question with which s/he is mainly concerned. How humans de facto think matters little, if at all, in this connection; but how humans ought to think is of the utmost importance. The de jure question of what constitutes a valid inference is here sharply separated from the de facto question of what occurs during the physiological, psychological, and other processes by which the business of inference is seemingly (or even actually) accomplished. Peirce’s early encounter with Kant’s text left a lasting impression: the normative character of our logical investigations cannot be gainsaid. Nor can it be reduced to, or derived from, merely contingent facts about human beings or other rational agents. Indeed, the character of logic as an irreducibly normative discourse needs to be safeguarded, especially since very influential philosophers, mostly in the empiricist tradition, have in effect reduced the de jure question of validity to the de facto question of how our minds tend to associate ideas with one another.13 In particular, psychological facts do not provide a justification for logical norms. While other forms of reductionism are not only possible but also exemplified in the history of thought (e.g., the attempt to reduce logical norms to anthropological or linguistic facts), the form with which Peirce was most directly concerned involved making logic a part of, and thereby subordinate to, psychology. In his review of volume II of James Mark Baldwin’s Dictionary of Philosophy and Psychology, Peirce played the role of the prophet in the archaic sense of this term, the prophet being in this sense not one who is able to foretell the future but one capable of reading the signs of the present. Among the signs of his own time, Peirce discerned, as characteristic of the philosophy in the first decade of the twentieth century, “the disposition to make psychology the key to philosophy — categories, Êsthetics, ethics, logic, and metaphysics” (CN III, 128). He is quick to point out that this is, to some extent, a characteristic feature of modern thought; yet, with the relatively recent rise of experimental psychology, it has become a more pronounced tendency: “Something of this has existed since Descartes; but since 1863 every student of philosophy, even though he be one of those who consider the present psychological tendency excessive, has placed a new and higher estimate than before upon the scientific value of psychology” (ibid.). It is helpful to recall that Psychologismus was first used in the first half of the nineteenth century by thinkers explicitly opposed to the Hegelian exaltation of a purportedly objective logic in which the indispensable

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

161

categories for knowing anything whatsoever are dialectically derived. For example, Friedrich Eduard Beneke (1798-1854) in Die Philosophie in ihrem Verhältnis zur Erfahrung (Berlin, 1833) argued: “With all of the concepts of the philosophical disciplines [including logic and metaphysics], only what is formed in the human soul according to the laws of its development can be thought; if these laws are understood with certainty and clarity, then a certain and clear knowledge of those disciplines is likewise achieved” (xv). Psychology is, in effect, prima philosophia: it is the discipline from which all of the other philosophical disciplines are derived.14 In the second half of the nineteenth century and the opening years of the twentieth, this tendency, at least according to Peirce, gathered momentum and came to be nothing less than one of the defining features of philosophical thought. He was convinced, however, that this tendency was as disastrous as it was widespread. Beneke was of course a minor figure in the history of philosophical thought, especially in connection with Anglo-American philosophy, however prescient he and other mostly forgotten authors (e.g., Jakob Friedrich Fries) were in their advocacy of Psychologismus. For Husserl as well as Peirce, it was a work by an English author from the second half of the nineteenth century, not one by a German writer from the first, that provided him with a target at which to shoot. For both of these opponents of psychologism, the most important historical representative of the opposite tendency (that of appealing to psychological facts as a warrant for our logical laws) was John Stuart Mill (see Dougherty, 87). They saw him as a part of a tradition and, in crucial respects, as the distillation of the principal methodological commitments of the British empiricist tradition and, closely allied with this, British associationist psychology (e.g., David Hartley, Thomas Brown, and of course James Mill as well as his son John Stuart). It is J.S. Mill [in Chapter XXI of his Examination of Sir William Hamilton’s Philosophy (1865)] who insists that how we ought to think can be ascertained in no other way than by reflection upon those psychological laws which teach us how we must needs think. But here we have to distinguish the case in which the compulsion attaches to that subconscious thought over which we have no control, and the case in which it attaches to conscious reasoning. (CP 2.47)

Logical compulsion needs to be sharply distinguished from other forms of compulsion. It pertains only to thought over which we can exercise a measure of control, thus only to conscious (and not subconscious) thought. When Mill or anyone else identifies a compulsion to which conscious thought might be subjected, “what he no doubt has in mind is, that a person ought to think in the way in which he would be compelled to think, if he duly reflected, and made his thoughts clear, and brought his whole knowledge to bear” (CP 2.47).

162

VINCENT COLAPIETRO

Such compulsion is not imposed upon the course of thought ab extra; rather we are compelled to think in this way partly as a result of our identification, however imperfect or wavering, with the norms of rationality.15 Any conclusion logically forced upon me is one I take myself to be compelled to accept, given my commitment to the canons of logic and, more generally, the norms of rationality.16 Logical compulsion and rational autonomy are, thus, of a piece. Peirce creatively appropriated the Kantian theme of autonomy, that is, self-legislation, and made it even more central to his account of knowledge, including his theory of logic, than did Kant. Just as physicians might be called upon to heal themselves, so judges might be challenged to judge themselves as severely as they have others. The cultivation of habits of self-criticism, so indispensable to the acquisition of a character in control of itself, is required of anyone who would assume the role of judge or critic in Peirce’s sense. Peirce also used Critic to designate one of the three branches of logic, that ‘which classifies arguments and determines the validity and degree of force of each kind” (1.191). What makes matters somewhat confusing is that Peirce used the term logic in a broad sense, one coextensive with semeiotic (or the general theory of signs), but also in a narrower sense, one limited to but one of the branches of logic-as-semeiotic. Just as logic-as-semeiotic is one of three normative sciences, so Critic is one of three branches of “logic” (neither the first nor the last, i.e., neither the branch of logic concerned with the most basic questions nor that one concerned with the most vital questions). Peirce was fond of correlating the three branches of “logic” broadly conceived (in other words, semeiotic) to the disciplines making up the medieval trivium. (What also leads to confusion here is that he did not settle on a set of terms for identifying the disciplines making up his semeiotic trivium, though speculative grammar, critical logic, and speculative rhetoric [see, e.g., 2.93], or, alternatively, stechiology, critic, and methodeutic [NEM 4, 21] — or some combinations of these terms or close variants! — are the ones most frequently used.) The grammar of thought in its logical sense encompasses principally a theory of signs, especially those indispensable for carrying out the work of inquiry. Since signs are the elements of thought, Peirce called this part of semeiotic stechiology as well as speculative grammar (the modifier here was of course used to distinguish grammar in this sense from its more traditional meaning). The logic of thought, again, in the irreducibly logical sense of thought concerns, in the main, criticism of reasoning as good or bad” (2.144), or inference as valid or invalid.17 Peirce occasionally called this branch of the trivium Critical Logic but also (more frequently) simply Critic. In Peirce’s case, the rhetoric of thought did not primarily concern persuasion so much as “the methods that ought to be pursued in the investigation, in the exposition, and in the application of

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

163

truth” (1.192). As a result of this emphasis, Peirce came to prefer the term methodeutic as a name for this part of the trivium.18 The relevance of recalling these points here is that the question of psychologism obviously arises in the context of logic. Hence, how one addresses this question will depend to some extent on how one conceives logic. Of course, it is also true that how one answers this question cannot help but carry implications for how one defines logic. In some crucial respects, it would be difficult (as already noted) to find a more militant opponent to psychologism than Peirce. But, given the direction in which his understanding evolved (Fisch 1986, 338-41), it can also only be difficult to see how both his speculative grammar and speculative rhetoric (or, stechiology and methodeutic) do not force him to betray his resolute commitment to advancing an “antipsychological” approach to logical topics. In semeiotic, Peirce was principally devoted to discovering “what must be the character of all signs used by a ‘scientific’ intelligence, that is to say, by an intelligence capable of learning from experience” (2.227). His critic of inferential forms was, thus, only a part of a more encompassing critical enterprise — a critic of scientific intelligence (or experimental reason). In sum, Peirce was committed to maintaining a sharp distinction between the de jure questions characteristic of logical investigation and the de facto ones definitive of such experimental investigations as psychology, anthropology, and linguistics. The grammars reconstructed by linguists are one thing, the grammar of thought at the fountainhead of logic quite another. Moreover, the laws governing our psychological processes are one thing, those directing our logical operations quite another, so much so that “laws” is being used here in an equivocal sense. Insofar as psychology is a natural science, its laws are akin to those of physics, not those of ethics. Insofar as logic is a normative science, its laws are not merely akin to, but truly instances of the laws or principles identifiable with ethics. “The phenomena of reasoning are, in their general features, parallel to those of moral conduct. For reasoning is essentially thought that is under self-control, just as moral conduct is conduct under self-control” (1.606).19 To speak of logical and moral endeavor being merely parallel, however, is potentially misleading, so Peirce immediately goes on to stress the identity he intends here: “Indeed, reasoning is a species of controlled conduct and as such necessarily partakes of the essential features of controlled conduct” (CP 1.606), that is, of ethical conduct (see 5.419-20). But, given this view of logical conduct, how is it possible to exclude from consideration the motives, judgments, actions, and habits of deliberative agents? And, given the necessity to include consideration of these factors, how is it possible to avoid lapsing into psychologism?

164

VINCENT COLAPIETRO

3. Peirce’s Critique of Psychologism This much is, at any rate, indisputable: Peirce emphatically rejected all attempts to ground logic in psychology. Here logic designates (as we have already noted) a normative discipline devoted to establishing (among other things) the laws of valid inference,20 whereas psychology signifies an experimental science, closely allied with other experimental inquiries (e.g., physiology and anatomy, on the one side, and linguistics and history, on the other), concerned principally with explaining the affective, volitional, and cognitive processes and dispositions of human and other organisms. The paradigm of psychologism was, for Peirce, to be found in John Stuart Mill’s classification of the sciences, as this was presented in An Examination of Sir William Hamilton’s Philosophy. Logic is, according to Mill, ...not a science distinct from, and coordinate to, Psychology. So far as it is a science at all, it is a part, or branch, of Psychology; differing from it, on the one hand as a part differs from the whole, and on the other, as an Art differs from a Science. Its theoretical grounds are wholly borrowed from Psychology, and include as much of that science as is required to justify the rules of the art.

The attempt to subsume the science of logic under the rubric of psychology21 and, moreover, that of reducing this normative science to an art of reasoning betrayed a fundamental misunderstanding of the status, scope, and character of this science. Whatever else it is, logic is not applied psychology. If anything, the reverse is true: psychologists are aided in their investigations by drawing upon the more abstract science of semeiotic, Peirce’s name for the study of signs in general (a study inclusive of what we ordinarily understand by logic).22 Beyond the fact of Peirce’s rejection of psychologism, much is open to dispute. Imre Lakatos has criticized Peirce for being psychologistic (1978; see Zheng). Indeed, as I highlighted in the Introduction, Peirce criticized various of his own writings for being not only too nominalistic but also too psychologistic! (Thompson, xii). Peirce’s critique of psychologism is intimately connected with his classification of the sciences and, thus, hardly comprehensible apart from the exact status Peirce accords to logic in this intricate scheme. But it is not very likely that those who are interested principally in the question of psychologism will have the patience to work through the details of this scheme. In any event, Peirce’s critique cannot be easily extracted from his most basic philosophical commitments regarding the nature of logical investigation in particular and that of the normative sciences (esthetics, ethics, and logic) in general; also those regarding his eventual identification of logic with semeiotic; and, finally, those regarding of course his conception of psychology. As is so often the case with him,23 it is far from clear whether his position is entirely consistent or coherent. Part of the difficulty is that “this

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

165

life-long student of reasonings” (3.415) seized countless occasions to denounce the reductionism implicit in psychologism, but never presented his brief against psychologism in a systematic, extended manner. So the interpreter of Peirce is required to reconstruct a position Peirce himself did not fully present. Another part of the difficulty is that, in reconstructing Peirce’s position, the expositor cannot avoid considering texts from quite distinct phases in Peirce’s intellectual career. Accordingly, apparent inconsistencies might be resolved by a consideration of the chronological development of Peirce’s sustained reflections on logical topics. But this difficulty itself points to yet another one: the still somewhat chaotic state of Peirce’s manuscripts often makes it difficult to reconstruct with the requisite accuracy the actual chronology of Peirce’s philosophical development, especially insofar as it bears upon a topic on which he frequently proclaims his stance but rarely details the bases of this stance. Regarding the topic of psychologism, there is, nonetheless, a decided tendency in Peirce’s voluminous writings. The main drift of his most considered judgment is clear enough to discern and, in some measure, to detail. My aim here, then, is the modest one of presenting the outline of what I take to be Peirce’s position regarding psychologism; and to do so in such a way as to help, on the one side, expositors of Peirce carrying this work forward and, on the other, those interested in the topic of psychologism to ascertain the relevance of Peirce to their interests. To repeat, Peirce staked his position regarding psychologism very early. In 1865, Peirce as a very young man presented at Harvard a series of lectures “On the Logic of Science.” In the course of these lectures, he made a point of criticizing Mill’s attempt to justify induction (see, e.g., W 1, 223). At the outset, he set in sharp contrast “antipsychological views” of logic (W 1, 163) from psychological conceptions of this discipline. “The one great source of error in all attempts to a Logic of Science has been the misconception of the nature and definition of logic” (W 1, 163). The next year Peirce gave a series of lectures at the Lowell Institute close to those he presented at Harvard University. “Logic has nothing at all to do with operations of the understanding, acts of the mind, or facts of the intellect” (W 1, 164). But insisting on this is not sufficient: we must go a step farther and “adopt a thoroughly unpsychological view of logic” (W 1, 164). This step is one of the most decisive in Peirce’s philosophical development, since it marks the direction in which he will move, though on occasion circuitously, until the end of his life. In addition, it points to an important affinity between Peirce and Wittgenstein. The physiological, psychological, and other processes accompanying or enabling the transformation of signs by which inferences are actually drawn by human beings in the actual circumstances of their intellectual lives are irrelevant to the assessment of the validity of these inferences. What is relevant is the thought embodied in some communicable

166

VINCENT COLAPIETRO

form such that it might be made a focal object of critical assessment. In order words, what is salient is the symbolization of thought.24 Thinking conceived as an inner or private process breathing life into signs and endowing inherently meaningless sounds or squiggles with externally ascribed meanings and implications falls utterly outside of the purview of logic. Thinking understood as a form of conduct (or species of endeavor), however, does not fall outside the scope of logic, at least as Peirce in time came to envision this discipline. But, in noting this, I am jumping ahead of the story. At this point what calls for emphasis is the distinction between thinking and thought as well as the connection between thought and symbolization. The logical process (or operation) of thinking is nothing other than that which is embodied in as sequence of signs, chief among these being icons and, specifically, diagrams, wherein logical relationships are cast in perspicuous form. Whatever other process of thought accompanies or underwrites what can, in principle, be displayed in publicly observable signs falls outside the scope of logic proper.25 This implies that thought is one with expression, though not identical with any particular system of expression; the forms of thought with which logician are preoccupied are ones open to being expressed in various ways and by different languages. In a late manuscript (MS 293; c. 1906), though one stressing the position to which Peirce was led four decades before by his rejection of psychologism, he imagines his reader asking him, “‘There is [from your perspective] no difference between Thought and Thinking, is there?’” (5).26 Peirce’s response is worth quoting in its entirety: I reply, There is indeed. ‘Thinking’ is a fabled ‘operation of the mind’ by which an imaginary object is brought before one’s gaze. If that object is a Sign upon which an argument may turn [i.e., if that imaginary object performs an illative function], we call it a Thought. All that we know of the ‘Thinking’ is that we afterwards remember that our attention was actively on the stretch, and that we seemed to be creating Objects of Transformations of Objects while noting their analogy to something supposed to be real, we choose to call an ‘operation of the mind’; and we are, of course, quite justified in doing so, provided it be well understood that its being consists merely in our so regarding it, just as Alexander, Hannibal, Caesar constitute a single quaterion or plural of four, as long as we put them together in thought. The ‘operation of the mind’ is an ens rationis. This is my insufficient excuse for speaking of it as ‘fabled.” (MS 293, 5-6)

At least these points merit comment. It is not absurd to suppose that a process akin to introspection allows us to glimpse something of our own processes of thinking. But, strictly speaking, introspection is a process of retrospection (“All that we know of the ‘Thinking’ is what we afterwards remember...”). Even so, the process of thinking is not so much observed as hypothesized: it is a guess put forth as a way of explaining the conjuring and transformation of objects in our imaginations. To use a distinction drawn by

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

167

Michael Polanyi, we are focally aware of imaginary objects and (at least) their more manifest transformations, but (at most) only peripherally aware of the processes by which these objects are brought to our attention and, then, transformed under our scrutiny. Insofar as these transformations are private, incommunicable processes, they are irrelevant to logic; however, insofar as they are, in principle, public, communicable procedures, they possibly fall within the scope of logic. The objects being considered here are only secondarily objects, entia to which our attention is drawn or ones capable of resisting, in some manner and measure, our efforts to make of them whatever we wish; they are primarily signs, entia drawing our attention to something other than themselves. Their being resides not in their opposition (or secondness), but in their mediation (or thirdness) — their capacity to link what otherwise would be disparate.27 These imaginary objects are, however, signs in a distinctive context, that of objective inquiry (“If that object is a Sign upon which an argument may turn [or an inference may hinge], we call it a Thought”). Yet this is thought in a logical, not a psychological, sense; for it is defined by its role or function in an inference. The explication of this late manuscript returns us, just at this point, to the additional step any antipsychological approach to logical investigations must take, if it is to move decisively beyond the province of psychologism. For the purposes of illustration, Peirce offers the syllogism: All conquerors are Butchers Napoleon is a conqueror Napoleon is a Butcher. Peirce informed his audience: “Now this [sequence of propositions] has a particular logical character to me as I write it; it has the same to you as you read it; it will have the same if you read it tomorrow; and while it remains on the board it will retain the same character to whoever can read it” (W 1, 164). In other words, the form of this inference is both trans-personal and trans-temporal. To what extent is this form dependent on thought, on a process of thinking by which this conclusion is inferred from these premises? Or, in Peirce’s words, “is this logical character a form of thought only?” (W 1, 164-65). Of course, this itself depends on how thought is understood in this context. But thought as a unique, psychological process is one locus of the logical character of this inferential form, not the origin of or warrant for this form. So, Peirce notes: “My thought when I wrote it was a different event from each one of your thoughts, and your thoughts will be different if you read it again from what they were when you read it just now. The thoughts were many, but the form was one” (W 1, 165). The words on the blackboard as much embody the logical form of this syllogistic inference as do the thoughts by which the propositions are understood and the conclusion drawn. Logical forms reside in whatever symbols they are or can

168

VINCENT COLAPIETRO

be instantiated, including of course those symbols indicative of the relationship among the other symbols. “The psychological view is that these forms are only realized in thought, and that language is essential to thought” (W 1, 165). In contrast, the “antipsychological view is that they are forms of all symbols whether internal or external [whether those employed by the imagination of some individual or those instituted by the members of some community] but that they only are [so] by virtue of possible thought” (ibid.). This decisive step in the direction of a thoroughly antipsychologistic logic is, thus, the recognition: the “logical form is already realized in the symbol itself; the psychologists [and, of greater relevance to us, the psychologists in logic] say that it is only realized when the symbol is understood” (W 1, 16566). The comprehension of a symbol or series of symbols would, upon the latter view, be a mental operation in the psychological sense. Note that Peirce includes a reference to thought (more precisely, to “possible thought”) in his characterization of antipsychologistic logic. But the form of thought to which he is referring here is that of thought in general, not any peculiar or particular manner or species of thought. Such thought turns out to be semiosis, or the activity of signs.28 It is by no means the thought peculiar to conscious, deliberative agents such as human beings are capable of becoming. In this sense, logical machines as much as conscious organisms are capable of thought, though (at least at this juncture) not the higher grades of reflexive criticism characteristic of language-using animals such as human beings. In a different context, Peirce suggests: A decapitated frog almost reasons [when it uses one leg to wipe acid off the other]. The habit that is in his cerebellum serves as a major premises. The excitation of a drop of acid is his minor premises. And his conclusion is the act of wiping it away. All that is of any value in the operation of ratiocination is there, except only one thing. What he lacks is the power of preparatory meditation. (6.286; see 2.711).

Peirce is fairly consistent in using the term reasoning to designate an inference over which a process of deliberate control has been exercised, in contrast to the more rudimentary process of inference, by which the truth of one or more propositions is taken as the warrant for the truth of another. A decapitated frog is, regarding inference properly understood, no worse off than a logical machine: the action of try to wipe the acid off its leg embodies nothing less than an inference. It lacks, however, the capacity to reason, for it cannot subject its inferences to ever higher levels of logical assessment. Thus, a seemingly mechanical action (the reflex of the frog in response to this stimulus and a straightforwardly mechanical operation (that of the computations done by machines) are ones in which logical forms, albeit generally ones of a quite rudimentary form, are realized. Thought in neither the distinctively human nor strictly psychological sense accounts for the logic of these operations.

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

169

What Peirce asserts regarding semiosis might, with equal justice, also be asserted of inference: “No doubt, intelligent consciousness must enter into the series” at some point (2.303), at least if the series is carried out to any great extent. In any case, Peirce’s conception of logic does not preclude reference to thought in every sense, though it does rule out reference to thought in its psychological sense. Intelligent consciousness in the form possessed by deliberative agents able to mind what they say and infer cannot be denied What threatens to muddy the waters, however, is a distinction drawn by Peirce for the sake of clarifying his mature understanding of logic as a normative science. Logic as a theory of inquiry (more fully, a normative account of objective investigation) is, at bottom, a branch of ethics. While logic is specifically concerned with self-controlled thought, ethics is more generally preoccupied with self-controlled action. That is, inquiry — thus, science — is a form of conduct and, as such, it is partly defined by its susceptibility to criticism. Criticism is to be exercised in the service of control; indeed, self-criticism makes self-control possible. Scientific investigation is a self-corrective practice because scientific investigators are self-critical, self-controlling agents, who mind carefully what they say and otherwise do, deliberately altering their conduct in light of the consequences of their exertions. To make sense out of any one of the normative sciences, it is necessary to see that the realization of this endeavor flows from the systematic deliberation of rational agents. In order to describe, interpret, or evaluate the deliberations of such agents, one must have recourse to such concepts as intention, belief, doubt, and investigation. But, from Peirce’s perspective, these are not in the strict sense psychological truths. To cite a passage quoted in the Introduction, logic “ does rest on certain facts of experience among which are facts about men” and women (5.110; emphasis added). Not all facts of experience pertaining to the mind of humans are psychological facts in the sense relevant to the question of psychologism. So Peirce insists upon a distinction: Under an appeal to psychology is not meant every appeal to any fact relating to mind. For it is, for logical purposes, important to discriminate between facts of that description which are supposed to be ascertained by the systematic study of mind [i.e., psychological truths in the strict sense], and facts the knowledge of which altogether antecedes such study, and is not in the least affected by it [truths Peirce sometimes identified as psychical, in contradistinction from psychological]; such as the fact that there is such a state of mind as doubt, and the fact that the mind struggles to escape from doubt. (2.210)

For the restricted purpose of classifying and evaluating forms of inference, there need be no recourse even to the vague, antecedent truths derived from our experience of ourselves as rational, deliberative agents.

170

VINCENT COLAPIETRO

For the broader purpose of offering a compelling account of our logical investigations as normative investigations, reliance on. such truths seems inescapable. In any event, Peirce’s early efforts to articulate the logic of science in time evolved into his mature project of formulating an ethics of inquiry, wherein the normative science of ethics is conceived in such a way as to provide the principles and norms for the normative science of logic.

4. The Space of Signs But, once Peirce has introduced into the province of logic such considerations as these, has he not completely compromised his antipsychologism? I do not suppose that he has, but I also do not know how to show the consistency of his opposition to psychologism, apart from offering a completely convincing account of his nuanced conception of logic, precisely in its status as a normative science. It might nonetheless be helpful to sketch some of the more salient features of his distinctive approach to this normative science. I will do so principally as a way of suggesting how a critique of psychologism, via a transformation of logic into semeiotic, entails nothing less than a radical reconceptualization of “the nature and definition of logic.” In turn, such a reconceptualization of logic cannot help but redefine the context in which the question of psychologism is most properly addressed. Peirce widened the scope of logic, making it coextensive with a general theory of signs. In addition, he redefined the formal object of experimental psychology, making it much wider than the study of consciousness. As a logician — indeed, as one whose acquaintance with logic went back to his youth and whose work as a philosopher was described principally as that of a logician — Peirce hesitantly modified the shape of the discipline with which he more deeply identified than any other. As someone who performed only a handful of psychological experiments, Peirce very reluctantly — but quite strenuously — proposed that psychologists were mistaken about the nature of their own undertaking. Peirce noted: “No doubt, it seems an extraordinary piece of presumption for a man to tell a large body of scientific men fro whom he professes high respect that they do not know what are the problems which they are endeavoring to solve; that while they think they trying to make clear the phenomena of consciousness, it is really something quite different that they are trying to do” (CP 7.367). Even so,

Peirce felt compelled to define the formal object of experimental psychology in a manner at odds with the disciplinary self-understanding of the psychologists of his own day: For if psychology were restricted to phenomena of consciousness, the establishment of mental associations, the taking of habits, which is the very

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

171

market-place of psychology, would be outside its boulevards. To say of such departments of psychology, — from every point of view, the most essential parts of it, — that they are studies of consciousness, is as if the ichthyologist were to define his science as a study of water. (7.367)

The normative science of logic was, in Peirce’s judgment, far removed from the experimental science of psychology. His advocacy of doctrines such as pragmatism and anthropomorphism, however, seem to compromise or even completely undermine his stance toward psychologism. He came to clarify his logical maxim of pragmatism and, in the course of doing so, he defended “a shocking learning toward anthropomorphic conceptions” (5.212). He went so far as to insist: “I hold...that man is so completely hemmed in by the bounds of his possible practical experience, his mind is so restricted to being the instrument of his needs, that he cannot, in the least, mean anything that transcends those limits” (5.536; see 5.47). Peirce’s anthropomorphism seems to entail psychologism. He appears to have sensed that this is the case. Peirce’s characterization of logic as the ethics of the intellect also appears to drive him in the direction of psychologism. He expressed misgivings about many of his own formulations precisely because they seem to imply that logic can be based on psychology. One of the ways he tried to avoid these embarrassing implications was by drawing a distinction between psychological and psychical truths. One of the most important underlying presumptions of his logical investigations was that of a deliberative agent able indirectly to control his conduct, because the cultivation of habits fell within the scope of this agent’s power. For certain purposes, we can abstract a purely formal definition of sign from the concrete phenomena with which we are so intimately familiar. Though the signs with which we are acquainted almost always involve minds as either their utterers or interpreters, signs as such are definable apart from any reference to mind. Peirce often felt the need to throw a sop to Cerebus, but he often took pains to define a sign without doing so. Perhaps his most famous definition of sign is, in truth, a compromised definition, for it explains the nature of a sign by appealing to the agency of mind: “A sign, or representamen, is something which stands to somebody for something in some respect or capacity. It addresses somebody, that is, creates in the mind of that person an equivalent sign, or perhaps a more developed sign” (CP 2.228). A text akin to this is found in one of Peirce’s letters to Victoria Lady Welby, though here he explicitly notes the limitation of his characterization: “I define a Sign as anything which is so determined by something else, called its Object, and so determines an effect upon a person, which effect I call its Interpretant, that the latter [the Interpretant] is mediately determined by the former [the Object]” (Wiener, 404). But, then, Peirce immediately notes: “My insertion of ‘upon a person’ is a sop to Cerebus, because I despair of making my own broader conception understood” (ibid.). But, given definitions both antecedent and subsequent to this one, definitions in

172

VINCENT COLAPIETRO

which all reference to mind is intentionally eliminated, it is important not to make too much of Peirce’s despair in this regard. His hope of crafting a purely formal and completely antipsychological definition of a sign never truly died. Hence, in his application for a grant from the Carnegie Foundation for the purpose of supporting his systematic elaboration of his entire philosophy, Peirce indicated: “A definition of a sign will be given which no more refers to human thought [thus to the human mind or any analogous agency] than does the definition of a line as the place which a particles occupies, part by part, during a lapse of time” (NEM 4, 20). This is close to the definition formulated in a letter to Lady Welby quoted above, except in one decisive respect (the nature of the interpretant is not identified as an effect upon any mind): “Namely, a sign is something, A, which brings something, B, its interpretant sign determined or created by it, into the same sort of correspondence with something, C, its object, as that in which [the sign] itself stands to C” (NEM 4, 20-21). Given the topic of this paper and thus the question of how to conceive logic (in particular, how to consider the source from which logic derives its conclusions, if not from psychology or any other such discipline), Peirce’s next assertion is of the utmost importance: “It is from this definition, together with a definition of ‘formal,’ that I deduce mathematically the principles of logic” (NEM 4, 21). This assertion cannot be fully unpacked here. For our purposes, let it suffice to highlight the following points. The logical notion of sign concerns those relationships (iconic, indexical, and above all illative29 relationships) implicated in inference, only insofar as these relationships obtain without reference to mind. The relationships constitutive of any sign in its most general form might be diagrammed, just as those pertaining to sequences of signs wherein some part of these is put forth as a warrant for another part might also be diagrammed. In these diagrams and other symbols, the logical form is fully realized (W 1, 165-66). But the sphere of inquiry in which the logical relationships obtaining among quite various signs are made into the focal object of scientific intelligence is, like any other such sphere, one carved out by means of what Peirce called prescissive (or precisive) abstraction. Yet, within this prescinded domain of logical investigation, hypostatic abstractions play an enormous role (see Zeman). T. L. Short helps us, with his characteristic precision and clarity, to discern this distinction: “Hypostatic differs from precisive abstraction: the latter consists in selecting one feature from a complex of features or a concrete object, as when we say of something that it is large, while the former [or hypostatic abstraction] makes this feature into an object of reference, as when we speak of something’s largeness (4.322; 2.364)” (1997: 295). By means of hypostatic abstraction, “a mathematician conceives the particle as occupying a point. The place is now made a subject of thought....When the mathematician regards an operation as itself the subject of operations, he is using abstraction...” (NEM 4, 11).

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

173

We are always already caught up in an exchange of signs, far outstripping our comprehension, even apart from our initiation or intention. We inhabit a space of signs, wherein what is manifest is ineluctably taken as evidence for what is hidden, absent, or in some other way (at least) provisionally inaccessible. Hence, there is a space of reasons inseparable from space of signs in which we live and move and have our being. This dimension of our world deserves to be considered in its own right. Mathematics and phenomenology enable us to do so. The operations of mathematics provide us with an intimate acquaintance with countless instances in which various aspects of the empirical world are considered in their own right, apart from all else. Phenomenology partly takes its inspiration from these operations, paying special attention to the recursive functions wherein operations may be performed on themselves or other operations (see Savan 1987-88: 12-14). There is, at any rate, a phenomenological derivation of a purely formal definition of sign. From this purely formal definition of sign, there is a “mathematically” deduction of the laws of logic. For certain purposes, then, we can abstract a purely formal definition of sign and, from this and allied conceptions, deduce the purely formal laws of logic. But the work of the logician is hardly exhausted by the execution of these tasks. It may properly extend to nothing less than a comprehensive, normative account of scientific intelligence in its most basic sense but also in its diverse forms. In any event, what we encounter in Peirce’s writings is a logician who, in some respects, protects the autonomy of logic as a formal, normative discipline but who, in other respects, insists upon offering a robust account of this and the other two normative disciplines (ethics and esthetics), such that securing this field of inquiry, precisely as an irreducibly normative undertaking, is the singular achievement of that historical community of deliberative agents known as logicians. The question of whether or not Peirce vitiates his critique of psychologism by incorporating within logic the array of considerations needed to make adequate sense out of self-controlled thought — this question cannot be answered without assessing the viability of Peirce’s inclusive, semeiotic conception of logic. This essay in its entirely, let alone at its very conclusion, cannot be the place for undertaking such an assessment. The consistency and force of Peirce’s critique of psychologism, hence, depends in the last analysis on the cogency and legitimacy of his understanding of logic. At this juncture, this can only mean that the question is an open one. Department of Philosophy The Pennsylvania State University, USA

174

VINCENT COLAPIETRO

NOTES 1 In “Peirce’s Place in American Thought,” Max H. Fisch quotes from a series of lectures given by Peirce in 1892 to the Lowell Institute: “I am above all things a student of logic; and have especially devoted myself to the historical study of the logic of science” (314). 2 In 1865, at the age of twenty-six, Peirce delivered at Harvard a series of lectures entitled “On the Logic of Science.” In the opening paragraphs of the first lecture, he sharply distinguished psychological from unpsychological approaches to logic, emphatically aligning himself with the latter (W 1, 162-64). 3 In “How to Make Our Ideas Clear,” Peirce argued for the need to go beyond abstract verbal definitions — positively, for pragmatic, thus experiential, clarification of the terms on which our investigations hinge. Famously, but not all too clearly, he proposed this maxim: “Consider what effects, that might conceivably have practical bearings, we conceive the object of our conception to have. Then, our conception of these effects is the whole of our conception of the object” (5.402). 4 Nicholas St. John Green, a lawyer in Boston, “often urged [upon Peirce and the other members of the Metaphysical Club, an informal discussion group,] the importance of applying Bain’s definition of belief, as ‘that upon which a man is prepared to act.’ From this definition, pragmatism is scarce more than a corollary; so that I am disposed to think of him [St. John Green] as the grandfather of pragmatism” (5.12). This definition of belief can be found in Bain’s The Emotions and the Will (1875), Chapter 11, 505. It is of course not insignificant that Bain was a psychologist and, moreover, that the definition of belief at the center of pragmatism is taken from a work in psychology. 5 This is the year in which William James presented at the University of California, Berkeley, his famous lecture “Philosophical Conceptions and Practical Results.” What Peirce’s essays in the Popular Science Monthly two decades earlier failed to accomplish James’s lecture dramatically did — a hearing for pragmatism. See Max H. Fisch’s “American Pragmatism Before and After 1898” (1986, chapter 15). 6 Peirce tended to think of James as a gifted psychologist but a suspect philosopher, suspect precisely because James lacked training in mathematics and hardly ever resisted expressing his antipathy toward logic. For Peirce, pragmatism was a logical doctrine; thus, the transformation of it into a psychological position distorted his position beyond recognition. Hence, he eventually proposed to call his own position pragmaticism, in order to differentiate it from the psychological, subjectivistic, and humanistic doctrines flying under the banner of pragmatism. He felt certain that his neologism was “ugly enough to be safe from kidnappers” (5.414). 7 “Among the things which the reader, as a rational person, does not doubt, is that he not merely has habits, but also can exert a measure of self-control over his future actions; which means, however, not that he can impart to them any arbitrarily assigned character, but, n the contrary, that a process of self-preparation will tend to impart to action (when the occasion for it shall arise), one fixed character” (5.418). “Now the theory of Pragmaticism was originally based, as anybody will see who examines the papers of November 1877 and January 1878 [“The Fixation of Belief” and “How to Make Our Ideas Clears,” respectively], upon the study of that experience of the phenomena of self-control which is common to all grown men and women; and it seems evident that to some extent, at least, it must always be so based. For it is to conceptions of deliberate conduct that Pragmaticism would trace the intellectual purport of symbols; and deliberate conduct is self-controlled conduct. Now control may itself be controlled, criticism itself subjected to criticism; and ideally there is no obvious definite limit to the sequence” (5.442). 8 “All knowledge whatever comes by observation; but different sciences are observational in such radically different senses that the kind of information derived from the observation of one department of science (say natural history) could not possibly afford the information

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

175

required of observation by another branch (say mathematics)” (1.238). In a different context, Peirce notes: “But the imaginary constructions of the mathematician, and even dreams, so far approximate to reality as to have a certain degree of fixity, in consequence of which they can be recognized and identified as individuals. In short, there is a degenerate form of observation which is directed to the creations of our own minds — using the word observation in its full sense as implying some degree of fixity and quasi-reality in the object to which it endeavours to conform. Accordingly, we find that indices are absolutely indispensable in mathematics” (2.305). 9 “Would it not, at any rate, in the present state of science, be good scientific policy, for those who have both a talent and a passion for eliciting the truth about such matters, to institute a cooperative cenoscpic attack upon the problems of the nature, properties, and varieties of Signs, in the spirit of twentieth-century science[?]” (EP 2: 462). 10 Signs are, in Peirce’s judgment, “the only things with which a human being can, without derogation, consent to have any transaction” (6.344). 11 Peirce once recalled: “when I was a babe in philosophy my bottle was filled from the udders of Kant” (CP 2.113). Even before having encountered Kant, however, Peirce read careful Friedrich Schiller’s Letters on Aesthetic Education. But the first philosopher on whom he truly cut his teeth (an apt metaphor here given Peirce’s own — “the udders of Kant”!) was the author of the Critique of Pure Reason. 12 “As for that phrase ‘studying in a literary spirit’ it is impossible to express how nauseating it is to any scientific man, yes even to the scientific linguist” (CP 1.33). See Haack (1998). 13 Logic is, from Peirce’s perspective, “the theory of right reasoning, of what reasoning ought to be, not of what it is. On that account, it used to be called a directive science, but of late years adjective normative has been generally substituted” (CP 2.7; see 2.52). 14 In a letter to William James dated October 3, 1904, Peirce stressed: “phenomenology is one science and psychology a very different one. I know that you are not inclined to see much value in distinguishing between one science and another. But my opinion is that it is absolutely necessary to any progress [in the sciences]. The standards of certainty must be different in different sciences, the principles to which one science appeals altogether different from those of the other. From the point of view of logic and methodological development the distinctions are of the greatest concern. Phenomenology has no right to appeal to logic, except to deductive logic [as exemplified in the practice of mathematical reasoning]. On the contrary, logic must be founded on phenomenology. Psychology, you may say, observes the same facts as phenomenology does. No. It does not observe the same facts. It looks upon the same world; — the same world that the astronomer looks at. But what it observes in that world is different. Psychology of all sciences stands most in need of the discoveries of the logician, which he makes by the aid of the phenomenologist” (8.297; emphasis added). 15 John E. Smith suggests it is imperative to distinguish between formal and living reason. There is no question that: “Rational assurance in the form of argument that is rationally compelling involves a measure of distance or separation of the individual self from the object of proof” (109). But it is equally unquestionable that the force of an argument is not forced upon rational agents ab extra. Hence, Smith stresses: “Living reason is not a power alien to the self, a purely universal and abstract set of rules which is indifferent both to the individual thinker and to the differences between [or among] subject matters thought about. Living reason is the process whereby the self seeks to trace out rational pattern in its experiences; the self recognizes and acknowledges its own nature as a rational being in that process. There is no sense [or, at most, very little sense] of external compulsion and no sense of being determined by a merely formal necessity” (113). 16 See Charles Taylor’s “Rationality” for why rationality is not reducible to logicality. 17 Peirce immediately goes on to stress: “Now it is idle so to criticize an operation which is beyond all control, correction, or improvement” (2.144). 18 Max H. Fisch suggests that Peirce’s semeiotic-as-logic evolved in the direction of making methodeutic (or speculative rhetoric) the most important branch of his logical trivium, though

176

VINCENT COLAPIETRO

Peirce did not live long enough to elaborate this branch as fully as he had the other two (1986, 292-93). This suggestion is, in my judgment, correct. 19 Thought in its logical is, for Peirce, a species of conduct, so there is something misleading about this passage insofar as it sets thought in contrast to conduct. 20 One of Peirce’s earliest but also most famous series of articles, the so-called “cognitionseries” appearing in The Journal of Speculative Philosophy, grew out of an invitation from William Harris to write a piece 21 It is interesting and important that the Swiss linguist Ferdinand de Saussure, who roughly at the same time as Peirce, imagined a comprehensive science of signs (of which his own more restricted study of language would be only a part), also subsumed semiologiy under psychology: “It is...possible to conceive of a science which studies the role of signs as part of social life. It would form part of social psychology, and hence of general psychology. We shall call it semiology (from the Greek semeîon, sign). It would investigate the nature of signs and the laws governing them. Since it does not yet exist, one cannot say for certain that it will exist. But it has a right to exist, a place ready for it in advance. Linguistics is but one branch of this general science. The laws which semiology will discover will be laws applicable in linguistics, and linguistics will thus be assigned to a clearly defined place in the field of human knowledge” (15-16). 22 “Of course, psychologists ought to make, as they are in point of fact they are making, their own invaluable studies of the sign-making and sign-using functions, — invaluable, I call them, in spite of the fact that they cannot possibly come to their final conclusions, until other more elementary studies have come to their first harvest, — studies that it is natural for the psychologist to regard...with something of the kind of disdain as he may naturally bestow upon the still more vacuous studies of the pure mathematician, or would, if mathematics were not so old a science that is rich granaries command respect” (EP 2: 461). These “more elementary studies” fall within logic-as-semeiotic. 23 What John E. Smith noted long ago in reference to himself remains somewhat true of many readers of Peirce today: “...I did [at first] confound the form in which his writings have come to us with the logical structure of his thought. To discover Peirce’s theory on any given topic one has to plow through volumes of papers, drafts and revisions of these drafts, reviews, and partially completed manuscripts” (1970, 80-1). But, in doing so, he discovered that ‘on any given topic there is a clear drift to his [Peirce’s] thought, even if he [Peirce] did not make it easy for us to find it” (81). 24 In a note on an article entitled “On the Natural Classification of an Argument” (1867), the mature Peirce looking back explains his preference for the term argument: “I was inclined to replace argument in this essay by inference (for, as Locke well says, ‘to infer is nothing but, by virtue of one proposition laid down as true, to draw in another as true’...), until I reflected that to do so would have the air of admitting what I could never admit, that logic is primarily conversant with unexpressed thought and only secondarily with language” (2.461nl). He suggests also: “all that we know of thought is but a reflection of what we know of its expression” (2.466n1). This connects with several an early characterization of logic: “Logic is objective symbolistic”” and, in turn, symbolistic “is the semiotic of symbols” (W 1, 303). This characterization was an attempt to articulate the anti-psychological approach to logic defended in the series of lectures “On the Logic of Science” (Harvard, 1865). 25 “There is no reason why ‘thought’...should be taken in that narrow sense in which silence and darkness are favorable to thought. It should rather be understood as covering all rational life, so that an experiment shall be an operation of thought” (5.420). 26 “How does this gibe with your former statement that reasoning is ‘self-controlled thought’? There is no distinction between Thought and Thinking, is there?” (MS 293, 5). 27 Peirce’s three most basic categories were, most frequently, identified as firstness, secondness, and thirdness. The function of these categories is principally heuristic: they were designed by him to guide and goad inquiry. One of the principal meanings of firstness is qualitative immediacy, what a thing is apart from anything else. Secondness designates brute

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

177

opposition, what anything is by virtue of its relation to something other than itself. Thirdness encompasses mediation, intelligibility, continuity, and a vast array of allied conceptions. 28 In “The Blue Book” Ludwig Wittgenstein proposes: “It is misleading then to talk of thinking as of a ‘mental activity.’ We may say that thinking is essentially the activity of operating with signs. The activity is performed by the hand, when we think by writing; by the mouth and larynx, when we think by speaking; and if you think by imagining signs or pictures, I can give you no agent that thinks. If then you say that in such cases the mind thinks, I would only draw your attention to the fact that you are using a metaphor, that here the mind is an agent in a different sense from that in which the hand can be said to be the agent in writing” (6-7). He goes on to assert: “If again we talk about the locality where thinking takes place we have a right to say that this locality is the paper on which we write or the mouth which speaks. And if we talk of the head or the brain as the locality of though, this is using the expression ‘locality of thinking’ in a different sense” (7). Peirce’s own position is very close to, if not identical with, Wittgenstein’s: “There is no reason why ‘thought’...should be taken in that narrow sense in which silence and darkness [i.e., the privacy of an allegedly inaccessible consciousness] are favorable to it. It should rather be understood as covering all rational life, so that an experiment shall be an operation of thought” (5.420). Elsewhere he insists “it is much more true that the thoughts of a living writer are in any printed copy of his book than they are in his brain” (7.364). He goes so far as to claim: “A psychologist cuts out a lobe of my brain (nihil animale me alienum puto) and then, when I find I cannot express myself, he says, ‘You see your faculty of language was localized in that lobe.’ No doubt it was; and, so if he had filched my inkstand, I should not have been able to continue my discussion until I had got another. Yea, the very thought would not come to me. So my faculty of discussion is equally localized in my inkstand. It is localization in the sense in which a thing may be in two places at once” (7.366). See Skagestad 1999. 29 Peirce claimed “the purpose of signs — which is the purpose of thought — is to bring truth to expression. The law under which a sign must be true is the law of inference; and the signs of a scientific intelligence must, above all other conditions, be such as to lend themselves to inference. Hence, the illative relation is the primary and paramount semiotic relation” (2.444n1).

REFERENCES

Abbagnano, Nicola. “Psychologism.” Encyclopedia of Philosophy. Paul Edwards, Editor in Chief. New York: Macmillian, 1967; volume 6, 52021. Anderson, Douglas R. Strands of System: The Philosophy of Charles S. Peirce. West Lafayette: Purdue University Press, 1995. Apel, Karl-Otto. Charles S. Peirce: From Pragmatism to Pragmaticism. Amherst: University of Massachusetts Press, 1981. Brent, Joseph. Charles Sanders Peirce: A Life, Revised and Enlarged Edition. Bloomington and Indianapolis: Indiana University Press, 1998. Buchler, Justus. Charles Peirce’s Empiricism. New York: Harcourt, Brace, 1939. Colapietro, Vincent M. Peirce’s Approach to the Self. Albany: SUNY Press, 1989. Colapietro, Vincent M. “The Dynamical Object and the Deliberative Subject.” The Rule of Reason: The Philosophy of Charles Sanders

178

VINCENT COLAPIETRO

Peirce, edited by Jacqueline Brunning and Paul Forster, 262-88. Toronto: University of Toronto Press, 1997. Delaney, C. F. Science, Knowledge, and Mind: A Study in the Philosophy of C. S. Peirce. Notre Dame: University of Notre Dame Press, 1993. Esposito, Joseph. Evolutionary Metaphysics: The Development of Peirce’s Theory of Categories. Athens: University of Ohio Press, 1980. Fisch, Max H. Peirce, Semeiotic, and Pragmatism: Essays by Max H. Fisch, Laine Ketner and Christian J. W. Kloesel. Bloomington: Indiana University Press, 1986. Gallie, W. B. Peirce and Pragmatism. Harmondsworth: Penguin, 1952. Goudge, Thomas A. The Thought of C. S. Peirce. Toronto: University of Toronto Press. Haack, Susan. Manifesto of a Passionate Moderate: Unfashionable Essays. Chicago: University of Chicago Press, 1998. Hausman, Carl R. Charles S. Peirce’s Evolutionary Philosophy. New York: Cambridge University Press, 1993. Hookway, Christopher. Peirce. London and New York: Routledge, 1985. Kent, Beverley. Charles S. Peirce: Logic and the Classification of the Sciences. Kingston & Montreal: McGill-Queen’s University Press, 1987. Lakatos, Imre. The Methodology of Scientific Research Programmes: Philosophical Papers, volume 1. Cambridge: Cambridge University Press, 1978. Misak, C. J. Truth and the End of Inquiry: A Peircean Account of Truth. Oxford: Clarendon Press, 1991. Murphey, Murray G. The Development of Peirce’s Philosophy. Indianapolis: Hackett, 1993. Peirce, Charles Sanders. The Collected Papers of Charles Sanders Peirce (CP), edited by Charles Hartshorne and Paul Weiss (volumes 1-6) and Arthur Burks (volumes 7-8). Cambridge: Harvard University Press, 1931-1935. Peirce, Charles Sanders. Contributions to the Nation (CN), compiled and annotated by Kenneth Laine Ketner and James Edward Cook. Lubbock: Texas Tech Press, 1975-1978. Peirce, Charles Sanders. The New Elements of Mathematics (NEM), edited by Carolyn Eisele. Atlantic Highlands: Humanities Press, 1976. Peirce, Charles Sanders. Writings of Charles S. Peirce: A Chronological Edition (W), edited by Max H. Fisch. Bloomington: Indiana University Press, 1982. Peirce, Charles Sanders. The Essential Peirce, volume 2, edited by the Peirce Edition Project. Indianapolis: Indiana University Press, 1998. Potter, Vincent G. Charles S. Peirce: On Norms and Ideals. New York: Fordham University Press, 1997).

C.S. PEIRCE’S CRITIQUE OF PSYCHOLOGISM

179

Potter, Vincent G. Peirce’s Philosophical Perspectives. New York: Fordham University Press, 1996. Saussure, Ferdinand de. Course in General Linguistics, translated by Roy Harris. La Salle: Open Court Press, 1991. Savan, David. An Introduction to C. S. Peirce’s Full System of Semeiotic. Toronto: Toronto Semiotic Circle, 1987-88. Short, T. L. “Hypostatic Abstraction in Self-Consciousness.” The Rule of Reason: The Philosophy of Charles Sanders Peirce, edited by Jacqueline Brunning and Paul Forster, 289-308. Toronto: University of Toronto Press, 1997. Skagestad, Peter. “Peirce’s Inkstand as an External Embodiment of Mind.” Transactions of the Charles S. Peirce Society, 35, 3, 1999, 551-61. Smith, John E. Themes in American Philosophy: Purpose, Experience and Community. New York: Harper, 1970. Smith, John E. Experience and God. New York: Fordham University Press, 1995. Thompson, Manley. The Pragmatic Philosophy of Charles S. Peirce. Chicago: University of Chicago Press, 1953. Wittgenstein, Ludwig. The Blue and Brown Books. New York: Harper, 1958. Zeman, J. Jay. “Peirce on Abstraction.” The Relevance of Charles Peirce, edited by Eugene Freeman, 292-311. La Salle: Open Court Press, 1983). Zheng, Lan. “Peirce, Lakatos, and Truth.” Living Doubt, edited by Guy Debrock and Menno Hulswitt, 197-208. Amsterdam: Kluwer Academic Publishers, 1994.

This page intentionally left blank

MICHAEL BRADIE

QUINEAN DREAMS OR, PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY

W.V.O. Quine was one of the leading philosophers of the twentieth century. Although best known for his work in logic and the philosophy of language, Quine contributed to many areas in philosophy. In epistemology, he championed a version of naturalistic epistemology that supporters found revolutionary and critics saw as beside the point. In this paper, I examine and evaluate Quine’s position, focusing, in particular, on the vexing question of the extent to which naturalized epistemologies can countenance norms. Section 1 provides a sketch of Quine’s naturalistic position and its epistemological implications. Section 2 explores the relationship between Quine’s naturalized epistemology and evolutionary epistemologies. Section 3 locates Quine within the tradition and its continuation. Section 4 deals with Quine and the skeptic. Section 5 evaluates the charge that Quine’s naturalized epistemology fails to deal with the normative concerns of the tradition and thus qualifies as “epistemology” in name only. Section 6 provides an overall assessment of the viability of Quine’s epistemological program and its claim to be an heir to the tradition.

1. Quine’s Naturalism Quine has been a practicing naturalist for most, if not all of his philosophical career. In one of his last statements concerning his view, he held that the gist of his naturalism was the endorsement of the traditional scientific method as the best methodology for the pursuit of truth about the world. This traditional method he took to be the hypothetico-deductive method (Quine, 1995b, 257). That same year, in From Stimulus to Science, Quine characterized naturalism, with a nod to Carnap’s Aufbau, as the “rational reconstruction of the individual’s and/or the race’s actual acquisition of a responsible theory of the external world.” (Quine 1995a, 16) 181 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 181-194. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

182

MICHAEL BRADIE

As such, it is an enterprise that is “part and parcel of empirical science”. The motivation, he insists, is still “philosophical” but “unlike the old epistemology, we seek no firmer basis for science than science itself; so we are free to use the very fruits of science in investigating its roots. It is a matter, as always in science, of tackling one problem with the help of our answers to others” (Quine 1995a, 16). Of course, Carnap’s program, unlike Quine’s, is a “foundational” one. The Aufbau stands as a continuation of the Cartesian program of Russell’s Our Knowledge of the External World. The idea was to “construct” human knowledge of physical objects out of the sensory data that constitutes our evidence of them. In 1969, Quine rejected this approach in favor of the “construction” of knowledge using psychology, evolutionary theory and whatever other scientific information we have at our disposal (Quine 1969). Roger Gibson has identified the two sources of Quine’s naturalism as [1] Quine’s holism, and [2] Quine’s “unregenerate realism” (Gibson 1987; see Quine 1981, 72). Quine’s holism derives from his qualified acceptance of Duhem’s thesis. His realism is a reflection of his acceptance of the results of natural science, along with the uncertainties inherent in them, coupled with his belief that nothing better is forthcoming. Every result is negotiable but not all at once. In early days, Quine argued that two “cardinal empiricist tenets remain unassailable”: [E1] “[W]hatever evidence there is for science is sensory evidence,” and [E2] “all inculcation of meanings of words must rest ultimately on sensory evidence.” (Quine 1969, 75) Gibson has labeled [E1] the “doctrinal” side of Quine’s modified empiricism and [E2] the “conceptual” side. But as Quine notes later (Quine, 1995c), these “unassailable theses” are, in principle, subject to revision as well. As far as our best science of the day tells us, [E1] is true, but telepathy, clairvoyance and other direct methods of mental implanting can not be ruled out tout court. This is not to say we need to take them seriously but if some wellconfirmed science of the future were to do so, we would need to jettison empiricism. This would not entail the rejection of naturalism, however, since we could still remain committed to the view that our best source of knowledge is whatever science allows us to accept. Quine rejects phenomenalism in favor of physicalism. Physicalism, however, like empiricism, is only contingently connected with his naturalism. Well-confirmed psychologies of the future may lead us to adopt some form of dualism. In that case, we would abandon physicalism but, again, not necessarily naturalism (Quine 1995b, 257). Is there anything that would lead us to abandon naturalism? For Quine, I take it, such rejection is possible if we came to endorse a methodology for discovering truth that rejected the hypothetico-deductive method that Quine sees as the hallmark of traditional science. In The Pursuit of Truth, Quine notes that testing hypotheses through predictions defines the ‘science game’

PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY

183

— à la Wittgenstein. This is not, for Quine, a norm of science but rather a defining characteristic of what it means to do science. To abandon predictivity as the test for hypotheses would be to play some other game (Quine 1990, 20). There is no guarantee that future truth seekers will not come to embrace methodologies that are widely at variance with currently accepted scientific practices. It is idle to speculate whether we would call such developments “science” or not. If not, then “naturalism” would fall by the wayside or re-emerge transformed by the adoption of new modes of understanding. So much for future speculations. Naturalism as we are to understand it today encompasses empiricism and physicalism as well as the methodological conviction that the scientific method is our surest guide to truth. What then of epistemology? In 1981, Quine wrote that “Naturalism does not repudiate epistemology, but assimilates it to empirical psychology” (Quine 1981, 72). The problem of epistemology becomes that of constructing a scientific understanding of how knowing organisms transform the meager information supplied by the evidence of the senses into scientific knowledge. The result will be a “scientific epistemology,” that is, a conjectural understanding of the formation of knowledge that draws upon the results of physics, physiology, and psychology. For the scientific epistemologist, the epistemological question becomes “how [is it] we human animals can have managed to arrive at science from such limited information [as is provided by the “irritations of our surfaces’ that constitute the ultimate source of our information about the world]” (Quine 1981, 72). Note that this can be parsed as a phylogenetic question or a proximate ontogenetic question. On the one hand, we want to know about the evolution of our human capacities to acquire and process information. On the other hand, and this is Quine’s central concern, we want to know how an individual, endowed with those inherited capacities, comes to acquire knowledge of the immediate circumstances of the world about her. Thus, Quine maintains, “I remain occupied — with what has been central to traditional epistemology, namely the relation of science to its sensory data” (Quine 1990, 19). The traditional concern is there but the goal is different. On Quine’s view, the tradition was plagued by what he called the “Cartesian dream” or the attempt to provide a foundation for science that is “firmer” than that which science itself can provide (Quine 1990, 19). Insofar as scientific epistemologists are constrained by the results of fallible science, the traditional Cartesian goal of finding a secure and certain foundation for science needs to be abandoned. “The naturalization of epistemology, as I have been sketching it, is both a limitation and a liberation. The old quest for a foundation for natural science, firmer than science itself, is abandoned: that much is the limitation. The liberation is free access to the resources of natural science, without fear of circularity. The naturalistic epistemologist

184

MICHAEL BRADIE

settles for what he can learn about the strategy, logic, and mechanics by which our elaborate theory of the physical world is in fact projected, or might be, or should be, from just that amorphous neural intake” (Quine 1995b, 256). One of the recurring complaints about Quine’s reconceptualization of epistemology is that it fails to do justice to questions of normativity. Quine denies this. On his view, what he calls “theoretical” epistemology becomes a “chapter of theoretical science”. Normative epistemology becomes a chapter of applied science or engineering, “the technology of anticipating sensory stimulation” (Quine 1990, 19). We shall return for a more detailed look at the question of norms in section 5 below.

2. Quine and Evolutionary Epistemology Quine does not say much about the phylogenetic question. He has, however, often remarked about the relevance of Darwin and the theory of natural selection to the problem of the phylogenetic emergence of cognitive structures in the human lineage (See, e.g., Quine 1969, 1975, 1981, 1995, 1996; Hahn and Schilpp 1998). Phylogenetic questions about the origin and development of our cognitive capacities, on the one hand, and the growth of scientific knowledge, on the other, are the province of evolutionary epistemology (Campbell 1974a; Bradie 1986, 1994). In 1980, Quine remarked that he liked Donald Campbell’s version of evolutionary epistemology (Hahn and Schilpp 1998, 732; see Quine 1968, 90). For Quine, the origins of knowledge are rooted in the animal capacities that enable us to make our way around the world. “Why should nature, however lawful, match up at all with the dog’s subjective similarity ratings? Here, at its most primitive, is the question ‘Why is science so successful?’” (Quine 1975, 70). We are, at heart, inductive creatures. As individuals we learn from experience. Over generations, lineages evolve improved capacities to cope with their experiences. Those that do not, suffer the consequences. “The survival value of primitive induction is anticipation of something edible, or of some creature by which one might be eaten. Thus it is that natural selection has endowed us with standards of perceptual similarity that mesh pretty well with natural trends, affording us better than random success in our expectations” (Quine 1995a, 19f). But science, the generation of testable hypotheses with predictive value, is just the perfecting of these natural endowments. “Prediction is verbalized expectation. Conditional expectation, when correct, has survival value. Natural selection has accordingly favored innate standards of perceptual similarity that have harmonized with trends in our environment. Natural science, finally, is conditional expectation hypertrophied” (Quine 1995b, 256).

PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY

185

With respect to the reliability of induction, Quine notes that “there is some encouragement in Darwin. If people’s innate spacing of qualities is a gene-linked trait, then the spacing that has made for the most successful inductions will have tended to predominate through natural selection. Creatures inveterately wrong in their inductions have a pathetic but praiseworthy tendency to die before reproducing their kind” (Quine 1969, 126). In Word and Object, Quine was concerned with the problem of determining the correct translation of observation sentences. Appeals to ‘sameness of meaning’ were eschewed. Quine settles, he tells us, for ‘sameness of stimulus meaning.’ But, this generates a problem since we cannot be assured that the receptors and neural networks of various observers exhibit evolved diversity. So, the problem becomes: “How does the mere sameness of the distal cause, the jointly observed object, prevail over the diversity of the proximal segments of the causal chains, inside the two observers, and still issue in agreeing response?” (Quine 1996,160). Quine’s solution is to appeal to a ‘pre-established harmony between perceptual similarity and the environment’ engendered by natural selection. This is in close agreement with Campbell’s general view that natural selection, working on cognitive capacities, helps “shape” our sensory organs, our perceptual apparatus, and ultimately our scientific conjectures to our environment (Campbell 1974b; Bradie 2001). Natural selection can be invoked to explain why we are successful inductive agents but can it provide a justification for our reliance on induction? Quine demurs. “I am not appealing to Darwinian biology to justify induction. This would be circular, since biological knowledge depends on induction. Rather I am granting the efficacy of induction, and the observing that Darwinian biology, if true, helps explain why induction is as efficacious as it is” (Quine 1975, 70). It is remarks such as these that have reinforced the view of his critics that whatever “naturalized epistemology” is supposed to be, it isn’t “epistemology”. For traditional epistemology is centrally concerned with questions of justification and the justification of induction is one of the central issues in the epistemology of science, traditionally understood. Quine’s solution to this problem, I shall suggest below, relies on his particular tweaking of the tradition.

3. Quine and the Tradition When Donald Campbell put forward his case for developing an evolutionary epistemology, he characterized it as “descriptive epistemology” because the focus was on describing the biological and psychological mechanisms and processes by means of which we acquire, evaluate and develop knowledge. The traditional focus of epistemology was deemed “prescriptive” because of its focus on grounds and norms. Quine’s naturalized epistemology would

186

MICHAEL BRADIE

appear to be of the descriptive genre. The question that arises is whether such deserves the name “epistemology”. Are descriptive epistemologies of the sort advocated by Campbell and Quine a continuation of the tradition or a change in subject matter? As we shall see below, Quine’s view is that norms can find a home in naturalized epistemologies. So, even if one argued, as many do, that the business of epistemology is the investigation and elaboration of epistemic norms, naturalized epistemologies cannot be ruled out of court, without further consideration, on the grounds that they ignore questions of normativity. The relation between descriptive or naturalized epistemologies and the tradition can be viewed in one of three ways. (1) Descriptive epistemologies might be construed as competitors to traditional epistemologies. On this view, it is assumed there is a core set of issues and questions to which both offer alternative solutions. Those who despair of integrating normative concerns into a naturalistic framework reject this as a viable option. It is not one that Quine would endorse in any case. Part of his objection to the tradition is an objection to the framework that lends itself to certain of those core questions. Abandoning the tradition means abandoning some questions and reformulating others (Dewey, 1910). (2) Descriptive epistemologies might be construed as complementary to traditional epistemologies. On this view, each has its proper province and the two together tell us all anyone would want to know about human knowledge. The traditional approach focuses on norms and justifications. The descriptive approach focuses on the empirical mechanisms and processes. This appears to be the Campbell’s early view (Campbell 1974a). It is not a view that would appeal to Quine. Although it sees the tradition as incomplete, it endorses the traditional issues and the need to resolve them. Quine was more inclined to see the tradition as, at best, misguided, and, at worst, corrupt. (3) Descriptive epistemologies might be construed as a successor to the tradition. Many defenders of naturalized epistemologies endorse this reading (e.g., Quine 1960, 1969, Davidson 1973, Dennett 1976, Harman 1982, Kornblith 1985, Hull 1988, Dewey 1910). This view sees the traditional issues as arising from a philosophical view of the world that sees epistemology as “first philosophy” whose aim is to provide secure foundations for science that, as Quine puts it, are firmer than those provided by science itself. In addition, these philosophical speculations are not to draw upon the results of science on pain of circularity. This vision of the relationship between science and philosophy provides the framework for what Dewey called “the search for certainty.” Naturalism rejects this picture and with it the sharp division between matters scientific and matters philosophical. Whether what emerges from the ashes of the tradition deserves to be called “epistemology” is a matter of some dispute (see Rorty 1979, Kim 1988, Quine 1969).

PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY

187

In “Epistemology Naturalized,” Quine quite clearly took what he was doing to be epistemology, albeit in a transformed sense. The essay opens with the proclamation that “Epistemology is concerned with the foundations of science” (Quine 1969, 69). The question is: where are we to find the “foundations”? The classical program, which Quine rejects, sought to ground science in sensations. This program had two sides, a “conceptual” side and a “doctrinal” side. The conceptual side involved the attempt to reduce or define physical objects in terms of sense-data. That side led from Hume through Russell’s Our Knowledge of the External World to Carnap’s 1928 Der logische Aufbau der Welt. It failed and was abandoned. The doctrinal side involved the attempt to prove or establish with certainty the truths of empirical science. Hume’s worries about induction were an early warning sign that something here was amiss. While Quine saw some progress on the first front, the inductive worry, he thought, was still just about where Hume had left it (Quine 1969, 74). The classical program had to be abandoned. Is this, then, the end of epistemology? Not necessarily. Epistemology goes on but “in a new setting and [with] a clarified status. Epistemology, or something like it, simply falls into place as a chapter of psychology and hence of natural science (Quine 1969, 82). Quine’s solution was not to foreswear foundations but rather to seek them within science itself as far as that is possible. In 1995, he put the point this way: “Where naturalistic renunciation shows itself most clearly and significantly is in naturalistic epistemology. Various epistemologists, from Descartes to Carnap, had sought a foundation for natural science in mental entities, the flux of raw sense data. It was as if we might first fashion a self-sufficient and infallible lore of sense-data, innocent of reference to physical things, and then build our theory of the external world somehow on that finished foundation. The naturalistic epistemologist dismisses this dream of prior sense-datum language, arguing that the positing of physical things is itself our indispensable tool for organizing and remembering what is otherwise, in James’ words, a ‘blooming, buzzing confusion” (Quine 1995b, 254). Does this mean that Quine does not see the justification of induction as a problem? Well, yes and no. No, if we are looking for a justification of induction that does not rely on the results of science. Yes, if we are satisfied with a more limited result. We take induction for granted and then note that Darwinian theory helps account for why it works so well. But suppose that in the future it begins not to work so well. Then we must be prepared to give an account using psychology, Darwinian theory or whatever else we can muster to account for that fact. The point, I take it, is that we never escape the circle of using some bits and pieces of our knowledge to give accounts of other bits and pieces. This is the game of ‘naturalism’. Someone who demands more of us wants to play a different game. We may question what the point of such a game might be and whether or in what way anything we

188

MICHAEL BRADIE

accomplish in that sphere contributes to what we do when we play the ‘game of science’ (see Hookway, 1994). So, in addition to giving up the search for a secure foundation for science that goes beyond the bounds of science, we must give up, or transform, some of the traditional questions. The result, Quine conceded, is “a far cry” from the tradition but not “a gratuitous change of subject matter” (Quine 1973, 3). The mature, and presumably wiser, Quine maintained that “I am of that large minority or small majority who repudiates the Cartesian dream of a foundation for scientific certainty firmer than scientific method itself. But I remain occupied, ...with what has been central to traditional epistemology, namely the relation of science to its sensory data. I approach it as an input-output relation within flesh-and-blood denizens of an antecedently acknowledged external world [short circuiting Russell’s worries in OKEW], a relation open to inquiry as a chapter of the science of that world. To emphasize my dissociation from the Cartesian dream, I have written of neural receptors and their stimulation rather than of sense and sensibilia. I call the pursuit naturalized epistemology, but I have no quarrel with traditionalists who protest my retention of the latter word. I agree with them that repudiation of the Cartesian dream is no minor deviation” (Quine 1990, 19; see Koppelberg 2000).

4. Quine and the Skeptic Quine’s naturalized epistemology is of a piece with natural science. As such, it is “contained” within science. But, it “contains” science as well because the transformed project of epistemology is to account for how human beings come to know what they do. The result, said Quine, is a “reciprocal containment” (Quine 1969, 83) This has led critics to charge that Quine’s position involves a “vicious circle” and leaves it with no recourse against the skeptic. Quine admits the circularity but rejects the claim that it is vicious. We employ the tools and results of science to construct an epistemological apparatus that can then, in turn, be used to criticize and correct scientific practice. As epistemologists, we become “busy sailors” on Neurath’s raft (Quine 1969, 84). Can such a naturalized epistemology answer the skeptic? Quine’s answer and the answer of any reasonable naturalist has to be, again, yes and no. Let me explain. Skepticism in traditional epistemology rests on two assumptions, both of which are undermined by the transformed epistemology that has risen in its place. The first is that the primary focus of the tradition is first person knowledge. But this is just a reflection of the traditional problem of accounting for knowledge of the world in terms of sensory data. Starting from this point leads to the Cartesian “veil of perception” which we may well despair of transcending. The starting point of the

PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY

189

scientific epistemologist is what Campbell has called “the epistemology of the other one” (Campbell 1974b, 141). As Quine put it, “We are studying how the human subject of our study [not ourselves] posits bodies and projects his physics from his data, and [then] we appreciate that our position in the world is just like his” (Quine 1969, 83). Of course, we could be mistaken about this latter point, but then the corrective is to apply more psychology, physiology, physics or the like. This leads to the second assumption. Skeptics can be classified as “global” or “local.” Global skeptics are prone to consider our entire knowledge corpus at risk. With no solid foundations for our knowledge and with no guarantee that the future will be like the past, we can be certain of nothing. The shadow of the tradition can clearly be seen. Such skeptics are hostages to Cartesian dreams. Disavowing such dreams undermines the rationale for such global worries. Such skeptics must be left at the crossroads, for there is nothing that will satisfy them. The alternative is the take the best we have, which, on Quine’s view, are the results of science and make of it what we will. Of course, as Quine notes, we are still in the “Humean predicament,” or at least some transformed version of it. For Hume, as an inheritor of the tradition, was a global skeptic. Scientific epistemologists, on the other hand, embrace the fallibilism that is the inheritance of the scientific revolution. Nothing is certain but no one expects that it should be. Any given hypothesis, theory or claim can be mistaken but not the whole corpus at once. The global skeptic argues from illusions that arise from intersubjective comparisons and errors in the first place. The naturalist response is to block this “transcendental” move and refuse to be seduced behind the “veil.” We are still in the in the “Humean predicament” but only locally so. From a God’s eye point of view, the worry that the knowers are using appropriate methods or have come to appropriate conclusions is a real and biting problem. When that image is discarded, as it gradually came to be with the rise of modern science, global skepticism loses its purchase. What we are left with is a natural universe in which some parts struggle to understand themselves and the other parts. Of course, we can be mistaken about we take ourselves to know. We can be mistaken about particular facts or the efficacy of particular methods but it no longer makes sense to worry about whether we are globally deceived. We are left with a healthy skepticism about particular results or methods that we can correct or adjudicate in the light of other particular results or methods.

5. Naturalism and Norms We turn at last to the vexing question of the status of norms in naturalized epistemologies. Quine is quite clear that his version of naturalized

190

MICHAEL BRADIE

epistemology countenances norms. Just as epistemology becomes a part of science, the normative bits turn out to be transformed into “chapters of engineering” (Quine 1990, 19). Norms emerge, in medias res, as technical maxims in the course of pursuing science. As technological rules, they embody a form of “instrumental normativity.” Christopher Hookway (Hookway 2000, 37) recently raised the question of how ‘typical’ is Quine’s treatment of norms as engineering problems. In other words, when Quine claims that his version of naturalized epistemology is akin to the normative tradition because it too includes norms, is he doing justice to the traditional normative issues? Hookway takes it that normative epistemological naturalism should be construed as the view that “questions of epistemic normativity raise no additional problems once we understand the nature of instrumental norms” (Hookway 2000, 37). This said, we can discern four different types of norm in Quine’s epistemology. First, there are what we may call “philosophical norms.” As an example, Quine notes that the ‘primary norm’ of naturalized epistemology is the ‘watchword of empiricism’: “There is nothing in the mind that is not first in the senses” (Quine 1990, 19). I call such norms “philosophical” guardedly since they do not share the cachet of traditional philosophical norms that lie beyond the scrutiny of science. Any norms, including the endorsement of empiricism itself, are subject to revision in the light of future developments. Any such future developments, Quine thinks, are at this time extravagant suppositions, at least with respect to our giving up empiricism. Given everything we know now, it is implausible to suppose that we might abandon empiricism in the light of evidence indicating the reality of telepathy, revelation or other modes of knowledge acquisition that would result in mental contents that bypassed the usual sensory gates. However, should evidence for such modes of acquiring information be forthcoming, we would have to rethink our commitment to empiricism. Second, there are what we may call “norms of discovery”. “Naturalized epistemology on its normative side is occupied with heuristics generally — with the whole strategy of rational conjecture in the framing of scientific hypotheses” (Quine 1990, 20). As specific examples of such norms, Quine cites “Be conservative” and “Look for simple theories” (Quine 1995b, 258). Third, there are what we may call “norms of justification”. “Normative epistemology is the art and technology not only of science, in the austere sense of the word, but of rational belief generally...Podiatry, appendectomy, and the surgical repair of hernias are technological correctives of bad side effects of natural selection, and such also in essence is normative epistemology in its correcting and refining of our innate propensities to expectation by induction...Correction of...[errors such as the ‘gambler’s fallacy’ are]...the therapeutic side of statistical theory, a substantial branch of applied mathematics that is part and parcel of normative epistemology” (Quine 1995a, 49f). There is, of course, no general analysis of crucial

PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY

191

justificational normative concepts such as ‘warrant’, but the statistical example alluded to by Quine indicates that he is happy with piecemeal and local determinations of warrant. This is consistent with Quine’s rejection of global skepticism in favor of local varieties. Fourth, particular norms emerge in the course of our scientific investigations. One such norm is “Be wary of astrologers, palmists, and other soothsayers” (Quine 1995b, 258). The force of this lies in the fact that such practitioners employ methods that remain uncertified and dubious according to the very best current science. As the science changes, what we should be wary of changes as well — we have a moveable feast of norms fashioned in accordance with the latest fashions of scientific thinking. So, one way in which norms are treated differently from the way they are treated by the tradition is that they are changeable. Regardless of whether the tradition is read as foundationalist or not, one might argue that classical approaches to questions of ‘justification,’ for example, are intended to be ‘once and for all’ however changing and ephemeral particular analyses might be. This may be another clue to the mystery of the curious way in which Quine does not seem to address the perennial worries of his critics from the tradition. They keep looking for eternal standards and attempts to formulate them. Quine ignores this quest in favor of mundane and prosaic norms derived from the best science along the lines of “Don’t try to fly by flapping your arms.” As I remarked above, one maxim that might be taken as a norm, “The test of a hypotheses is prediction or observation,” is not a norm on Quine’s view. Instead it is, he says “the name of the game”. To abandon this would be to abandon science. Moreover, Quine notes, “naturalism has no special claims on the principle, which is rather the crux of empiricism.” And, empiricism, as we have seen, like physicalism, although part and parcel of Quine’s vision of naturalism, can be decoupled from naturalism proper. Well, what then is naturalism? When Quine characterizes it, he seems to try to explain it by appealing either to physicalism or to empiricism or both but then he takes it away by insisting that the three views are independent (Quine 1995b). If we identify ‘naturalism ‘with’ remaining within the bounds of science,’ then we just deflect the problem from ‘what is naturalism?’ to ‘what counts as science?’ Science, as we know and love it, is the application of the hypothetico-deductive method to the formulation and testing of hypotheses by means of comparing predictions with observations. But, who knows whether this concept of “traditional science” will survive in the future any better than the concept of “traditional epistemology” has for us? Finally, Hookway (Hookway 2000, 40) raises the deep question: Can naturalized epistemology, in particular Quine’s version of it, give a naturalistic account of normativity as such? This echoes objections raised by McDowell and O’Hear, among others (McDowell 1994, O’Hear 1997). Can

192

MICHAEL BRADIE

a naturalistic account provide a model of human agency that makes “sense of how there can be responsible active users of norms at all”? Perhaps not, but even before we concede this point we need to settle what the limits of a pure naturalism are. So, what are the limits of naturalism? That is, what can and cannot be included in an account that professes to be naturalistic? Suppose we admit that a naturalistic account cannot fully explain “what it is to follow norms and be subject to normativity.” Does it follow that the naturalist has nothing to say about these questions? What would a ‘full explanation’ of normativity amount to? The first question is easily answered. Surely, appeals to Darwin and natural selection are relevant to our understanding of how norm following creatures came to be and how they manifest this behavior. Just as surely, this tells us something about “what it is to follow norms and be subject to normativity.” As for the second question, I leave that for another occasion.

6. Conclusion Quine’s naturalized epistemology is a logical outgrowth of his commitment to a world view in which science provides us with the best information about the world as well as the best means for generating norms for the gathering and assessing of that information. He has rejected the “Cartesian dream” of searching for a secure foundation for our knowledge that transcend the methods and procedures of science. He has replaced it with a “Quinean dream” of a self-contained, self-correcting mechanism for the generation and assessment of what we know. Epistemology has been transformed into a branch of natural science. As such, some of the traditional questions of epistemology are rendered irrelevant while others survive in a transformed guise. If Quine’s reading of the tradition is correct and he is right in his assumption that the central question of the tradition is “the relation of science [or knowledge in general] to its sensory data” then it is clear that Quine is continuing in the tradition. But the goals are more circumscribed and local. Norms emerge in the practice of science and can be handled by the very methods and procedures that they are used to evaluate. The result is a non-vicious circle. We use bits of pieces of parts of what we know to evaluate other bits and pieces. There is no escape from integrated corpus of fact, theories, procedures and norms. Each part contributes to the stability of the whole but no single part can be claimed to be unchangeable. Quine’s invocation of Neurath’s vivid image is particularly apt. The hallmark of modern science is its commitment to the fallibility of everything we take to be known. The hallmark of a Quinean version of naturalized epistemology is the recognition that the norms we employ and the analyses we provide of them are in the same boat. It is a bracing perspective that may not appeal to everyone but the alternative is to languish

PROSPECTS FOR A SCIENTIFIC EPISTEMOLOGY

193

in impossible dreams that, at present, we have no hope of seeing realized. Department of Philosophy Bowling Green State University, USA

REFERENCES

Bradie, M. “Assessing Evolutionary Epistemology”, Biology & Philosophy, 4, 1986, 401-59. Bradie, M. “Epistemology from an Evolutionary Point of View.” In Conceptual Issues in Evolutionary Biology, edited by Elliott Sober, 45375. Cambridge: The MIT Press, 1994. Bradie, M. “The Metaphysical Foundation of Campbell’s Selectionist Epistemology.” In Selection Theory and Social Construction: The Evolutionary Naturalistic Epistemology of Donald T. Campbell, edited by and Cecilia Heyes and David Hull. Albany: SUNY Press, 2001. Campbell, D. T. “Evolutionary Epistemology.” In The Philosophy of Karl Popper, edited by P. A. Schilpp, 413-63. LaSalle, IL: Open Court, 1974a. Campbell, D. T. “Unjustified Variation and Selective Retention in Scientific Discovery.” In Studies in the Philosophy of Biology, edited by F. J. Ayala and T. Dobzhansky, 139-61. London: Macmillan, 1974b. Davidson, Donald. “On the Very Idea of a Conceptual Scheme.” In Proceedings of the American Philosophical Association, 1973, 5-20. Dennett, Daniel. Brainstorms. Cambridge: The MIT Press, 1978. Dewey, John. The Influence of Darwin on Philosophy and Other Essays in Contemporary Thought. New York: Henry Holt & Co., 1910. Gibson, Roger. “Quine on Naturalism and Epistemology”, Erkenntnis, 27, 1987, 57-78. Gibson, Roger. “Quine on the Naturalizing of Epistemology.” In On Quine: New Essays, edited by Paolo Leonard and Marco Santambroglio, 89103. Cambridge: Cambridge University Press, 1995. Hahn, Lewis Edward, and Paul Arthur Schilpp (eds.). The Philosophy of W. V. Quine. Expanded Edition, The Library of Living Philosophers. Chicago: Open Court, 1998. Harman, G. “Metaphysical Realism and Moral Relativism”, The Journal of Philosophy, 79, 1982, 568-75. Hookway, Christopher. “Naturalism and Rationality.” In Quine: Naturalized Epistemology, Perceptual Knowledge and Ontology, edited by Lieven Decock and Leon Horsten, 35-55. Amsterdam: Rodolpi, 2000. Hookway, Christopher. “Naturalized Epistemology and Epistemic Evaluation.” Inquiry, 37, 1994, 465-85. Hull, David. Science as a Process: An Evolutionary Account of the Social

194

MICHAEL BRADIE

and Conceptual Development of Science. Chicago: University of Chicago Press, 1988. Kim, J. “What Is ‘Naturalized Epistemology’?”, Philosophical Perspectives 2. Epistemology, 1988, 381-405. Koppelberg, Dirk. “Quine and Davidson on the Structure of Empirical Knowledge.” In Quine: Naturalized Epistemology, Perceptual Knowledge and Ontology, edited by Lieven Decock and Leon Horsten, 101-19. Amsterdam: Rodopi, 2000. Kornblith, H., ed. Naturalizing Epistemology. Cambridge: The MIT Press, 1985. McDowell, John. Mind and World. Cambridge: Harvard University Press, 1994. O’Hear, Anthony. Beyond Evolution: Human Nature and the Limits of Evolutionary Explanation. Oxford: The Clarendon Press, 1997. Quine, W. V. From Stimulus to Science. Cambridge: Harvard University Press, 1995a. Quine, W. V. “Naturalism; or, Living within One’s Means”, Dialectica, 49, 1995b, 251-61. Quine, W. V. “The Nature of Natural Knowledge.” In Mind & Language, edited by S. Guttenplan, 68-81. Oxford: Oxford University Press, 1975. Quine, W. V. Ontological Relativity and Other Essays. New York: Columbia University Press, 1969. Quine, W. V. “Progress on Two Fronts”, The Journal of Philosophy, 93, 1996, 159-63. Quine, W. V. Pursuit of Truth. Cambridge, MA: Harvard University Press, 1990. Quine, W. V. “Reactions.” In On Quine: New Essays, edited by Paolo Leonardi and Marco Santambroglio, 347-61. Cambridge: Cambridge University Press, 1995c. Quine, W. V. The Roots of Reference. LaSalle: Open Court, 1973. Quine, W. V. Theories and Things. Cambridge: Harvard University Press, 1981. Rorty, R. “The Unnaturalness of Epistemology.” In Body, Mind and Method, edited by D.F. Gustafson and B. Tapscott, 77-92. Dordrecht: D. Reidel, 1979.

JOSEPH MARGOLIS

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM

I The following two dicta bear directly on the fortunes of psychologism: (1) if Platonism is true, psychologism is false; (2) whenever a principled disjunction between epistemological, ontological, and psychological analyses of the “mental” obtains, insoluble and unacceptable forms of skepticism follow; hence, no such distinction can be plausibly invoked by psychologism’s opponents. Optimists believe that psychologism can be attacked effectively without championing Platonism and that the unwanted disjunction can be avoided consistently by distinguishing between “mental” processes and the “laws of thought.” I take both countermoves to be doubtful — possibly demonstrably mistaken — but I do not believe that the best strategy for vindicating psychologism lies in the direction of proving Platonism false or proving that a principled disjunction would yield skepticism. There are better arguments: I suggest that ‘psychologism’ is no longer a central philosophical concern, not because it has won the day, but because the claims on which its supposed defeat were once postulated are no longer secure. It has lost its dialectical importance for the same reasons the original claims of Frege and Husserl have lost their importance. The trail is not altogether straightforward. Broadly speaking, psychologism is a theory about the concceptual and cognitive resources on which truth claims of any sort depend: it is a theory of what we should understand by the “mental” or “psychological” or by the structure and content of “thought” insofar as “thought” designates certain “subjective” resources assigned cognizing subjects (persons, selves) that are apt for grasping what is true of the independent world or of some putatively independent order of reality (arithmetic, for instance) that is said not to be part of the natural world. The resources in question may be no more than necessary for knowledge (as in being embedded in sensory experience) or may be sufficient for knowledge (as in grasping directly what is true of 195 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 195-214. © 2003 All rights reserved. Printed by Kluwer Academic Publishers, the Netherlands.

196

JOSEPH MARGOLIS

numbers). Psychologism, then, is the denial that there is any principled disjunction, epistemically and subjectively, between the cognizing source or scope or epistemic certainty imputed to distinctive “elements” or aspects of human cognition or, short of that, any determinate difference in scope or power between would-be elements that, though not explicitly disjoint, may be shown to contribute in different modular ways to what should count as knowledge. I put the point with some caution because, although Husserl (who is the arch opponent of psychologism), with Frege, argues in favor of a very strong disjunction between naturalistic and phenomenological resources of the epistemic kind (see Husserl 1960, Med. 5; 1970, p. 148), there are more recent phenomenologists (also Fregeans) who mean to defeat psychologism by leaner means — by disallowing, in fact, any strong disjunction of Husserl’s (or of any other cognate) sort. These phenomenologists may be said to concede a “weak” form of psychologism or to redefine psychologism (partly against Husserl and partly against what I have just laid out as the psychologistic thesis). Possibly the best and most straightforward of these phenomenologists is J.N. Mohanty, who offers the following recommendation in a broadly Husserlian spirit: The logical and the mental cannot be radically sundered completely from each other...our thoughts and theories are products of our acts of thinking ...all meaningfulness must lie in our being able to perform such operations ...The radical anti-psychologistic thinker is wrong in reifying thoughts into things, the psychologistic thinker erred in reifying thoughts into subjective ideas. Each saw one side of the truth; the noetic act and the objective Sinn are but two sides of a structure, separable only by abstraction. (1989, 8)

Mohanty’s strategy is to replace the general claim (differently formulated in Husserl and Frege — and, by extension, differently formulated in Kant and Cassirer) of assigning the logician and the empirical psychologist “different domains” of inquiry — hence, different epistemically pertinent resources, which, Mohanty says, yields both “a poor philosophy of mind and a poor philosophy of logic.” Mohanty wishes to “bring them closer, to ground [them, that is, mental processes and the norms of rational thought] in a philosophy of mind — without having to court the ruinous consequences of psychologism” (4). His strategy comes to this: admit the link between “the rules of thought” and mental “acts and operations,” but make the latter “dispensable” relative to isolating the “rules”; construe mental acts as possessing a “structure” ranging over the “content” of such acts, where discerning such structures answers to our discerning the rules mentioned; and, then, concede the “normative autonomy” of the strategy of analysis that discerns such rules — in effect, adopt a ‘transcendental psychology’ — functioning within the

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM

197

naturalistic terms of the mental or psychological. Mohanty calls such a policy “an enlightened psychologism” (1989, 4-5). I find this a heroic but completely indefensible maneuver. I shall, however, allow the demonstration of that finding to be inferred from a more general attack on antipsychologism. What is most important lies elsewhere. As I see matters, insisting on psychologism has at least two objectives: (a) subverting any claims of de re or de dicto or de cogitatione necessity and (b) subverting all forms of extensionalism or attempts to validate extensionalized strategies that are not subaltern to, or epistemically encumbered by, natural psychological resources (thinking, speaking) said to yield complex forms of intentionality. However, admitting all that entails additional qualifications. First, by dictum 2 psychologism cannot be cast as merely epistemic: it cannot fail to have “ontic,” “logical,” “semantic” import as well. (This goes against Mohanty’s view [see 1989, 2].) Second, psychologism need not be restricted to any one or another epistemology, say, some thing like British empiricism or Carnapian or Diltheyan Erlebnisse. (This goes against Mohanty again [see 1989, 4-5].) Third, there may be many reasons for attacking modal necessity and extensionalism, or particular epistemologies, that do not directly address the fate of psychologism. For example, one may be antipsychologistic and yet mean to put in doubt the viability of any inclusive extensionalism. (In Mohanty’s view, that might actually be the upshot of Husserl’s own thesis.1) Or, one may intend to “psychologize” logic and knowledge and yet support some form of extensionalism. (This is, precisely, Quine’s objective [see 1969a, 1992].) Finally, antipsychologistic measures may come in so many different forms that they are not easily detected. We need to appreciate how protean both doctrines may be.

II Let me offer a specimen — a current Fregean thesis — to show you how difficult it is to spot antipsychologism even when it stares you in the face. Consider a plausible recommendation of Michael Dummett: it hardly appears to engage psychologism at all. Apart from invoking mathematical intuitionism, it seeks to secure bivalence, though not excluded middle (or, it seeks to replace the “classical” logic in which bivalence entails excluded middle) and to secure by “semantic” means an extensional reading of the preconditions of truth (of what is “assertorically” uttered — in effect, the conditions of meaning) along the lines of the necessity of adhering to tertium non datur (the restoration of bivalence): The thesis of the priority of language over thought in the order of explanation is, obviously, important in itself; but its acceptance or rejection makes

198

JOSEPH MARGOLIS comparatively little difference to overall philosophical strategy, because doctrines concerning meaning can be fairly readily transposed into doctrines concerning thought, and vice versa. An analysis of the logical structure of sentences can be converted into a parallel analysis of the structure of thought. That is why Frege was able to claim that the structure of sentences reflected the structure of thought. Thus the thesis, in the philosophy of language, that the meaning of a sentence is determined by the conditions for it to be true, can at once be transposed into the thesis, in the philosophy of thought, that the content of a thought is determined by the conditions for it to be true: in either mode, arguments for and against the thesis are to a large extent the same. (1991b, 3)

Dummett takes Frege to have been the first to have made explicit the connection between meaning and truth. Fine. He goes on, however, to construe the indifference of prioritizing logic and thought vis-à-vis one another (in accord with Frege’s discovery) as signifying (in the Fregean manner) the priority of semantics over metaphysics. This is the result of a deep non sequitur ultimately tethered to Frege’s own antipsychologism (which Dummett shares), but it does not surface explicitly. In fact, Dummett’s argument focuses on the relative unmanageability of metaphysics in comparison with semantics. Nevertheless, the argument is, at bottom, antipsychologistic. For one thing, Dummett nowhere shows that ‘semantics’ (by which he means the formal semantics of sentences, ultimately rendered in bivalent and extentionalist terms) is not itself a form of metaphysics “by other means.” Second, he nowhere shows that the syntax of natural language sentences possessing semantic content (or the “think they represent) can be decided prior to and independent of deciding how “the world is” and what our cognitive access to it is; hence, he has not shown that its analysis yields (or must yield) along bivalent and extensionalist lines. Third, he nowhere shows that the analysis of the conditions of meaning or the conditions of truth can be given in a way that is actually prior to our analysis of metaphysics or (to use Mohanty’s term) our analysis of the ‘mental processes’ that enter into epistemic claims. Finally, he nowhere shows that the rules (or laws) of thought (or logic) are so obviously separable (or distinct) from our natural mental processes that they could not be deemed to be projections from those same processes. Concede only that there is no principled disjunction between the “subjective” and the “objective” (without invoking any particular epistemology — for instance, without invoking any seventeenth- or eighteenth century rationalism or empiricism), concede only that cognition and the cognized world are social constructions of some sort (in the sense of realism, not of reality), and you see at once that Dummett’s presumption could not possibly escape “psychologistic” difficulties.2 This, of course, is the true import of dictum 2. Dummett complains that to begin “top down,” to “resolve the metaphysical problem first, then...derive from the solution to it the correct model of meaning, and the appropriate notion of truth, for the sentences in

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM

199

dispute, and [then] deduce the logic we ought to accept as governing them” runs into the difficulty that “we do not know how to resolve these disputes.” By contrast: To approach these problems from the bottom up is to start with the disagreement between the realist and the various brands of anti-realist over the correct model of meaning for statements of the disputed class, ignoring the metaphysical problems at the outset. We are dealing, after all, with forms of statements which we actually employ and which, with the exception of the statements of math ics and scientific theory, are familiar to all human beings. Their meanings are already known to us. No hidden power confers these meanings on them: they mean what they mean in virtue of the way we use them, and of nothing else. (1991b, 12-13)

But you have to read these sensible-sounding remarks in accord with Dummett’s deeper commitment, for Dummett subscribes to Frege’s thesis that “the laws of logic...are not the laws of nature but the laws of the laws of nature. It makes no sense to observe the world to discover whether or not it obeys some given logical law...it is our thinking about reality that obeys such a law or flouts it” (1-2). This is antipsychologism. But where is the supporting evidence? I cannot see that it is easier to determine whether our thinking about nature conforms with the “laws of thought” or whether nature itself (about which we think) conforms to the laws of logic. I cannot see any difference between these two options. Also, if we mean to adhere to the actual processes of thinking, then it is not clear that any straightforward realism (of Dummett’s sort) will work or that our theory of meaning must be squared with a realist theory of truth (a fortiori, with an exceptionless bivalent or extensional logic — for instance, one committed to excluded middle or at least tertium non datur). I cannot see that, as far as defeating psychologism is concerned, there is much difference between a canny Husserlian like Mohanty and a canny Fregean like Dummett. I cannot see how either’s claim may be vindicated. Each assumes a fixity that he has not secured. Psychologism is the custodian of the expose of all such maneuvers. Broadly speaking, both Mohanty and Dummett claim (with out evidence) that there must be a certain invariance or universal structure to reason or thought or meaning and that we may then infer (without further labor) that the cognitive competence to discern such structures must be distinct from their embedding forms of “natural” thought (the “mental”) that can yield no more than contingent regularities. Simplicity itself!

III Now, it’s rather easy to formulate a coherent and plausible argument that would entail psychologism; and it is even possible to offer reasons in its

200

JOSEPH MARGOLIS

favor that, since they are likely to be promising, make any form of antipsychologism doubtful on its face. The argument would touch on the following: (i) that self-consciousness is the paradigm of consciousness, that is, the ability to report, reflexively (linguistically), what one is conscious of; all other attributions of consciousness will, as a consequence, be anthropomorphized in terms of the linguistic model; (ii) that it is in virtue of enculturation that human beings first acquire the capacity for selfconsciousness; they are, in that sense, artifacts of social history; (iii) that there is no algorithmic or criterial resolution of the linked puzzles of reference and predication; whatever evidentiary grounds we may invoke here will themselves be grounded in the holistic (noncriterial) tolerance of the linguistic practices we acquire; (iv) that our second-order capacity to legitimate first-order powers (as of reference and predication) arise from the same social sources as the other and are continuous with them. Ergo, the “laws of thought” are idealized artifacts of the same mental gifts by which we acquire our first-order linguistic powers. Ergo, there is no alternative to psychologism. There is enough that is controversial here to warn us against drawing too hasty a conclusion. But the linchpin of every successful argument against antipsychologism will, I believe, come to terms with claim (iii). If that is true, then surely the two dicta with which I began will prove to be much more strategically placed than their mere mention suggests, for the analysis of predication (a fortiori, reference and denotation) will preclude Platonism; hence it will encumber thought (and speech) in such a way that the would-be laws or rules of logic could not be but an idealization from, and regularization of, the informal (not criterial) consensual play of the linguistic practices of our society. Simply put: If thinking is historicized — the emergent social artifact of enculturation — then, from the side of Frege and Husserl, psychologism will prove unavoidable; yet, to admit all that would (pace Mohanty and Putnam [1994b]3) not be to fall back to any eighteenth-century subjectivism or representationalism. It would be reasonable to hold (as Thomas Kuhn [1970] and Hans Georg Gadamer [1975] make clear) that the paradigm forms of the mental (not, of course, all forms of mental life) are already social or socialized or collective — in the sense of having internalized the traditional practices of language (and what that entails). (Husserl, I might add, tried — unconvincingly — to separate thought and its “laws” from the enncumbrances of contingent natural language.4 Hardly any contemporary antipsychologist would willingly follow Husserl in this. Certainly, no Fregean would.) Admit the historical contingency of our conceptual categories, no matter how plausibly regularized they prove to be: psychologism will be vindicated. Admit some significant conceptual divergence — even a measure of incommensurability — and a tolerance for vindicating apparently conflicting truth claims: you call into question the necessary

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 201

ubiquity of bivalence. Deny any principled disjunction between the analysis of the powers of cognizing agents and the analysis of the structures of the cognized world: you undermine the supposed ground for any exceptionless bivalence and exceptionless extensional treatment of the conditions of meaning of ordinary sentences. Admit that the paradigm of the mental is modeled, one way or another, on considerations such as those just mentioned: you preclude every possibility that the would-be laws of thought or reason or logic or meaning can be anything but constructed from some survey of our “natural” (our “second-natured”) mental or psychological processes. But that is psychologism! To clinch the argument, I claim that neither Mohanty nor Dummett (committed spokesmen for Husserlian and Fregean antipsychologisms) have successfully met this challenge — or could. From what I have already reported about Mohanty’s view, you can see that his project does not so much deny psychologism as convey the expectation that, within its terms, the invariant laws of thought can still be asymptotically approached. I say that is impossible, if reference and predication behave as I suggest. (I shall come to the argument shortly.) In any case, there is (here) little difference be tween the Husserlian and the Fregean arguments. What there is weakens the Husserlian (the more developed) alternative, for Mohanty very plausibly objects to the “Fregean reading of Husserl that Dreyfus offers, on the grounds that Dreyfus fails to recognize that “the rules [of the Husserlian noema] are...semantic as well as syntactical,” that is, such that the syntax of the noema cannot be disjoined from, or as signed priority over, its “meaning” (1989, 5).5 But then, what’s good for the noema is good for the processes of ordinary thinking and for the relationship between phenomenological and naturalistic reflection. Mohanty does not address the connection explicitly, unless (by default) to affirm his confidence that the isolation of the laws of noematic thought are confirmably distinct from the regularities of the other. Ultimately, it may be that the charge against psychologism is simply that psychologism is tantamount to relativism and that relativism is (necessarily) self-contradictory or paradoxical in self-referential terms. Construed thus, Mohanty’s argument signifies that a failure to retrieve the laws of thought is tantamount to adhering to a self-defeating relativism. But that is itself a contested claim, which is no more than tangentially linked to the fate of psychologism.

IV A tighter connection between the fate of psychologism and the recovery of the laws of thought is affirmed by Dummett. Dummett’s argument also fails, I claim, but its lacunae are more readily detected because they are more

202

JOSEPH MARGOLIS

explicit and more nearly in accord with the general drift of Anglo-American analytic philosophy. Dummett proceeds by considering how to recover (for his own purposes) Tarski’s so-called “(T) schema” (Tarski 1983, notably, the Postscript). He is thinking primarily of natural language discourse, in which, that is, “the metalanguage, the language in which the statement [of the truth conditions of an uttered sentence] is given, is [itself] an extension of the object-language.” The principle underlying the (T) schema, he says, affirms that “to assert a sentence is tantamount to asserting that the condition for it to be true obtains” (Dummett 1991b, 52). Fine. From this point on, however, Dummett’s theory becomes quite murky with regard to the fortunes of psychologism: here, at least, the question of psychologism does not take explicit form (but see Dummett 1991a, chapters 13, 14). You must bear in mind that Tarski’s schema is not philosophically “neutral,” both in the obvious sense that it is uncompromisingly extensionalist in providing for the truth conditions (the conditions of “meaning,” in Dummett’s view) of any “assertorically” uttered sentences and in the sense that Tarski himself became increasingly skeptical about its applicability to natural language discourse6 (also, of course, in the sense that it is not compatible with psychologism). Dummett offers an ingenious alternative: To claim that a Tarski-like theory of truth can be seen to be the right characterization of the condition for a sentence to be true, for the purpose for which we need the notion of truth in a meaning theory for an actual language (in advance of deciding wheth er that language has a classical logic), is to claim that such a theory of truth is neutral as between different logics” (1991b, 64; emphasis added).7 Dummett is prepared to concede that: The principle of bivalence cannot hold in any semantics for an intuitionistic language [as perhaps also in other cases that Dummett does not consider], but that is not enough to show that under the appropriate notion of truth there will be counter-examples to the (T) schema, since, if a sentence is false just in case its negation is true, and if truth commutes negation, it would be contradictory to say that there was a sentence which was neither true nor false. What matters is not whether bivalence holds but whether there are sentences that violate it. (64)

This concession obviously violates dictum 2, offered at the start of this essay: that is precisely the point of the italicized expression in advance of. The champions of psychologism would never allow it: it is, in effect, the characteristic theme of antipsychologism in the English language tradition. Consider two objections to the thesis. For one, although, on Dummett’s admission, the principle of excluded middle (the “classical” logic) may be abandoned on intuitionistic grounds, it nevertheless remains true that once a statement’s truth conditions are determined (so that the statement is shown to be “decidable”), the principle of tertium non datur cannot be violated (see

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 203

Dummett 1978). Hence, the bipolar values, True and False, are restored with a vengeance. The rule of bivalent inference (and conformable inferences) proves necessary and remains entirely uncontested. If, however, the analysis of discourse about certain sectors of the world required or invoked, or was best represented by, replacing a bivalent logic with a many-valued logic — in particular, a “relativistic” logic, one that would replace True by a run of many values (reasonable, say) and thus countenance validating claims that on a bivalent logic but not now would yield contradictories or contraries — then Dummett’s position would be either false or a priori (or, in any case, unsupported). It would, of course, confirm the good sense of dictum 2 and it would threaten the commutation of negation. For the second objection, it is in some sectors of inquiry demonstrably more reasonable to prefer a relativistic logic to a bivalent logic. For example, in interpreting artworks and cultural phenomena — say, alternative interpretations of Hamlet — it looks as if we are prepared to argue that certain interpretations are false but that, congruent with admitted textual facts, we cannot preclude the possibility that contrary interpretations are jointly “valid” (let us say, plausible or reasonable) in spite of the fact that they cannot be jointly true.8 Truth and falsity do not behave symmetrically here. Hence, Dummett’s logic may be (reasonably) eclipsed. What I hold is that, if this argument is sustained, then it may be confirmed as well in the sciences, in philosophy at large, and elsewhere. In Dummett’s view, we should have to admit a “violation” of bivalence that we cannot show to be self-contradictory, illogical, contrary to reason, or anything of the sort. It’s for this reason that I suggest that the treatment of bivalence is part of the “closet” antipsychologism of the analytic tradition. This is the form in which the issue arises for such important figures as Carnap and Quine, though their connection with the psychologism issue is far from clear. In any case, we have here an argument purporting to show that bivalence is not demonstrably “inviolable” or necessarily true in any sense independent of our study of a particular empirical domain or our psychological practice with regard to it. But that is to defeat the antipsychologist. Turn, then, to Quine.

V W. V. Quine is, of course, the preeminent American philosopher of the second half of our century — bar none. He is also, effectively, the natural opponent and natural (anachronistic) target of Frege’s attack on psychologism. The curious thing is that Quine is not an altogether straightforward champion of psychologism, although he is of course the redoubtable champion of “naturalizing” epistemology and, in particular, of naturalizing epistemology by psychologizing it. The question remains, What

204

JOSEPH MARGOLIS

is the difference between psychologism and psychologizing or naturalizing epistemology? The simple answer is that psychologism is a species of naturalism, but naturalism does not entail psychologism. So it is entirely possible for Quine to advocate naturalism but not psychologism in spite of the fact that he does “psychologize” both epistemology and logic. If you admit this line of reasoning, certain embarrassments begin to surface. For one thing, it becomes clear that Frege may have been of two distinct minds about psychologism: on the one hand, he may have regarded any effort to “mentalize” (in effect, psychologize) logic as tantamount to psychologism; on the other, he may have treated the denial of the autonomous discovery of the invariant “laws of thought” as tantamount to psychologism (see, particularly, G. P. Baker and P. M. S. Hacker 1989). These are not equivalent doctrines and their difference makes it difficult to classify certain key discussants perspicuously. The American analysts — certainly Quine — do not see themselves as risking relativism, for instance, in psychologizing logic or epistemology (see, e.g., Goldman 1992); and they do not see themselves as favoring psychology in the “mentalist” sense Frege opposed — for instance, as being inaccessibly private, inaccessible to public scrutiny. Hence, although they are obviously committed to psychologism in some sense, they do not fit neatly. In effect, they oppose the Husserlian and Fregean lines for naturalistic, not psychologistic, reasons. Alternatively put, Quinean naturalizers” and “psychologizers” have, antecedently, already disposed of the privileged standing of modal and essentialist claims. Psychologism is one among many doctrines (like Mill’s, for instance) that attacks the supposed resources of the laws of thought; psychologizing is an epistemic adjustment, well after any such fact, that presumes that privilege has been dethroned. However, the arguments against necessary truth and the laws of thought need not be psychologistic in any important sense. By contrast, the phenomenologists — Mohanty, for instance (also Thomas Seebohm) — tend, rather surprisingly, to conflate naturalism and psychologism and the advocacy of relativism. There’s a considerable muddle. The question of whether, say, Quine or Carnap embraced psychologism cannot be entirely straightforward. There are two foci in Quine’s psychologizing efforts: one is provided in “Epistemology Naturalized”; the other in the famous “Two Dogmas” paper. The first, taken by itself, has nothing to offer psychologism, unless Quine’s opposition to “mentalism” is sufficient to tag Quine (which would be a mistake). To appreciate the proper force of the first, one should perhaps examine its application to the puzzle of reference and denotation and the interpretation it yields on the indeterminacy of translation issue. The second bears directly on Carnap’s treatment of logic and logical necessity and may be thought to betray Quine’s and Carnap’s stands on psychologism. But that

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 205

viewpoint turns out to be problematic. (I should mention that “Epistemology Naturalized” was originally subtitled “The Case for Psychologism” [reported by Willard 1989, 288].) To lessen the mystery, I will cite a number of distinctions that should be kept in mind. I think there can be little doubt that the best way to construe the attack on psychologism that Husserl and Frege share, a way that would illuminate the plausibility of regarding either Carnap or Quine or both as favorably or unfavorably disposed to psychologism, would be to read the attack thus: (a) the laws of thought or logic are cognizable in a way that is suitably distinct, or separable, from whatever may be said to describe directly the processes of natural language discourse or languaged thought or natural thought (if that be deemed to be distinct from the use of language or “mental” or “subjective” processes, as with Husserl); (b) the empirical or naturalistic description of mental or psychological processes may indeed benignly prompt, or even house, the discoveries intended in (a) but cannot form their evidentiary ground; and (c) the difference between the two lies in the discovery (or power to discern) the strict (or modal) necessity of the first and the contingency of the second.9 We have already (implicitly) seen that the force of (a)-(c) is considerably obscured or blunted by Mohanty’s and Dummett’s accounts. The question remains whether we can do better. Having been associated with the Marburg school, Carnap was still disposed to think in terms of the Kantian distinction between natural and transcendental reason (see Coffa 1991, chapter 11), but Quine was not; of course, in the “Two Dogmas” paper, Quine opposes Carnap, but not on either naturalistic or psychologistic grounds. Beyond this, I add two strategic remarks, one by Carnap and one by Quine, which I believe will orient us correctly to their respective intentions on the psychologism issue. I do not take what they say to be entirely transparent in this regard; that is their charm and also the mark of an important change in perception. In the foreword to The Logical Syntax of Language, Carnap emphatically affirms: “Philosophy is to be replaced by the logic of science...for the logic of science is nothing other than the logical syntax of the language of science” (1959, xiii; original in italics). For his part, Quine declares, equally emphatically, “Semantics is vitiated by a pernicious mentalism as long as we regard a man’s semantics as somehow determinate in his mind beyond what might be implicit in his dispositions to overt behavior. It is the very facts about meaning, not the entities meant, that must be construed in terms of behavior” (1969b, 27). These are eccentrically selected remarks, for Carnap eventually gave up, first of all, any purely syntactic account of the “logic” of science and, more interestingly, of analyticity (1956a, 1956b). In addition, Quine’s remark does not touch at all on the supposed status of the “laws” of thought or logic. While Carnap’s remark is a fair clue to his generally linguistic strategy in specifying the distinction between the analytic and the synthetic,

206

JOSEPH MARGOLIS

Quine’s remark shows unmistakably that his “behavioristic” account of meaning is, although certainly meant to “psychologize” meaning and knowledge, expressly opposed to any “mentalism” of the private sort Frege repudiated. These remarks, therefore, help to explain the sense in which Carnap was distinctly sanguine about the analytic/synthetic distinction well after conceding the inadequacy of a purely syntactic (semantically uninterpreted) strategy of analysis and well after Quine’s attack in the “Two Dogmas” paper and beyond (see Quine 1963 and Carnap 1963b). Carnap’s “liberalizing” of empiricism, particularly (along semantic lines) with regard to logical truth (in accord with Leibnizian and Wittgensteinian notions [see Carnap 1963a]), effectively demonstrates that if empiricism is taken to be tantamount to psychologism, then it is not at all obvious that the recovery of logical truth and the necessity of analytic truths by “empiricist” (or psychologistic) means must prove impossible (pace Quine). Of course, Carnap’s linguistic reading of empiricism plainly is not meant to be mentalistic (on his own reading). Clearly, Fregean objections do not apply in any straightforward way to Carnap. In fact, the conceptual connection between the search for the “laws of thought” (or logic) and the fate of psychologism, empiricism, naturalism, and allied doctrines becomes increasingly problematic and uncertain as we bring the issue to bear on Carnap and Quine. We begin to see, therefore, that Quine’s attack on Carnap and on logical truth and analyticity, as well as on “mentalism,” cannot not quite be reconciled with Frege’s and Husserl’s strategies. It seems that sheer apriorism or Platonism or privileged noetic competence is as much anathema in the recovery of necessary truth (if recovery be possible) as the advocacy of anything like the mentalism that originally provoked Frege. It begins to dawn on us that psychologism may no longer have a clear role to play in deciding the fate of necessary truths, regardless of which way the verdict may fall: privileged resources are no longer convincing; modal necessity must be earned. That is the new contest Carnap and Quine compel us to acknowledge. We glimpse how far Quine is from Frege and Husserl when, countering an attack on his own attack on Carnap and analyticity, Quine remarks: “I part company with the essentialist and the modal logician only when they accord these modes a place in the austere and enduring description of reality. Once we abstract from passing concerns of the moment, I can recognize only gradations of obviousness, gradations of consensus, gradations of platitiude, rather than any intelligible demarcation between the necessary and the contingent” (1986b, 94; see also Bohnert 1986). Quine dismisses any realist reading of necessity. I suppose one might say he expresses himself in broadly pragmatist terms; but to admit that much is not yet tantamount to holding that the objection to modal necessity in thought or logic or reality is a consequence of his adherence to pragmatism or

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 207

naturalism or some form of psychologizing knowledge. Quine simply finds no grounds for anything like Husserl’s or Frege’s epistemic confidence and, as a consequence, while adhering to the same broadly linguistic focus Carnap espoused, Quine finds no support for Carnap to continue to draw the contested disjunction in any way that would secure the modal claim. This line of reasoning strikes me as not at all psychologistic, though it is indeed linked to the psychologizing of logic and knowledge. The trick is this. Frege (and, of course, Husserl) wrongly supposed that necessity and strict universality regarding the laws of thought (or conceptual essences, in Husserl’s sense) were epistemically assured — incontrovertible, in fact. Believing that, they naturally believed that psychologism, that is, any treatment of the processes of thought that made necessity problematic or impossible — as psychologizing or naturalizing our mental life — must be philosophically indefensible. This same conviction motivates the canny strategies of Dummett (as a Fregean) and Mohanty (as a Husserlian). Neither is entirely willing to redeem the frontal claims of either Frege or Husserl, but they still insist on their benefits. In that sense, Dummett and Mohanty are very skillful transitional figures committed to the defense of a lost cause: Mohanty actually redeems psychologism, since he is so confident about the Husserlian thesis; and Dummett plainly favors a linguistic (or semantic) strategy that brings neo-positivism into accord with Frege’s original conviction, which then allows the apriorism of Dummett’s own doctrine of bivalence to go almost unnoticed. Once we move to Carnap and Quine, there is no place to hide: apriorism and epistemic privilege fall away; necessity and strict universality are put at risk; and, as a consequence, empiricism, naturalism, “psychologizing” are seen to be no more than options within a generous array of cognitive competences by which the fixed laws of thought are to be recovered — if, indeed, they can be recovered. Carnap thinks they can be, as far as logical truth and analyticity are concerned; Quine believes it is a lost cause. Nonetheless, psychologism no longer plays a spoiler role in the way Mohanty and Dummett insinuate. If you read the “Two Dogmas” paper carefully, you will find the following to be true: (a) In his critique of Carnap and, independently, in his analysis of analyticity, Quine nowhere invokes psychologism; he merely reports (and defends the finding) that there is no adequate criterial treatment of analyticity that would (or could) support our discerning the necessary truth of would-be analytic sentences. And, (b) his general attack on empiricism is an attack, compatible with (a), on any reliance on the wouldbe privilege of construing the meaning of “each meaningful statement” to be “equivalent to some logical construction upon terms which refer to immediate experience” (Quine 1953, 20). That means that Quine is doubly not an advocate of psychologism: first, because he attacks the doctrine of necessary truth on operative linguistic

208

JOSEPH MARGOLIS

grounds without reference to the mental; second, be cause he attacks what might otherwise be a form of psychologism (Carnap’s thesis), even if it is meant to recover (by such means) the laws of thought. Notice, incidentally, that it is not essential that we take sides for or against Quine or Carnap on the criterial treatment of analyticity (see Quine 1953, 33).10 The quarrel between them is not psychologistic at any point, even though they both proceed semantically. The result is the implicit defeat — well beyond mere jeopardy — of Husserlian and Fregean claims about the resources for discerning the laws of thought, particularly with respect to logical truth, analyticity, and essentialism; of course, it also places in direct jeopardy (perhaps defeats) Carnap’s would-be reclamation of an analyticity. Only after this labor is completed does Quine move to the psychologizing of epistemology, but that is already too late for the charge of psychologism to take hold (Quine 1953, 41). Besides, Quine reminds us, “mathematics reduces only to set theory and not to logic proper” and so reductionism (in the foundations of mathematics) “does not reveal the ground of mathematical knowledge, it does not show how mathematical certainty is possible” (1969a, 70). That, too, is not psychologism, though it defeats Frege — and, by extension, Carnap (see Quine 1969a, 77-8).

VI Apart from Carnap’s obvious regard for Frege’s pioneering efforts, the brief discussion of psychologism in the relatively late book Logical Foundations of Probability confirms that Carnap is, by and large, a Fregean on the issue in spite of his empiricism. In fact, departing from Russell, Carnap treats induction as “inductive logic,” hence not in terms of inductive hypotheses about the world, but as “the statement of the inductive relation between the hypothesis [any empirical hypothesis] and the evidence,” which Carnap holds is “purely logical” — hence, if true, analytic. He concludes that we may “regard the inductive method as valid without abandoning empiricism” (1950, 181). In fact, the shift in Carnap’s thinking from The Logical Syntax of Language to Meaning and Necessity does not really affect his general opposition to psychologism (see Carnap 1956a). Any doubts on this score are expressly dispelled by Carnap himself, with respect not only to deductive logic, but also to inductive logic. The following summary statement, matched by another regarding induction, leaves little ground for doubt: Logical relations, e.g., logical consequence, are (i) logical, i.e., nonfactual, based merely upon meanings, (ii) objective, i.e., not dependent upon anybody’s thinking about them. Most logicians treat them within their systems as objective relations, but, in spite of this, many characterize them in their

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 209 general preliminary remarks in subjectivist terms, e.g., with reference to actual thinking or believing. We call this discrepancy primitive psychologism in (deductive) logic. A qualified psychologism refers, not to actual, but to correct or rational thinking. This is usually meant in an objectivist sense; in this case, the reference to thinking is gratuitous. (1950, 37)

Carnap treats Dewey’s Logic rather tactfully as not being a form of psychologism “because there is no mixture of heterogeneous components” in it; that is, there is no mixture of Fregean-like laws of thought or reason and sociologically specified forms of thinking (42). This makes for a courtly reading, but the intention is plain enough. Carnap explicitly invokes Frege and Husserl in this regard (37-43). Plainly, ‘empiricism’ (in Carnap’s sense) is not a form of psychologism at all. Still, I cannot find an answer in Carnap to Quine’s objection to the recovery of necessity with respect to either logical truth or analyticity. This is not to endorse the whole of Quine’s argument (against Carnap) regarding the adequacy of enumerating instances of analytic truths. Surely, however, Quine is right in not losing sight of the absence of any principled demarcation of analyticity: We at present lack any tenable general suggestion, either rough and practical or remotely theoretical, as to what it is to be an analytic sentence. All we have [as in Carnap] are purported illustrations and claims that the truths of elementary logic, with or without the rest of mathematics, should be counted in...But when we would supplement the logical truths by the rest of the socalled analytic truths, true by essential predication, then we are no longer even able to say what we are talking about. The distinction itself, and not merely an epistemological question concerning it, is what is then in question. (Quine 1963, 403-4)

In Quine’s view, logical truth (setting aside the essential predications required for analyticity) can be managed well enough if we have in hand “a fixed logical notation,” that is, a logical vocabulary suited to elementary logic. But there is no such fixed vocabulary (Quine claims) to serve as an analogue with regard to analyticity (404). There’s the difficulty. I think Quine’s right; but if he is, then I cannot see that Husserl’s or Frege’s projects can be any more effectively recovered than Carnap’s. Of course, the reclamation of logical truth in ways internal to a given fixed logical vocabulary, in the sense Quine is willing to concede to Carnap, will not secure the necessary truth of the principle of inductive logic that Carnap favors and will not secure anything like Dummett’s recovery of bivalence construed in terms of tertium non datur or Husserlian essences or Frege’s principle of identity or the principle of noncontradiction in any form thought to depart from Wittgenstein’s notion of tautology. These are what are needed if the attack on psychologism is to have any point at all. But, of course, if individuation (hence, numerical identity) cannot be captured in purely predicative terms, which seems reasonable enough and which no one

210

JOSEPH MARGOLIS

has ever shown to be false, then the identity of indiscernibles is itself false and the Leibnizian “law” of identity is no more than a limited convenience (as Leibniz himself realized; see Russell [1992]). It may well be a mistake to argue, with Mill, that logic and mathematics depend in some essential way on empirical generalization; however, the segregation of uninterpreted logical truths from mathematical truths, the further disjunction between domains like arithmetic and the uninterpreted geometries of Euclid and Riemann (and their reception in the context of relativity physics), and the attempt to test empirically (as a consequence) the validity of putatively synthetic a priori truths (see Coffa 1991, chapter 10) leave little in the way of doubt that the antipsychologistic case for the “laws of thought” is very seriously — possibly fatally — weakened by this time. The mere admission of necessity is not, by itself, a serious hurdle for the opponent of Fregean and Husserlian antipsychologism. This is already clear from Quine’s concession to Carnap regarding logical truth. The serious question arises only with respect to necessity de re or de cogitatione or to some similar claim (say, regarding analyticity). Quine is absolutely straightforward here: “Along with the notion of logical or mathematical necessity I reject also the notion of physical or natural necessity, and thus also the distinction between law and accidental generalization” (1986a, 39798). The nerve of the argument seems clear enough: “I see pure mathematics,” he explains, “as an integral part of our system of the world” (398). If that be admitted, then it appears that mathematical necessity cannot be any more compelling than natural necessity or logical necessity. I can do no better here than report my intuitions: I cannot pretend to mathematical competence. But if mathematics is expressed in terms of sets, then the problematic standing of sets cannot be ignored; and if the necessity of relations between numbers is an abstraction from the necessity of relations between sets, then the doctrine of mathematical necessity begins to appear thin indeed. At the very least, the burden of proof belongs to Husserl and Frege and their followers. But, now, the option does not include only — or principally — the stalwarts of psychologism.11 Department of Philosophy Temple University, Philadelphia, USA

NOTES 1

See Mohanty 1989, particularly his opposition to Dreyfus’s treatment of the Husserlian noema (5, 7); see also Dreyfus (1984, Introduction). 2 I take this to be the sense in which Putnam is committed to something like psychologism, although Putnam apparently now believes his “internal realism” implicates some eighteenthcentury “representationalism” (which may be easily excised, if true) that requires a more general adjustment. The full replacement (what Putnam now calls “natural realism”) is not yet

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 211 in place (see Putnam [1987], Lecture 2; 1994b). I may say that I have formulated a constructive (a constructivist) form of realism that is (for that reason, psychologistic or hospitable to psychologism) that does not commit to anything like eighteenth-century representationalism (and is not a form of idealism). It also rejects modal necessities and shows how to avoid, coherently, any unconditional commitment to bivalence, whether in accord with excluded middle or (as with Dummett) tertium non datur. (See my Historied Thought, Constructed World [1995].) 3 See also Putnam’s “Rethinking Mathematical Necessity” (1994a) for an excellent (Wittgensteinian) gloss on the problem of the a priori among the laws of thought. 4 For a discussion that touches on this matter, see my “Deferring to Derrida’s Difference” (1994). 5 Here, Mohanty invokes Foucault’s mention of the “empirico-transcendental doublet” and Seebohm’s argument to the effect that psychologism is a form of relativism. I think the argument is a non sequitur, in the obvious sense that psychologism is a form of constructivism and that constructivism is hospitable to, but not tantamount to, relativism. In any case, I take relativism to be recoverable in a coherent and self-consistent form. For Foucault’s usage, see The Order of Things (1970 chapter 10). But, of course, Foucault was hardly a phenomenologist; on the contrary, part of his inexplicit labor was to undermine Husserl’s pretensions. See, also, Seebohm’s “The More Dangerous Disease” (1989) and Margolis’s “Entailments from ‘Naturalism = Phenomenology’” (1989). On relativism, see Margolis (2000). 6 This goes directly contrary to Davidson’s (1984) well-known advocacy of Tarski’s schema applied to natural languages. Here, Davidson and Dummett converge. 7 I have italicized the expression in advance of to show the sense in which Dummett’s argument is aprioristic. 8 The details are given in my The Truth about Relativism (1991). 9 For this reading of Frege and Husserl, see Willard (1989), Dummett (1981, 204), and Skorupski (1984, 240). (The references to Dummett and Skorupski are given in Willard 1989). 10 This is the site of Carnap’s option to Quine’s complaint. 11 On completing this paper, I came across Kusch’s Psychologism: A Case Study in the Sociology of Knowledge (1995). (I must thank Robert Nola, University of Auckland, for drawing my attention to this.) There can be no doubt that Kusch has assembled the fullest array of materials bearing on the reception of Frege’s and Husserl’s original treatments of psychologism. I would be less than candid, however, if I did not acknowledge my puzzlement at Kusch’s strategy. First of all, following Bloor (1991), Kusch favors the “strong program” of the sociology of knowledge; that is, he accepts “impartiality” (regarding matters of truth and falsity) and “methodological symmetry” (requiring no assessments of objective rationality). But this appears to ignore the reflexive import of the sociology of knowledge itself. Second, and more seriously, if the “strong program” is viable, then I see no reason to confine the discussion of psychologism to its terms; that is, I see no reason not to go on to the philosophical appraisal of psychologism itself.

REFERENCES

Baker, G. P., and P. M. S. Hacker. “Frege’s Anti-Psychologism.” In Perspectives on Psychologism, 75-137. See Notturno 1989. Bloor, David. Knowledge and Social Imagery. 2d ed. Chicago: University of Chicago Press, 1991. Bohnert, Herbert G. “Quine on Analyticity.” In The Philosophy of W. V.

212

JOSEPH MARGOLIS

Quine, 77-91. See Hahn and Schilpp 1986. Carnap, Rudolf. Logical Foundations of Probability. London: Routledge & Kegan Paul, 1950. Carnap, Rudolf. Meaning and Necessity. 2d ed., enlarged. Chicago: University of Chicago Press, 1956a. Carnap, Rudolf. “Meaning and Synonymy in Natural Languages.” In American Philosophers at Work: The Philosophic Scene in the United States, ed. Sidney Hook, 58-74. New York: Criterion, 1956b. Carnap, Rudolf. The Logical Syntax of Language. Trans. Amethe Smeaton. Paterson: Littlefield, Adams, 1959. Carnap, Rudolf. “Intellectual Autobiography.” In The Philosophy of Rudolf Carnap, 1-84. 1963a. See Schilpp 1963. Carnap, Rudolf. “W. V. Quine on Logical Truth.” In The Philosophy of Rudolf Carnap, 915-23. 1963b. See Schilpp 1963. Coffa, Alberto. The Semantic Tradition from Kant to Carnap: To the Vienna Station. Ed. Linda Wessels. Cambridge: Cambridge University Press, 1991. Davidson, Donald. “In Defense of Convetion T.” In Inquiries into Truth and Interpretation, 65-75. Oxford: The Clarendon Press, 1984. Dreyfus, Hubert L., ed. Husserl, Intentionality and Cognitive Science. Cambridge: MIT Press, 1984. Dummett, Michael. Truth and Other Enigmas. Cambridge: Harvard University Press, 1978. Dummett, Michael. Frege: Philosophy of Language. 2d ed. Cambridge: Harvard University Press, 1981. Dummett, Michael. Frege and Other Philosophers. Oxford: The Clarendon Press, 1991a. Dummett, Michael. The Logical Basis of Metaphysics. Cambridge: Harvard University Press, 1991b. Foucault, Michel. The Order of Things: An Archaeology of the Human Sciences. New York: Vintage, 1970. Gadamer, Hans-Georg. Truth and Method. Trans. Garrett Barden and John Cumming. New York: Seabury Press, 1975. Goldman, Alvin I. Liaisons: Philosophy Meets the Cognitive and Social Sciences. Cambridge: The MIT Press, 1992. Hahn, Lewis Edwin, and Paul Arthur Schilpp, eds. The Philosophy of W. V. Quine. LaSalle: Open Court, 1986. Husserl, Edmund. Cartesian Meditations: An Introduction to Phenomenology. Trans. Dorion Cairns. The Hague: Martinus Nijhoff, 1960. Husserl, Edmund. The Crisis of European Sciences and Transcendental Phenomenology: An Introduction to Phenomenological Philosophy. Trans. David Carr. Evanston: North ern University Press, 1970. Kuhn, Thomas S. The Structure of Scientific Revolutions. 2d ed., enlarged.

LATE FORMS OF PSYCHOLOGISM AND ANTIPSYCHOLOGISM 213

Chicago: University of Chicago Press, 1970. Kusch, Martin, Psychologism: A Case Study in the Sociology of Knowledge. London: Routledge, 1995. Margolis, Joseph. “Entailments from ‘Naturalism = Phenomenology’”. In Perspectives on Psychologism, 32-57. See Notturno 1989. Margolis, Joseph. The Truth about Relativism. Oxford: Basil Blackwell, 1991. Margolis, Joseph. “Deferring to Derrida’s Difference.” In European Philosophy and the American Academy, ed. Barry Smith, 195-226. LaSalle: Monist Library of Philosophy, 1994. Margolis, Joseph. Historied Thought, Constructed World: A Conceptual Primer for the Turn of the Millennium. Berkeley: University of California Press, 1995. Margolis, Joseph. “Relativism and Interpretive Objectivity”, Metaphilosophy, 31, 2000, 200-26. Mohanty, J.N. “Psychologism.” In Perspectives on Psychologism, 1-10. See Notturno 1989. Notturno, Mark A., ed. Perpectives on Psychologism. Leiden: E. J. Brill, 1989. Putnam, Hilary. The Many Faces of Realism. LaSalle: Open Court, 1987. Putnam, Hilary. “Rethinking Mathematical Necessity.” In Words and Life, ed. James Conant, 245-63. Cambridge: Harvard University Press, 1994a. Putnam, Hilary. “Sense, Nonsense, and the Senses: An Inquiry into the Powers of the Mind.” (The Dewey Lectures) The Journal of Philosophy, 91, 1994b, 445-517. Quine, W. V. “Two Dogmas of Empiricism.” In From a Logical Point of View, 30-46. Cambridge: Harvard University Press, 1953. Quine, W. V. Carnap and Logical Truth.” In The Philosophy of Rudolf Carnap, 385-406. See Schilpp 1963. Quine, W. V. “Epistemology Naturalized.” In Ontological Relativity and Other Essays, 69-90. New York: Columbia University Press, 1969a. Quine, W. V. “Ontological Relativity.” In Ontological Relativity and Other Essays, 26-68. New York: Columbia University Press, 1969b. Quine, W. V. “Reply to Charles Parsons.” In The Philosophy of W. V. Quine, 396-403. 1986a. See Hahn and Schilpp 1986. Quine, W. V. “Reply to Herbert G. Bohnert.” In The Philosophy of W. V. Quine, 93-95. 1986b. See Hahn and Schilpp 1986. Quine, W. V. Pursuit of Truth. Enlarged ed. Cambridge: Harvard University Press, 1992. Russell, Bertrand. A Critical Exposition of the Philosophy of Leibniz. 2d ed. London: Routledge, 1992. Schilpp, Paul Arthur, ed. The Philosophy of Rudolf Carnap. LaSalle: Open Court, 1963. Seebohm, Thomas S. “The More Dangerous Disease: Transcendental

214

JOSEPH MARGOLIS

Psychologism, Anthropologism, and Historism.” In Perspectives on Psychologism, 11-31. See Notturno 1989. Skorupski, John. “Dummett’s Frege.” In Frege: Tradition and Influence, ed. Crispin Wright, 227-41. Oxford: Basil Blackwell, 1984. Tarksi, Alfred. “The Concept of Truth in Formalized Languages”. In Logic, Semantics, Mathematics, trans. J. H. Woodger, 2d ed., ed. John Corcoran, 152-278. Indianapolis: Hackett, 1983. Willard, Dallas. “The Case against Quine’s Case for Psychologism.” In Perspectives on Psychologism, 286-95. See Notturno 1989.

MICHAEL JUBIEN

PROPOSITIONS AND THE OBJECTS OF THOUGHT

1. Introduction Nowadays it’s very common to think of Fregean thoughts as propositions. In this view, propositions are taken to be those occupants of Frege’s “third realm” that are the ultimate bearers of truth values and the objects of belief and other attitudes. So they’re seen to be objects of thought in the modern sense of that term. They are natural realm-mates of Platonic properties and relations, and as such they are held to exist independently of minds and the spatiotemporal realm. On this Fregean picture, to have a propositional thought, say a belief, is to stand in a certain special relation to a specific proposition. This proposition is either true or false, and gets to be so according to whether it accurately represents the world (or part of the world). In this way our (propositional) thoughts are, in general, about the world outside us, and the overall picture therefore conforms to Frege’s famous rejection of “psychologism”. This is certainly an elegant picture, one that reflects a very appealing interpretation of how natural language works, and also reflects our everyday notion of truth as (in general) correspondence to a mind-independent reality. But I don’t think the picture is correct, because I don’t think there really are any propositions. If this is right, then propositions aren’t available to be the literal objects of thought. But then we have to ask whether Frege may have been wrong to reject psychologism. I don’t think he was. The reason is that I think the objects of thought are the very entities that are traditionally conceived to be the constituents of propositions, namely Platonic properties and relations. On this account, to have a propositional thought is to stand in an appropriate intentional relation to some of these entities. If it isn’t already obvious, I think it will emerge clearly in what follows that this view isn’t psychologistic. My goals in this paper are, first, to explain why I reject propositions, and second, to suggest how wemay accommodate various undoubted propositional phenomena without benefit of propositions. 215 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 215-227. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

216

MICHAEL JUBIEN

2. Problems with Propositions It may seem odd that someone would favor Platonic properties and relations, but then draw the line at propositions. People normally think these three types of entities either stand or fall as a single “intensional” package. The usual argument against intensional entities certainly applies across the board, and people either accept it or reject it. If they reject it, they usually celebrate by embracing all three. The usual argument, of course, is Quine’s complaint that intensionalentities have obscure or otherwise deficient “identity conditions”, or perhaps have none at all.1 I believe this is a very serious and unfortunate error, and I’ve said so in print. In particular, I’ve made the following four claims: (1) that the demand for identity conditions is ill-conceived; (2) that the very notion of identity conditions is confused and has never been satisfactorily explicated; (3) that the relation of identity is simple and should be taken as a metaphysical primitive; and (4) that if entities of any sort whatsoever really do exist, then each of them is identical with itself automatically (and notably without having to satisfy any condition at all).2 I won’t repeat my arguments here. Instead, I’ll just assume that Quine’s very influential argument is ineffective. So my complaint about propositions isn’t Quine’s, and in fact it simply doesn’t apply to properties or relations. Actually, I have two complaints, but they’re best presented in connection with theories of propositions, so let’s turn to that topic first. I think there are three essentially different sorts of theories of propositions around these days. In one, the entities that play the role of propositions really aren’t propositions. Instead they are set-theoretic or other mathematical constructions, sometimes involving “possible worlds”, truth values, maybe properties and relations (but usually just “extensions”), often “logical operations”, and sometimes even linguistic entities. A simple early example is Montague’s (Kripke-inspired) treatment of propositions as functions from possible worlds to truth values. George Bealer has urged that “most of us have difficulty honestly believing that the very propositions we believe and assert are really functions or ordered sets...”3 And here is a related point. Setting aside doubts about the notion of a possible world, the Montague approach is easily seen to be equivalent to treating propositions as sets of possible worlds.4 But a set of possible worlds is simply not a function from possible worlds to truth values. How could the entities of one type have a stronger claim to being the propositions than those of the other? I believe only if there was an independent reason for thinking that it’s those entities that really are the objects of attitudes or the bearers of truth values. But neither sort of entity is an intuitively plausible candidate for either role. Similar points may be made with any other theory of this same general sort, which we may as well call mathematical theories. A long time ago Paul

PROPOSITIONS AND THE OBJECTS OF THOUGHT

217

Benacerraf taught us that the multiplicity of theoretically adequate “reductions” of natural numbers to sets is powerful evidence that numbers aren’t sets.5 I like to make the same point with ordered pairs since it’s much easier to think, at the outset, that ordered pairs are sets than it is to think numbers are. (Ordered pairs will reappear soon.) But after all, most who offer theories like Montague’s probably don’t think the mathematical entities that play the proposition-role really are propositions in the first place. These theorists probably see themselves as merely offering a model for their theory — one in which the entities of the domain are proposition-surrogates that mimic the structure of the real propositions, a structure they think their theory captures. In fact, it’s entirely consistent to offer such a theory while rejecting genuine Platonic propositions. So one and the same mathematical theory could be propounded with or without ontological pretensions, and the central point here is that if the pretensions are present, then they’re thwarted by a Benacerraf dilemma. A second type of theory would try to provide a more plausible account of the intrinsic nature of Platonic propositions, one that doesn’t treat them as mathematical entities of some antecedently familiar sort. To succeed, a theory of this kind must mesh well with our intuitions about propositional constituency, and hence with our conception of propositions as complex entities of their own special kind. Inevitably, it would include an account of how propositions are composed of entities like properties and relations, which after all are the intuitive constituents ofpropositions. I confess that in the past I’ve flirted with theories of this general sort.6 Let’s call such theories ontological. Then, finally, the third type of theory takes propositions as primitive, regarding them as sui generis Platonic entities. Let’s call these primitiveentity theories. Bealer is a leading proponent of such a theory.7 My two complaints about propositions are best lodged against ontological theories, because they make claims about how propositions are in themselves. But, as we’ll see, the second complaint also applies to primitive-entity theories. The first complaint is that despite the firm intentions of their proponents to avoid the pitfalls of mathematical accounts, ontological theories also face a Benacerraf dilemma. To see this, let’s think about how one might go about developing an ontological theory. The essence of a proposition is to represent. Propositions represent the world as being one way or another. If they didn’t represent in this way, it would be utterly implausible to view them as the ultimate bearers of truth values. Now, in an ontological account, the representing that a proposition does must be aproduct of its internal makeup — it must somehow emerge from the constituents of the proposition itself. To ground the representational force outside the proposition, say in the intentions of conscious agents, would sabotage the motivating “ontological” idea that the

218

MICHAEL JUBIEN

entity really is a proposition, not a mere surrogate. So it must emerge from a successful ontological account that no other entity could be the proposition in question. If its “propositionality” were derived from representational activity outside the entity, then the hope of uniqueness would be lost. In effect we would merely have a philosophically more enticing mathematical account, with parallel surrogate accounts in the wings. Now consider the intuitive proposition that some dogs are swimming. If this is going to be a genuinely Platonic entity, it better have the property being a dog and the property swimming as constituents. It also better have some specific “logical glue” that sticks these properties together in the right way, so that the result doesn’t represent all dogs as swimming, or everything as being a swimming dog, etc. How might this work? An idea that once struck me as promising goes like this. Among the relations that one property will bear to another is one we can call coinstantiation. Two properties are coinstantiated iff something is an instance of both. Now consider the mereological sum of the property of being a dog and the property of swimming together with the coinstantiation relation. I believe this is a perfectly legitimate Platonic entity.8 Could it be the proposition that some dogs are swimming? First let’s see what the idea has in its favor. The sum has these properties as parts, and perhaps they represent their instances: dogs and swimmers. Since these instances are in the spatiotemporal world, perhaps the sum also represents them and, further, does so in a way that has the kind of representational force we require in a genuine proposition. But how? The sum has the coinstantiation relation as a part. If we proceed on strict analogy with the properties, this relation will represent any properties that may happen to bear it to each other. As a result, the sum would not only represent dogs and swimmers, but also all instances of properties in the field of the coinstantiation relation. Since all instantiated properties are coinstantiated with themselves, the field includes every property that has an instance. Of course this boggles the mind. So the representational contribution of coinstantiation must work some other way. Intuitively, we want it just to apply the logical glue. But how might this work? Perhaps we were wrong to think that, in the sum, being a dog represented all dogs. Maybe, thanks to the presence of swimming and coinstantiation, all that got represented were the swimming dogs. In short, maybe the sum does its representing not piecemeal but in one compositional swoop. But there are several problems with this idea. I will only mention a couple. First, suppose the realm of properties is sufficiently abundant as to include being a swimming dog. Very plausibly, this property is simply the merelogical sum of being a dog and swimming. But then our sum of two properties and one relation is identical with a sum of one property and that same relation.9 Then why is this sum not the proposition that some

PROPOSITIONS AND THE OBJECTS OF THOUGHT

219

swimming dogs are swimming dogs? After all, the original sum is identical with being a swimming dog plus being a swimming dog plus coinstantiation. Here’s the second problem, and if anything it’s more serious. Consider an “A-form” proposition, say the proposition that all canines are dogs, and call the characteristic logical relation of the A-form subextensiveness. Then the proposition that all canines are dogs — for a genuine Platonist — is just the proposition that being canine is subextensive to being a dog. Of course, unlike coinstantiation, subextensiveness isn’t symmetric. So at the very least we would need some further ingredient to get things glued together in the right way, enabling us to distinguish that all canines are dogs from that all dogs are canines. This is eerily reminiscent of the Benacerraf dilemma for the reduction of ordered pairs to sets. The ordered pair of a and b can’t just be the pair set whose members are a and b, for then we lose the order. The solution is to code in the order by an artificial device. Thus is usually regarded as {{a,b},{a}}. The singleton member of this set encodes the fact that the firstelement of the ordered pair is a. But we could just as easily have used the same basic strategy, except with the singleton member coding the second element of the ordered pair. So we have a Benacerraf dilemma, and the idea that ordered pairs really are sets can no longer be defended. We might resort to an ordered-pair-like strategy to encode which of the two properties is subextensive to the other. For example, we could take the proposition that all canines are dogs to be the sum of subextensiveness with the property of being the property of being canine and the property of being a dog. The property of being the property of being canine — having being canine as its only instance — is to the property of being canine as the singleton of a is to a. This works, but — no surprise, I’m sure — it won’t generalize smoothly to all propositions. But the point is that any such strategy, including those that might generalize, is just an artificial coding. Now, this sort of situation is going to arise whether we take constituency mereologically or not. For consider again the proposition that all canines are dogs. We’re trying to see it as a Platonic entity having the properties being canine and being a dog as constituents. Of course we would finally have to say something about constituency. But whatever we might say, the constituents would have to be bound together in the right way with logical glue, glue that is itself Platonic in nature. So the glue would have to include something more or less like the subextensiveness relation. It is clear, for the reason just given in the mereological case, that it won’t be enough merely for this relation to be a further constituent. For each of the two properties has to be associated with the correct position in the relation. What’s crucial here is that no particular way to effect the positional association is mandated by our prior conceptions of properties, relations, and the Platonic realm in general. We’re faced with a very general, sophisticated coding problem, essentially like the simpler ordered-pair

220

MICHAEL JUBIEN

problem, and thereforeone to which our ingenuity may be applied in any number of different but interchangeable ways. This calls for a bright red flag. For the ontological project is not just to find a modeling of propositions among Platonic entities. We aren’t just looking for Platonic propositional roleplayers. We’re looking for the actual propositions! If there is a multiplicity of interchangeable systems of proposition-like complexes, then we have a Benacerraf dilemma, and the fact that we need positional coding guarantees such a multiplicity. That’s my first complaint against propositions: the ontological effort to construe them as complex Platonic entities with their intuitive constituents creates a Benacerraf dilemma that is just as bad as the one that confronts mathematical theories of propositions. Now for the second complaint. Although the propositional system of any ontological theory would consist of Platonic entities, no such entity would be a plausible candidate for being a proposition even if there were no Benacerraf dilemma. The reason is that our active intentionality is required in order for there to be a coding in the first place. These Platonic entities, in and of themselves, do not determine positional association. They lack the requisite intentionality. When we “see” such an entity “as a proposition” we confer a strictly derivative dose of intentionality upon something that is otherwise intentionally inert. It maybe very convenient to suppress this and proceed as if the entity had the intentional feature on its own, but doing so is a matter of efficiency of thought, not ontology. The fate of the ontological effort prompts a rethinking of its origins. The leading idea was that certain entities having constituents that represent things in the world might parlay the representational force of these constituents into the kind of representation we require of propositions, namely, truth-value-supporting representation. But we found that the resulting complex entities, on their own, have no representational force at all. What went wrong? I think the project was doomed from the start by the illusion that, for example, the property of being canine represents canines. It doesn’t follow from the fact that a property has instances, that it represents those instances (whether individually or collectively). After all, a property like being canine needn’t have had any instances at all and might have had entirely different ones. Of course, if we like, we may utilize a property to represent its instances, but then the representational oomph is coming from us. We, after all, can use anything to represent anything else. If, employing a prearranged code, a spy displays a lemon to signal the presence of dogs, there’s no illusion that the lemon itself is doing any representing on its own. With a property and its instances, the illusion is just more likely, but it’s still an illusion. But there’s nothing special here about properties. In the end, it’s implausible to think that any genuine Platonic entity could represent on its own cuff. Representation is an “intentional” or “outer-directed” relation. If x

PROPOSITIONS AND THE OBJECTS OF THOUGHT

221

represents y, then x has a part that “stands for”, “refers to”, or is otherwise “about” y. This is the admittedly vague but undeniable heart of the notion. It borders on the absurd to suppose that any inert, non-spatiotemporal entity could have a part that, in itself, plays any such referential or quasireferential role, especially with respect to contingent, concrete entities. Representation is ultimately the business of beings with intentional capacities, in short, thinkers. So Bealer’s primitive Platonic propositions are no better off than those of the ontological theorist. To hold, as he does, that propositions are sui generis doesn’t relieve them of the burden of representation, and this is a burden they simply can’t bear.

3. Propositionality Without Propositions So suppose these considerations are accepted and we conclude there really are no Platonic propositions. It is still undeniable that there are many everyday phenomena, such as belief, that seemingly involve propositions. I think these phenomena may be convincingly accounted for without propositions. Let me now sketch the main idea for a theory that promises to achieve this. The three traditional roles that propositions are supposed to play are: (1) objects of the attitudes; (2) ultimate bearers of truth values; and (3) meanings (or semantic values) of certain declarative sentences. Let’s take them up in that order. (1) Human beings certainly do seem to be sources of intentionality. We evidently represent things that are entirely independent of ourselves, and we do so with great regularity and with an impressive degree of apparent success. So let’s assume that this really happens. The question of how it happens is one of the most compelling problems in philosophy of mind. Unfortunately, I think every well-known position on this problem is unsatisfying in one way or another. So the view I’ll sketch about propositional representation will avoid the question of how it takes place. Now consider an example from the realm of belief. Suppose a believes that all dogs bark. (I have deliberately chosen a case of false belief.) Then a is in a state that is somehow or other intentional. It’s about dogs and barking. It would be nice if we could see this state as one in which a stands in a certain relation, belief, to a Platonic proposition, where the proposition carried the intentional burden. This classic picture of propositional belief as a binary relation between agents and propositions is very powerful and compelling. But if there aren’t any Platonic propositions, it’s still wrong. And even if there are various Platonic models of propositions, it’s utterly implausible to think that their proposition-surrogates are involved in belief. Most people have no knowledge of such

222

MICHAEL JUBIEN

esoteric matters and, after all, one and the same surrogate will inevitably wind up doing duty for different propositions in different systems. The intuitive constituents of the would-be proposition that all dogs bark are the properties being a dog and barking, the subextensiveness relation, and no others.10 Of course these are the same as the constituents of the different proposition, that all barkers are dogs. Now I want to suggest that the most fundamental phenomena of propositionality can be entirely accounted for “contextually”, appealing to such would-be constituents and, of course, without propositions. This proposal will place the burden of intentionality squarely on our shoulders, where the previous discussion strongly suggested it belongs. The central idea, which goes back at least to Russell,11 is that in its most fundamental form, belief (and any other propositional attitude) is a multigrade relation that may hold between an appropriate subject and some properties and relations. Let Bel stand for this relation and let d and b stand for being a dog and barking. Also let S stand for subextensiveness. Then a’s believing that all dogs bark consists in Bel’s holding in a certain way among a, d, S, and b, which we may represent in standard first-order fashion by Bel(a,d,S,b)’, letting theorder capture the “way” in question. (And a’s hoping that some dogs don’t bark is represented by H(a,d,C,not-b)’, with H for hope and C for coinstantiation, etc.) Thus the propositional attitudes are seen as relations, but not as binary relations between subjects and propositions. Instead they are relations between subjects and the natural constituents of the relevant would-be propositions. This treatment credits subjects with, first, the intentional ability to hook together the constituents in the right way, and second, the capacity to have a variety of different attitudes toward the same constituents (in the same order). I claimed earlier that the intuitive propositions, that all dogs bark and that being a dog is subextensive to barking, are identical. Thus we may read Bel(a,d,S,b)’ in a potentially illuminating way: “a believes being a dog to be subextensive to barking.” Here there is no sentential component crying out for a proposition as referent. In fact, this makes it look like the attribution of a belief to a that is de re with respect to the two properties and subextensiveness. This is a picture I would like to encourage. It grants to a the ability to be related intentionally to each of the three “constituents” and, again intentionally, to take one of them as relating the others in a particular way. A key problem with propositions was this: Although there’s no difficulty in making the “logical glue” a propositional constituent, it’s impossible to construe any glue-containing candidate proposition as applying the glue in the right way to the other constituents without relying on some external convention. This devastating problem has no analogue in the present theory simply because the glue is applied directly by the agent. This example suggests how to deal with cases in which an agent

PROPOSITIONS AND THE OBJECTS OF THOUGHT

223

seemingly stands in a relation to a specific proposition. What about cases in which we apparently quantify over propositions? For example, it might be that Bill believes everything Hillary does. With propositions and binary belief, this just means that for every proposition p, if Hillary believes p, then Bill believes p. In the proposition-free, multigrade approach things can’t be quite this simple.12 But they’re not that complicated either. We’ve dispensed with propositions but not with propositional attitudes. And we have reconstrued the attitudes as multigrade rather than binary relations. Let’s call the entities that stand in propositional attitude relations with agents — entities like d, S, and b in the example — propositional relata (taking care not to understand this as presupposing propositions). Although I won’t try to defend it here, I believe that propositional relata are always properties and relations, and never individuals.13 Then the natural idea is to replace apparent quantification over propositions with quantification over propositional relata. The problem is that the attitudes are multigrade relations. An agent can have beliefs involving arbitrary numbers of propositional relata. The solution to this problem involves “plural” quantification, a phenomenon that is common in ordinary English but relatively unexplored in logic. The classic example of plural quantification is provided by the sentence “Some critics admire only each other.” It isn’t too hard to see that ordinary first-order quantifiers are inadequate to capture the everyday meaning of this sentence. (If you’re skeptical, have fun trying!) Yet it’s very clear that the sentence makes perfectly good sense and we know exactly what it takes for it to be true. I can’t go into details about plural quantification here.14 So let’s just rely on the intuitive clarity of the notion and announce the solution to our problem as follows: For any propositional relata, if Hillary stands in the belief relation to them (in any specific way), then so does Bill. The plural quantifier finesses the belief relation’s lack of a specific index. But what is it to stand in the belief relation to some propositional relata in a specific way? Well, it’s possible to believe being a dog to be subextensive to barking, but it’s also possible to believe barking to be subextensive to being a dog. In these different beliefs a person would bear the belief relation to the same relata but in different ways. On the present account, to believe that all dogs bark is nothing more nor less than to believe being a dog to be subextensive to barking. If you have this belief, then you have a certain relational property: believing being a dog to be subextensive to barking. This property is obviously different from believing barking to be subextensive to being a dog. We may think of such properties as belief-properties. Then to be related by the belief relation to some propositional relata in a specific way is just to have a specific beliefproperty. With belief-properties at our disposal we may compress the solution to the Hillary-Bill problem as follows: For any belief property, if

224

MICHAEL JUBIEN

Hillary has it, then so does Bill.15 I hope it is now reasonable to think that the multigrade approach may be exploited to account for the full range of propositional-attitude phenomena without appealing to propositions. So let’s turn to the “ultimate-truthbearer” role. (2) It is often said that propositions are the ultimate bearers of the truth values. Of course, if there are no propositions, then this can’t be right, whatever it may mean. What it is usually taken to mean is that our attributions of truth values to entities other than propositions depend on there being propositions with suitable truth values. A typical picture is this. Someone utters a sentence like “Some dogs bark” and this leads us to say a couple of things. First, we’re inclined to attribute truth to the utterance, that is, to the specific sentence-token that was produced. Second, we’re inclined to say that what was said is true. The thing that was said is supposed to be a proposition, one that happens to be true. This is what licenses us to call the utterance true. In other words, the utterance expressed the proposition, and got to be true derivatively because the proposition was true in the most basic sense of the word. Hence the present question is whether we can get the utterance to be true without benefit of the proposition. We ought to be able to do this, since we’ve now seen that we already do the representing in our thought without the aid of propositions. The clues are all there in the example of belief. Where once we said that the utterance of “Some dogs bark” expresses the proposition that some dogs bark, we now say that it represents being a dog and barking as being coinstantiated. Then we declare the utterance true iff these two properties in fact are coinstantiated. We don’t need a proposition as an intermediary between our utterance and the world. We are capable of asserting these properties to be coinstantiated, and we do just this when we utter “Some dogs bark” (assertorically). We have the capacity, in our thought, to represent the world as satisfying various conditions, and we have the coordinate capacity to express such representations in our language. If the world cooperates by satisfying the conditions, then the thought that does the representing is true, and so is any assertion of the thought we may produce. Truth is now seen as a property that applies in the first instance to our propositional representations, whether they occur only in our thought or also in linguistic (or other) expressions of our thought. Whether truth applies to such a representation of course depends on how things are in the world at large, so the proposition-less notion of truth is every bit as realistic and nonsubjective as the traditional, proposition-based account. It is for this reason that we don’t automatically embrace psychologism when we abandon propositions. Someone might think this view of truth deprives the pre-intentional world of truths. It invites questions like Billions of years ago, before there

PROPOSITIONS AND THE OBJECTS OF THOUGHT

225

was intelligent life, wasn’t it true that stars emitted light? Of course we want to say that it was true that stars emitted light at many times in the distant past, and the theory of propositions gives us a convenient way to cash this out. For the proposition that stars are emitting light at t (for any t) is an atemporal Platonic entity, and according to the theory (and of course the facts) for countless long gone times t, such propositions are (and were) true. I’ll just make a couple of brief points in response. First, suppose that the present theory did deprive the pre-intentional world of entities instantiating truth values. Why panic? It wouldn’t follow that, billions of years ago, stars weren’t emitting light. They were. The properties of being a star and emitting light were indeed coinstantiated in those otherwise dark times. The second point is that certain rather plausible Platonic accounts of linguistic entities have the initially surprising consequence that the sentences of our language are atemporal entities themselves, because they’re properties (whose instances, when they have them, are sentence tokens).16 Most of these sentences never have had and never will have tokens, and of course in the darktimes, none of them did. But this doesn’t entail that none of them had truth values. (Roughly, any univocal sentence is true iff any representation that would be appropriately expressed by an utterance of the sentence would be true.) (3) Now let’s turn to the matter of meanings and semantic values. The latter term strongly suggests technical semantics. In this setting it can of course be very useful to enlist specific entities to serve as semantic values of sentences (etc.). Many different choices are possible, and different theorists have opted for one or another in a variety of theoretical contexts. Montague’s functions from possible worlds to truth values come to mind. Such theoretical strategies are surely above reproach, but as we saw, the chosen entities just aren’t propositions. They’re proposition-surrogates, featuring only derived intentionality, and systematically replaceable by entirely different entities. But, technical semantics aside, there is the informal idea that sentences (or sentence-tokens) have meanings, and that these meanings are often propositions. For instance, it’s very natural to think that an utterance of “Some dogs bark” means that some dogs bark. But, taking a lead from Quine, we don’t have to reify propositions in order to sustain the meaningfulness of the utterance. Its meaningfulness doesn’t consist in there being some specific entity that is its meaning. Thanks to the intentional capacities of the speaker, the utterance represents the world as being a certain way (as described earlier), and that’s where its meaningfulness lies. That we commonly report such representations with the aid of (propositional) that-clauses merely reflects the fact that ordinary language carries a heavy superficial commitment to propositions. But I believe we’ve now seen substantial philosophical reasons for rejecting propositions, along

226

MICHAEL JUBIEN

with a basic sketch of how to live without them.17 Department of Philosophy University of California, Davis, USA

NOTES 1 2

See, for example, Quine (1960), esp. 200-9. For details, see Jubien (1996). This paper was written before I began to doubt the existence of Platonic propositions. But the general conclusions don’t depend on this in any way. Soon we’ll see that the present suspicions about propositions have nothing to do with identity conditions. 3 Bealer (1998,2). 4 Just replace each function from worlds to truth values by the set of worlds mapped to the truth value T by that function (and make appropriate adjustments in the semantics). 5 See Benacerraf (1965). 6 For example, Jubien (1993). 7 Bealer (1998). 8 For a spirited defense of the idea that mereology extends beyond the spatiotemporal realm, see Lewis (1991). 9 And very likely is also a sum of a number of different pluralities of properties and that relation as well, reflecting further analyses of the two starting properties. 10 Recall that for a genuine Platonist, the intuitive propositions that all dogs bark and that being a dog is subextensive to barking are one and the same. It might be objected that this identification is incorrect, that the former proposition concerns dogs, not properties, and the latter concerns properties, not dogs. But this is a superficial objection. The first-mentioned proposition concerns dogs in the sense that, in general, dogs help make it true. But beinga dog’s being subextensive to barking, if it were, would consist in all the instances of the one being instances of the other. And the instances (if any) of the one would be dogs. On the other hand, the first-mentioned proposition doesn’t “transparently” concern whatever entities might happen to instantiate being a dog, but somehow independently of their doing so. If being canine and being a dog were coextensive, this intuitive proposition would not be the intuitive proposition that all canines bark. The property, being a dog, is hence an essential constituent of the first intuitive proposition, and any reasons for distinguishing the first from the second now evaporate. We simply have both “property-level” and “object-level” ways of expressing the same (would-be) proposition. 11 Russell had somewhat different reasons. See Chapter XII of his (1912) (and elsewhere). He says, 125: “The relation involved in judging or believing must, if falsehood is to be duly allowed for, be taken to be a relation between several terms, not between two. When Othello believes that Desdemona loves Cassio, he must not have before his mind a single object, ‘Desdemona’s love for Cassio’, or ‘that Desdemona loves Cassio’, for that would require that there should be objective falsehoods, which subsist independently of any minds; and this, though not logically refutable, is a theory to be avoided if possible. Thus it is easier to account for falsehood if we take judgement to be a relation in which the mind and the various objects concerned all occur severally; that is to say, Desdemona and loving and Cassio must all be terms in the relation which subsists when Othello believes that Desdemona loves Cassio”. Russell thought individuals like Cassio were constituents of intuitive propositions and I don’t. But that’s a different story. 12 A number of people, including Ruth Marcus and Ted Sider, have suggested that there may be quantificational difficulties in the present account. 13 This, of course, is a significant departure from Russell. See note 11.

PROPOSITIONS AND THE OBJECTS OF THOUGHT

227

14 Lewis discusses and appeals to plural quantification in (1991). An important early discussion may be found in Boolos (1984). 15 Some contexts will require further resources. For example, it may be that Bush’s greatest fear is Gore’s greatest hope. Then there are some propositional relata to which Bush bears the fear relation in a certain way and Gore bears the hope relation in the same way. So Bush will have a fear-property and Gore will have a hope-property, and these relational properties will be structurally identical but will differ in (and only in) their base relations. 16 For example, I sketch this kind of an account in Chapter 3 of Jubien (1997). 17 An earlier version of this paper was presented at an APA symposium on psychologism at the 2000 Pacific Division meetings in Albuquerque. The session was arranged by Dallas Willard and the other speakers were John Dreher and Dale Jacquette — I thank all three for their remarks. I also thank David Copp, Greg Fitch, Jay Garfield, Mark Moffett, Terence Parsons, and Scott Soames for several valuable comments. Finally, I am grateful to Jeffrey King for many conversations on these and related topics. His own original account of propositons counts as “ontological” in my classification, but sometimes I think there lurks behind it a view that is really more akin to my own. King’s recent thinking on the matter may be found in his (1995).

REFERENCES

Bealer, George. “Propositions”, Mind, 107, 1998. Benacerraf, Paul. “What Numbers Could Not Be”, The Philosophical Review, 74, 1965, 47-73. Boolos, George. “To Be is To Be the Value of a Variable (or To Be Some Values of Some Variables)”, The Journal of Philosophy, 81, 1984, 43049. Jubien, Michael. Ontology, Modality, and the Fallacy of Reference. Cambridge: Cambridge University Press, 1993. Jubien, Michael. “The Myth of Identity Conditions”, Philosophical Perspectives, 10 (Metaphysics), 1996, 343-56. Jubien, Michael. Contemporary Metaphysics. Oxford: Basil Blackwell, 1997. King, Jeffrey. “Structured Propositions and Complex Predicates”, Noûs, 29, 1995, 516-35. Lewis, David. Parts of Classes. Oxford: Basil Blackwell, 1991. Quine, W.V.O. Word and Object. Cambridge: The MIT Press, 1960. Russell, Bertrand. The Problems of Philosophy. Oxford: Oxford University Press, 1912.

This page intentionally left blank

JOHN H. DREHER

THE CONCEPTS OF TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

INTRODUCTION

Although virtually no one now holds psychologism in its original, nineteenth-century form, I shall argue that there are current views that are psychologistic in very important respects. In this paper I want to consider the concepts of truth and knowledge as they figure in those theories. I shall further argue that psychologistic views involve subjective conceptions of truth and knowledge that are inconsistent with the prima facie correct understanding of these notions. Much anti-realism in fields like ethics is due to relativist conceptions of truth. Indeed, in its most radical form, the one with which I shall be principally concerned, psychologism is now the view called ‘cultural relativism.’ Although I see much of merit in psychologism and even cultural relativism, I conclude by arguing that a completely psychologistic understanding of human knowledge and belief would take us too far in the direction of subjectivism. What I value in psychologistic views is that they have acted as a continuing restraint upon uncritical idealism. Indeed, psychologism originated in the nineteenth century, the century of idealism, as a theory about the nature of philosophical thought. According to nineteenth-century psychologism, philosophy is a branch of empirical psychology. This means that philosophical truths are essentially truths about how our minds work. The argument for this claim is based on the belief that the methodology of philosophy is essentially introspective, but the fact that philosophy employs an introspective methodology does not by itself imply that its results are facts of empirical psychology. Indeed, Kantians maintain that the methodology of philosophy is essentially introspective, but insist that at least some of its results are a priori and hence not facts of empirical psychology. Nonetheless, early psychologism claimed that the fact that the method of philosophy is introspective made it properly a subject of 229 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 229-244. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

230

JOHN H. DREHER

empirical psychology, however its results might be characterized. By thinking of introspection as an ordinary matter to be investigated empirically, early psychologism sought to counter the Kantian line that the introspective character of philosophical method supported the claim of aprioricity of its results. The doctrine that both the activity of philosophy and its results are proper objects of an empirical science of the mind had little following in the nineteenth century, principally because Kant’s philosophy was dominant. By the end of the nineteenth and the beginning of the twentieth-century cracks in the Kantian view became visible, but philosophers did not respond to those stresses by resorting to the earlier, somewhat crude psychologistic accounts of philosophy. On the contrary, the decline of Kantian philosophy initially elicited a realist response within a predominantly empiricist culture. The difficulties that the realist response encountered, among other things, opened the door to new and ever more sophisticated forms of psychologism. To see all this it might be useful to begin by reminding ourselves of the events that led to the rejection of Kantian idealism and the re-emergence of realism.

I According to Kant the world-as-experienced, viz. the phenomenal world, has a structure that can be known a priori. Insofar as the phenomenal world is the object of perception, it is characterized by ‘pure forms of intuition.’ The pure forms of intuition structure (or, perhaps, are the structure) of both ‘inner’ and ‘outer’ experience. ‘Outer’ experience is both spatial and temporal. ‘Inner’ experience is temporal. The structure of the spatial component of outer experience is Euclidean because the ‘form of intuition’ that necessarily characterizes it is Euclidean geometry. The principles of Euclidean geometry can be known a priori. The structure of both ‘inner’ and ‘outer’ experience is also revealed by (or in) arithmetic, because experience is temporal. It is not entirely clear why it should be that the truths of arithmetic should characterize temporal experience. Surely much derives from the fact that temporal relations, (e.g., before and after) are, like arithmetic relations, (e.g., less than and greater than) in that each is irreflexive, anti-symmetrical and transitive. Like the truths of Euclidean geometry, arithmetic truths can be known a priori. According to Kant, we also find a priori knowledge on the conceptual side. The pure categories of the understanding, by which we conceive the world as characterized by contingencies and necessities, parts and wholes, causes and effects and so forth, are applied to the world through their schemata. This too allows for a priori knowledge of phenomenal reality, for example, the knowledge that every event has a cause.

TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

231

There are still many philosophers who accept the Kantian view, but it is fairly clear that his view cannot be held in its original form. The Euclidean view of physical space was rejected by Einstein and, indeed, the Einsteinian view that supplanted Euclidean geometry was confirmed by the very famous experiment performed by Arthur Eddington during a solar eclipse at the end of World War I. The story that Kant told about arithmetic also proved to be implausible, although it has had ingenious reinterpretations and defenses. Kant’s view stresses the essentially dynamic (hence, he thought, the synthetic) nature of arithmetic that is perhaps revealed in plotting data. It has nevertheless proved to be difficult to think of arithmetic as both synthetic and a priori. After all, even physical joining or mixing (an analogue of addition) is not always additive. Two quarts of alcohol and two quarts of water make fewer than four quarts of liquid. The idea that metaphysical truths or claims, like the claim that every event has a cause, are a priori is even more problematic. In the first place, whether or not a certain event is a cause of another seems to depend in part on the way in which the events are conceived or described. As Davidson stresses, events can be conceived in a variety of ways (viz. can be considered under various descriptions). On his view, how they are causally related (or even whether or not they are causally related) will depend upon the descriptions under which they are considered.1 What is even more important is that there are processes that appear to be genuinely random. For example, in the case of radioactive decay it is difficult to see what would count as the cause of the emission of a particular particle. Heisenberg’s Principle also suggests that there simply isn’t a fact of the matter about the location and momentum at any given time of very small particles. (Its not quite clear that this amounts to saying that the movements of electrons, for example, have a random component, but it certainly suggests indeterminism.) Many philosophers argue nonetheless vigorously for causal determinism, and perhaps they are right, but surely the matter is more complicated than intimated by Kant. None of this is meant to disparage Kant. After all, Kant’s philosophy made grand claims that withstood a century of criticism, and no doubt there is much truth in his philosophy. Indeed, the main argument of this paper has a Kantian feel to it. My point is only that Kant’s main idea (to wit, that we can discover, a priori, the structure of phenomenal reality) seems less plausible than ever, and it is important to remember that even before the revolution in physics that was brought on by Einstein and Heisenberg, many found the Kantian view difficult to accept. After all, how can the principles that describe the world be discovered without examining the world? Truths about the way we think appear to be truths about ourselves (viz., about the ways in which we are disposed to construe reality) and not necessary truths about the world (except, of course, to the extent to which we and our thinking are parts of the world). There appears to be logical space for the

232

JOHN H. DREHER

conjecture that we might completely misconceive the world and hold mistaken views about it. Perhaps just such reflections gave rise to psychologism in its earliest form. Kant himself would have been appalled by the suggestion that philosophy is merely an empirical discipline and would have completely rejected the idea that the truths discovered by philosophy are facts about the operations of our own minds. Although not all of us are favorably disposed to Kantian critical philosophy, it is unlikely that any of us shall find ourselves much happier than Kant with the suggestion that the truths of philosophy are essentially truths about the way we in fact think. As intimated, late nineteenth-century philosophers found themselves in a similar predicament, and they responded to it in much the way we do. They were neither open to Kantian critical philosophy nor receptive to the thought that philosophy is ordinary empirical science. Take logic, for example. We do not view it as a description of how we in fact think but rather of how we must refrain from thinking if we are to avoid inferences that might carry us from truths to falsehoods. It was just this sort of consideration that led Frege and Husserl to reject psychologistic accounts of logic. Reality constrains what can be intelligibly thought, and hence, asserted. Kantian and Berkeleyan idealists of the nineteenth century were just wrong in their understanding of mathematics and logic. Although early twentieth-century British philosophers like Russell and Moore rejected Kant, neither was ready to embrace psychologism. In fact, they not only held virtually the same view as Frege and Husserl about logic, but also tried for realist views about perception and even morality. Perhaps we cannot perceive a figure except as defined by a boundary of colors, and perhaps colors are not in the world but rather are of the mind, but that does not show that shapes are also objects of the mind. Some early twentiethcentury philosophers like Carnap and Russell, advocated sophisticated attempts to reconstruct the external world from the sense data, but they would have rejected the idea that the external world is essentially mental or that it is characterized by structures of the mind. This means that they accepted the idea that the immediate objects of consciousness and hence thought are mental, but rejected the Kantian notion that reality is constituted by mind. Kant could claim that by internal acts of mind we come to know the structure of the external world precisely because he thought that the phenomenal world is constituted by the mental. The work of twentiethcentury realism begins at the very point at which the work of nineteenthcentury idealism ends. For the twentieth century, the basic philosophical puzzle is to show how it is possible to have knowledge about the world on the basis of an inner life that is at best contingently connected to the world. The severity of this problem is illustrated by the fact that the twentieth century was obsessed by the problem of others minds and ultimately solipsism. Russell, Ryle and Wittgenstein all addressed this issue into mid-

TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

233

twentieth century. Twentieth-century philosophy rejected the a priori and resisted idealism. At first the turn towards realism seemed to reintroduce common sense (and indeed good sense) into philosophy. For example, consider your belief that I wrote these words. It is plausible to explain your belief by the fact that the objects of your belief (my actions) are in fact that causes of those very beliefs. This line of thought suggests that a claim is known and hence true if its object is its cause. In this way beliefs can be justified even though the connections between them and the world are contingent. Many philosophers have thought (or perhaps fantasized) that this familiar mode of explanation can be extended in obvious ways to accommodate more and more complicated cases, where, for example, causes are remote (i.e., only their effects — or effects of their effects — are observed). However, reconstructed empiricism appears to collapse when it comes to knowledge of entities that are unobservable, like numbers or values. Even some objects claimed to be causally active (like God) will be rejected by empiricists on the ground that they cannot be observed (i.e., in public, repeatable events) but are at best experienced by a few and, what is worse, in essentially private experiences. Perhaps many philosophers have been just as happy to see God go, who might after all have reminded them of an overbearing and possessive parent or spouse. It requires more fortitude to let numbers and morality go. Indeed, much of the twentieth century was spent struggling with the issues of mathematics and morality, blocks on which empiricism and realism have always stumbled. If mathematics is both a priori and, yet, applicable only as a matter of fact to physical reality, it was natural to conclude that mathematics itself, viz. pure mathematics, must be analytic. This answer seemed to suit philosophers of the thirties, like Carnap and Tarski. Mathematics and mathematical logic would in a sense structure thought but would not contain any genuinely substantive claims. Of course mathematics appears to make existence claims, but Carnap taught that those existence claims are ‘internal’ to the theories that they inform. Tarski even argued that as far as the requirements of logic are concerned, the applicability of the concept of truth itself to experience need not have any real content. The rejection of the a priori had ominous implications for ethics as well as mathematics; indeed ethics foundered in the rejection of the a priori, but in the case of ethics it proved difficult to beat a quick retreat to analyticity and an ‘internal’ sense of existence. Ethical claims are substantive, in just the way that analytic claims are not. Ethics directs action; to say that something it good, like showing kindness to strangers, just isn’t definitional. Denying the goodness of showing kindness is surely a mistake, but it is surely not a mistake of language. At outset of the twentieth century, born-again realists turned their attention to ethics. Moore, Prichard, Ross and even Russell all figure

234

JOHN H. DREHER

prominently in the revival. However, even they found it difficult to ignore persistent moral disagreement, and, after the First World War, even more difficult to believe that the moral views of their own culture were really so much more reasonable than others. A loss of confidence in Western moral supremacy marked an abrupt ending to the realist revival in ethics. It was only after the Second World War, late into the fifties, that it was yet again introduced by philosophers like Geach and Foot. But ethics has not regained its earlier confidence even to this day. The rejection of the a priori ultimately led to various forms of noncognitivism and anti-realism — noncognitivism being the doctrine that the strictly ethical content of ethical claims is really not descriptive at all but rather evocative or commendatory (e.g., Ayer, Stevenson and Hare), the anti-realist view being that ethical terms simply fail of reference or fail to have adequate criteria of applicability (e.g., Mackie and Schiffer). It is important to see that noncognitivist and anti-realist reactions to the decline of Kantian idealism and the rejection of moral realism are not themselves reversions to psychologism. To reconceive ethical discourse as essentially emotive or prescriptive is not to claim that ethics is a branch of empirical science involving knowledge of the operations of our own minds. The retreat to noncognitivism is a denial that ethical utterances are really claims of any kind and, a fortiori, claims about the operations of our minds. Likewise, anti-realism need not be psychologistic. It acknowledges that ethics has the form of assertoric discourse; it just denies that it is or can be successful. Anti-realism does not make a positive ethical claim. A reversion to psychologism in ethics would claim that ethics really is about us and that its claims really are true, at least to the extent that they accurately reflect, describe or perhaps express our propensities to evaluate. Whether or not mathematics can be understood psychologistically is less clear. If mathematical claims are, contrary to Kant, analytic, might we not say that they are true in virtue of the way in which we use language (viz. in virtue of their form and meaning)? Shouldn’t we then acknowledge that the propositions of mathematics depend for their truth upon the way in which we think? If so, wouldn’t the necessity of mathematics show that there is a certain kind of necessity in nature, discoverable by empirical means, to wit, a necessity about the way in which we think? This line of reasoning, plausible as it is, cannot be quite correct as it stands. We could not choose to mean what we mean by ‘2’ and ‘1’ and ‘=’ and ‘3’ and deny that 2 + 1 = 3 . It is the necessary character of mathematical truth keeps it from being a truth of empirical psychology. What makes a mathematical claim true is not that it is a correct description or expression of the way we think but rather that it is a genuine constraint on the way in which we can think. Despite all this, the twists and turns in twentieth-century philosophy that led to Tarski and Carnap in mathematics and logic and to Stevenson, Ayer, Hare, Mackie and Schiffer in ethics, also led to a revitalization of

TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

235

psychologism, which has re-emerged at the end of the twentieth century in a form more virulent and comprehensive than its nineteenth-century original. Twentieth-century psychologism is at the root of what has come to be called the post-modernist view of the world, which trumpets a radical from of cultural relativism that is both daring and mysterious. According to this line of thought, truths of mathematics, physics, as well as ethics (indeed truths of every sort) are relativized to culture. This view is the dominant philosophical view of our time, although we analytic philosophers were late in recognizing its ascendancy. We can see most easily see what led to the re-emergence of psychologism and its ultimate transformation into cultural relativism by examining the work of one of our own, namely Quine, who borrowed much from the philosophy of science of Duhem. In fact, there are three interrelated developments that have supported latter-day psychologism. The first is the recognition, emphasized by Duhem, that our beliefs about the world are radically underdetermined by our experience of the world. The second is what appears to be a corollary of Duhem, namely, that what we mean by the words we used is likewise underdetermined by our experiences of the world, i.e., our experiences of the objects to which our words refer. These two are familiar Quinean doctrines. The third is the relativization of truth itself to conceptual schemes and hence to cultures. It may seem that the third of these views, at least in its extreme form, is obviously false and even ridiculous, but I shall argue that it is not at all easy to refute — however ridiculous it may appear. In one of its forms (depending upon what is meant by the ‘relativization’ of truth to a conceptual scheme) the third view arguably follows from the first two. (That is because it is difficult to see how truth can be defined without referring to the languages in which truths are expressed.) In any case, I want to begin by discussing a fairly easy case, namely ethics, and shall go later to talk about more difficult and more important cases.

II Much of what I have to say about ethics has been inspired by Bernard Williams’s influential book, Ethics and the Limits of Philosophy, but I do not offer the following as an interpretation of his writing. In his book, Williams insists that there is a sense of knowledge in which it is possible for us to have ethical knowledge and for it to be destroyed by reflection.2 I want to begin my discussion of contemporary psychologism by focusing on Williams’s controversial claim and an example drawn from the ancient world to which Williams alludes. The ancient world was cruel in many ways that we find hardly comprehensible, much less defensible. Ancient Romans, for example,

236

JOHN H. DREHER

delighted in cruel spectacles that were staged not only in Rome but also in coliseums throughout the Hellenistic world. Those ancient events were surely indefensible. If any action is morally indefensible, it is causing another pain for one’s own pleasure. Yet there are also examples of ancient cruelty that are not as obviously indefensible. Take, for example, the practices of maiming and humiliating defeated enemies. Warfare in the ancient world was more personal than it is now, where we kill at a distance. In today’s world, defeated enemies can be controlled simply by denying them access to the instruments of modern technological warfare. In ancient times, however, a defeated enemy needed only a sword and a shield to regain parity with his captors. Defeated enemies needed to be controlled, and the easiest means to control was no doubt to disable them. Extreme forms of humiliation were also common in the ancient world, and perhaps had their role to play in dispiriting enemies. Think of Hector being dragged around Troy. Defeated enemies, of course, suffered far worse than humiliation. They were enslaved, a virtual death sentence, and even slaughtered. In ancient times young warriors were trained in the ways of cruelty. The suggestion that a defeated enemy should be treated mercifully would have been rejected as foolish and weak-minded and, even worse, as womanly. However defensible these ancient practices might have been by their own lights, they deeply offend modern sensibilities. We have conventions of war prohibiting the inhumane and degrading treatment of the defeated. They are not always observed, but that of course is beside the point. Our ideals are high, and we are especially outraged by practices that were commonplace in the ancient world. I think, for example, recent examples from Sierra Leone in which the hands, feet and ears of captives are severed, ‘to make a political point,’ as they say. Surely there is a disagreement between the ancient view of cruelty and our own, modern view. The philosophical question is how to characterize that disagreement. Realists will claim that there is a truth of the matter when it comes to ethics, and they will claim that we are right and that the ancients were wrong about cruelty, and perhaps this is so. It is difficult, however, to imagine a serious defense of realism in this context that does not have an a priori component. (I am obviously thinking of something along Platonist or Kantian lines.3) There is no doubt that philosophers who are willing to turn to the a priori have nothing to fear from relativism. Those less confident about the underpinnings of philosophical truth, however, will have a difficult time arguing from experience against the ancient view and for the modern view. Do we really think that there is a fact of the matter about the way in which defeated enemies should be kept, and that somehow we are in a better position than the ancients because we are better at finding out those facts?

TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

237

A different take is that the views of the ancients and moderns only appear to conflict. The facts of each era, it might be argued, make all the difference. Might not those facts justify the cheerful conclusion that there really isn’t a difference in the ancient and modern view of the proper treatment of defeated enemies? If we found ourselves in the circumstances of the ancient world, perhaps we would take up their view; on the other hand, if they found themselves in our circumstances, they would take up ours. Granted: Ancient warfare was intimate and fought up close. True: Ancients had better reason for crippling their enemies than moderns have, but it is important to remember that cruelty has its purposes in modern warfare as well. Can anyone really deny that there are occasions in which cruelty is useful in crushing a despised and fearful enemy even in the modern era of automatic weaponry? That, of course, is just what is happening in Sierra Leone. It may nonetheless be tempting to persist in the thought that ancient and modern views differ only in their intended range of applicability, and therefore do not really conflict. Indeed, it is unlikely that the ancients meant their views to be extended to modern contexts, whose ways of war would no doubt have eluded even the most imaginative ancient mind. Unfortunately, however, it is virtually certain that we, whose views about morality have been informed by the Enlightenment, intend our ideas about right and wrong, good and bad to be applied to all people everywhere of all times. This is surely true concerning the treatment of enemies; that is, we claim not merely that it is good for us to treat the defeated mercifully and to respect their human rights, but we claim that it is good for everyone, at all times. We think that our ways are better than the ways of the ancients; we think that moral practice has improved. I believe that our differences with the ancients about the treatment of enemies reveal even deeper differences about cruelty itself. The matter of public spectacles is important, because it is difficult to believe that a widespread practice can really be regarded as wrong by its entire culture. The cruelty that pervaded the ancient world gives evidence that the ancient world did not find cruelty to be offensive in the way we do. On the matter of cruelty, there is what Stevenson would have called a ‘rock-bottom attitudinal difference’ between us and ancient cultures. At any rate, I ask that you suppose it so, for the sake of my arguments concerning psychologism. What should empirically-minded philosophers, that is, philosophers who reject the a priori, say about cases like this? It is time to return to our three principles. Any belief, Quine insists, may be held come way may. This includes beliefs about cruelty. Ancient and modern claims about cruelty need to be considered within the context of all their moral beliefs, and their moral claims need to be considered within the wider contexts of their entire culture. It is not implausible to think that ancient moral beliefs about cruelty

238

JOHN H. DREHER

could have been maintained into our own era by redistributing truth values over other moral beliefs. Likewise, the ancients might have seen merits to mercy if they had had taken a different view of heroism and manliness. In this relativist sense, ancient belief about cruelty is a kind of knowledge. It is not unreasonable to suggest that cruel treatment of the defeated can be reasonably evaluated only within (that is, relative to) a wider cultural context or world-view. Reflection might destroy ‘knowledge’ about cruelty, because it might destroy the world-view in which that knowledge took root. It is important to see that even on this relativist line of thought, both we and the ancients are using ‘good’ in its usual commendatory sense. It is especially important to remember that we therefore really do differ about cruelty in differing about whether it is good or bad; that is, in commending different, incompatible actions. Relativism concedes that we differ in opinion but paradoxically claims that we both hold correct opinions, which is possible because our opinions are relativized to our respective worldviews. So far this discussion can be located within a fairly familiar if controversial and somewhat confusing discussion. And doubtless the cultural relativism about ethics defended by Williams is taken for granted by many. However, there is a form of cultural relativism that goes much further. It attempts to extend the relativist line of argument beyond ethics (where many find it difficult to see that there is any empirical evidence about the facts of the matter) to cases where there apparently are empirically discoverable facts of the matter, like natural science itself. Most analytic philosophers regard these attempts as wrong-headed, even preposterous. Williams stoutly denies that cultural relativism can be plausibly applied to science.4 However, once the ball gets rolling in the direction of relativism, it is more difficult to stop it than some might think.

III If any belief can be held come way may, including beliefs about the meanings of words, and even the meaning of the truth predicate itself, then a view might very well open us to that I shall call radical or full-blown cultural relativism, to wit, the view that the meaning of the truth predicate itself can be understood only relative to a culture. This would mean that knowledge about the nature of truth might be destroyed by reflection, in the sense that ethical knowledge can be destroyed by reflection. The first thing that I want to point out is how very difficult it is to take a completely non-relativist view of truth. Sentences (or a statements made by using sentences) are sentences or statements of languages. So, the truth of a sentence or statement is already defined relative to the language in which that sentence or statement occurs. Beyond that, there are genuine and

TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

239

substantial differences about how truth should be conceived. For example, we end up with different conceptions according to whether we hold a substitutional or referential view of the quantifiers. Truth, like any predicate, has an extension, and insofar as interpretations of a predicate assign it different possible extensions, the interpretations yield different notions. The nerve of my argument is this. If we may hold any view come way may, whether about fact or meaning, then we can hold any view about truth that we choose come what may, if only we are prepared to make sufficient adjustments to our remaining beliefs. I suggest that there are a cluster of relativist views about truth that can be held come what may. They can be defined by a common constraint. Each conception will require all its current beliefs to fall within the extension of its truth predicate. There are obviously two forms of this view. One is idiosyncratic, requiring truth to comprehend all one’s owns beliefs; the other is cultural, requiring the truth predicate to comprehend all the beliefs of one’s own culture. Views of this sort more or less come down to the assertion that whatever else I mean by ‘true,’ I mean all the things I (or we) now believe.5 Surely, you will remonstrate, this view, spectacularly bold and crude, cannot be taken seriously. Yet, I claim not only that it has been taken seriously by philosophers but also that it is now the dominant view of truth. Let’s begin by considering how full-blown cultural relativism might be defeated. Keep in mind that appeals to the a priori are not going to be allowed, because they will simply amount to the rejection rather than the refutation of psychologism. And keep in mind that ruling out appeals to the a priori does not lessen but rather increases the interest in the argument — after all, Quine and Williams are hardly inclined to resort to the a priori. One problem with full-blown cultural relativism might be that it will open the door to massive recalcitrant experience. If we constrain our truth conception so that everything we now believe counts as a truth, our conception of truth will need to be continually revised because our beliefs will stand in need of continual revision, for the obvious reason that not all our current beliefs really are true. This means that the extension of the truth predicate will be changed retroactively and hence continually redefined. I myself think that this objection would be absolutely devastating if we could mount a good defense of the necessary premise, namely that full-blown cultural relativism has and is likely to continue to meet with massive recalcitrant experience. But how do we know that? What complicate the argument are Hume’s problem and the prohibition on the appeal to Kantian resources to get around it. Keep in mind that it does not matter how badly full-blown cultural relativism fared in the past. If there are neither arguments from experience nor from reason that suggest that the past is indicative of the future, then past failures cannot be deemed to predict future failures. Quine himself faces up to this problem. He claims that empirical science

240

JOHN H. DREHER

endures not because we must hold to it, but because we are disposed to hold to it, despite recalcitrant experience. The criteria that make empirical science a reasonable activity do not include a guarantee or even promise of continued success. Rather they are the criteria of conservativeness and simplicity.6 Perhaps that is what is wrong with full-blown cultural relativism: It just isn’t conservative or simple, but we should ask: Why should we accept conservativeness and simplicity as criteria for theory revision? If they lead to the truth, we have good reason, but that conservativeness and simplicity lead to the truth is just what we do not know, and if Hume is right, cannot know. Conservativeness and simplicity are science-friendly, but then why should we seek criteria that are friendly to science if we do not know that science will yield true results? We like science, and perhaps have gone so far as to incorporate its methods into our very conception of rationality. Are we in a position, however, to criticize others who have a more relaxed view of rationality that perhaps emphasizes ‘intuitive, unscientific’ methods or even those who have authoritarian standards of truth-seeking that rely upon ancient texts? Perhaps certain cultural relativists will favor criteria that reinforce ‘traditional’ beliefs rather than those that validate the cold, hitherto reliable, findings of indifferent empirical science. Another objection to cultural relativism might be that its conception of truth is so remote that it just doesn’t come close to according with ordinary usage. Perhaps, you are thinking at this very moment, that although many philosophers have accepted full-blown cultural relativism and is strange notion of truth, not one of them is using the word ‘true’ correctly. But, we need to be careful here. Cultural relativism surely does not use the word ‘true’ in the way we analytic philosophers do, but it does not immediately follow from that fact that they use it incorrectly. We are inclined to think of it — to use an old-fashioned phrase — as a meaning postulate that our conception of truth must be independent of belief, since our ideal that governs the formation of belief is to seek the truth. If so, the fact that we believe something surely cannot be among the criteria (except in the weird, exceptional case) for identifying the truths that we seek. However, this argument presupposes what the radical cultural relativist is determined to deny, namely, that there can be criteria of truth-seeking that are not subjective, that is, that are in principle separable from our beliefs and dispositions to believe. That, obviously, leaves us in an awkward position. What objection do we have to cultural relativism that does not presuppose a positive answer to Hume’s question or the stipulation of an ideal that is impossible according to the relativist? Two answers come to mind. Unfortunately, while the first may appeal to empirically-minded philosophers, it is inconclusive. The second is more nearly conclusive, but it is very unlikely to appeal to empirically-minded philosophers.

TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

241

IV One thing that full-blown cultural relativism cannot do is to question coherently its own notion of truth. Within any full-blown cultural relativism, R, the sentence ‘The criterion of truth of R is false in R’ will itself be false in R because a constraint on every theory of truth in R is that it includes all the beliefs of R, and therefore, itself. (Recall that the criterion of truth of R is the defined to include all the current beliefs of R.) This is the sort of point that empirically-minded analytic philosophers like to make, and it is a good point, but it is not conclusive. It does not rule out versions of cultural relativism that simply refuse to screen for objectionable forms of selfreference, perhaps on the ground that the disposition of theories to question their own notion of truth is exactly a request for transcendence that is impossible to achieve. The second argument against cultural relativism is more nearly conclusive. It goes much deeper than the first, but it will be of little comfort to most empirically-minded philosophers, who deny that there is a transcendent perspective from which theories may be properly assessed. This second argument begins by considering the reaction we should have if faced by two perfectly well confirmed, complete (as can be) but incompatible theories of the world. One reaction (the distinctively positivist) reaction would be to claim that the competing theories are really one, i.e., that they are empirically equivalent and hence that they mean the same and really are the same.7 A second reaction would be to accept a frank dualism. This would be the response that both theories are true. Since we have described this as a frank dualism, perhaps we should add, ‘but true of different things.’ It is a brave response, and it does not multiply entities (i.e., worlds) without reason; indeed, it increases them for the apparently good reason that we need an extra world so that each perfectly well-confirmed, complete-as-can-be theory will have something to be true of. I think that an obviously better response to this thought-experiment is more modest. Faced with two perfectly well confirmed, complete-as-can-be but incompatible theories of the world, I think that it would be time to respond modestly, to concede simply that there isn’t an adequate basis for choosing between the two and that we just do not know which, if either, is true. This, of course, is not a response open to full-blown cultural relativism, for there is ex hypothesi, nothing for this modest belief to be relativized to. The radical cultural relativist and positivist may therefore find themselves denying that there is even a genuine thought to be articulated in this case. Indeed, my response to the thought experiment is also a brave one in that it presupposes a transcendent notion of truth (not relativized to any theory) and a transcendent perspective from which theories can be evaluated, especially perfectly well confirmed, complete-as-can-be, incompatible

242

JOHN H. DREHER

theories. It will be objected that it is only by a priori means that we could know that there are transcendent perspectives. I claim, however, that there is empirical evidence for the existence of a transcendent perspective, and my evidence is that we have just made a judgment from that perspective, but not everyone, I concede, will find that so obvious.8 Fortunately, there is a further argument. Suppose T and T* are perfectly well confirmed, complete-as-can-be, and incompatible. Denying that there is a transcendent perspective from which to view their comparative merits seems to require us to invent a further theory T* in which to claim that there isn’t adequate reason to choose between T and T* and to state that we do not know which of the two, if either, is correct. Moreover, T*, if it is to be perfectly well confirmed and complete-as-can-be, will need to include T and T’ and hence their claims. In that case, however, T* itself will not be perfectly well confirmed much less true, as the claims of T and T’ conflict by hypothesis. This shows that those who want to maintain that the proper response to the ‘two-perfectly-wellconfirmed theories-thought-experiment’ is the modest response (that is, that we do not know which of T and T’ is true, if either) will not be able to find a wider, perfectly well confirmed theory T* in which to make that judgment. They consequently will be driven to the conclusion that that there is a transcendent perspective (that is a perspective that has not been relativized to any particular theory) from which to evaluate perfectly well confirmed, complete-as-can-be theories, which, ex hypothesi, exist as a matter of fact. Of course those who think that truth is necessarily relativized may flatly deny that there is any way at all of articulating what I have called the modest response without a new and more inclusive theory, namely T* — but I do not think that there denial would be plausible. After all, we have just articulated the modest response without actually constructing T* or anything remotely like it. In other words, we may reasonably suppose that there is a transcendent perspective without actually having access to it.

V In this paper I have tried to sustain three points. The first among them is that psychologism is still a serious philosophical position and enjoys wide acceptance when it comes to fields like ethics. Second, although relativism has been rejected in its extreme form, which I call full-blown cultural relativism, it is not obvious just how full-blown cultural relativism is to be countered. Third, I have suggested that full-blown cultural relativism cannot account for an important thought-experiment that I claim requires a transcendent notion of truth. I concede that many important philosophers will not be impressed or happy with this response, but they might find themselves unable to avoid radical relativism without resorting to positivism

TRUTH AND KNOWLEDGE IN PSYCHOLOGISM

243

or to an implausible dualism. Department of Philosophy University of Southern California, Los Angeles, USA

NOTES 1 2 3

Davidson (1966), 93. See also Davidson (1967; 1980). Williams (1985), 167-69. By this I do not mean to ignore attempts that purport to be objectivist (or at least nonsubjectivist) but that refuse to resort to the a priori. I am thinking here of views like the ones presented by Wiggins in such papers as “Truth as Predicated of Moral Judgments” in Wiggins (1987), or Scanlon (1998). Of course, each of these writers refuses to be drawn into the realism/anti-realism controversy. 4 Williams (1985), 135-55. 5 In “Truth as Predicated of Moral Judgments”, 151f, Wiggins specially excludes this extremely relativist view of truth as predicated of moral judgments on the theory that it does violence to the very meaning of our notion of truth. Many other writers take a similar view. Whatever the merits of this approach, it does not, without further argument, refute a relativist notion of truth; it appears to reject it on the ground that it is obviously absurd. 6 Quine (1960), 20f. 7 Quine seems to be sympathetic to this line of thought in (1990), 95-101. 8 Perhaps here our concept of empirical evidence has been strained beyond the limits of its application. There is, however, a weaker notion to which to retreat, a pragmatic, regulative principle of inquiry. On this view, we might argue that rationality requires that we assume an ‘objective’ point of view or perspective in order to consider competing world-views, even though that perspective (that is, the propositions that describe its relation to the competing world-views) cannot be incorporated into a further world-view. In this way we might strive to take account of conflict that cannot be settled by ordinary standards. Rationality, some may argue, is better served in this way than it is by a silence that surrenders the debate to radical cultural relativism.

REFERENCES

Davidson, Donald. “The Logical Form of Action Sentences”. In Nicholas Rescher, ed., The Logic of Decision and Action. Pittsburgh: University of Pittsburgh Press, 1966. Davidson, Donald. “Causal Relations”, The Journal of Philosophy, 64, 1967; rpt. in Davidson, Essays on Actions and Events. Oxford: The Clarendon Press, 1980. Quine, W.V.O. Word and Object. Cambridge: The MIT Press, 1960. Quine, W.V.O. Pursuit of Truth. Cambridge: Harvard University Press, 1990. Scanlon, Thomas. What We Owe Each Other. Cambridge: Harvard University Press, 1998 Wiggins, David. “Truth as Predicated of Moral Judgments”. In Wiggins,

244

JOHN H. DREHER

Needs, Values and Truth. Oxford: Basil Blackwell, 1987. Williams, Bernard. Ethics and the Limits of Philosophy. Cambridge: Harvard University Press, 1985.

DALE JACQUETTE

PSYCHOLOGISM REVISITED IN LOGIC, METAPHYSICS, AND EPISTEMOLOGY

1. Psychology and Psychologism

The word ‘psychologism’ historically has meant many things, but has mostly been considered a term of abuse leveled against efforts to explain philosophical concepts or address philosophical problems from the standpoint of subjective psychological experience. The same degree of opprobrium, notably, does not attach to psychology itself, which has generally been regarded as a legitimate subject of scientific and philosophical investigation. This is true even for opponents of conventional psychological theories, ranging from Cartesian dualism to phenomenology to behaviorism, computationalism, and other branches of cognitive science. The trouble, according to critics, occurs specifically when the attempt is made to turn psychology of any form and following any methodology into a philosophical ideology, whenever and to whatever extent psychology becomes an ‘ism’.1 2. Kant’s Anti-Psychologism in Logic and Ethics

Although the mere mention of psychologism in some circles is enough to discredit a philosophical program, there does not seem to be any single objection universally raised against every version of psychologism in every philosophical subdiscipline. It may nevertheless be fair to say that what concerns many theorists about psychologism is that it appears to threaten the objectivity of philosophical reasoning. Kant’s criticisms of empirical considerations in any aspect of what he calls ‘formal’ as opposed to ‘material’ rational knowledge are historically representative of this attitude. In Foundations of the Metaphysics of Morals, Kant proposes to purify logic and ethics alike of ‘empirical’, by which he clearly means to include 245 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 245-262. © 2003 All rights reserved. Printed by Kluwer Academic Publishers, the Netherlands.

246

DALE JACQUETTE

psychological, elements. Following the ‘ancient Greek’ Stoic philosophy which he takes as his model, Kant first distinguishes between the three philosophical ‘sciences’ of physics, ethics, and logic. Then he argues: Logic can have no empirical part — a part in which univeral and necessary laws of thinking would rest upon grounds taken from experience. For in that case it would not be logic, i.e., a canon for understanding or reason which is valid for all thinking and which must be demonstrated...All philosophy, so far as it is based on experience, may be called empirical; but, so far as it presents its doctrines solely on the basis of a priori principles, it may be called pure philosophy. The latter, when merely formal, is logic; when limited to definite objects of understanding, it is metaphysics. (Kant 1959, 3-4; see Beck 1960, 53-5)

Kant in this passage sounds all the major themes of anti-psychologism. Logic must be valid for all thinking, and hence cannot be derived from empirical study of particular subjective experience. A pure as opposed to applied philosophy, according to Kant, must identity and try to articulate exclusively a priori principles that are grasped independently of empirical contingencies as universally true. Kant has undoubtedly put his finger on what is troublesome about psychologism to most of its critics. Philosophy has an a priori conceptual subject matter that is adulterated and compromised by the intrusion of empirical and especially subjective psychological factors and methodologies. Kant then extends the same argument to pure moral theory, considered as a branch of formal philosophy. He asks: Is it not of the utmost necessity to construct a pure moral philosophy which is completely freed from everything which may be only empirical and thus belong to anthropology? That there must be such a philosophy is self-evident from the common idea of duty and moral laws. Everyone must admit that a law, if it is to hold morally, i.e., as a ground of obligation, must imply absolute necessity; he must admit that the command, “Thou shalt not lie,” does not apply to men only, as if other rational beings had no need to observe it...He must concede that the ground of obligation here must not be sought in the nature of man or in the circumstances in which he is placed, but sought a priori solely in the concepts of pure reason... / Thus not only are moral laws together with their principles essentially different from all practical knowledge in which there is anything empirical, but all moral philosophy rests solely on its pure part. Applied to man, it borrows nothing from knowledge of him (anthropology) but gives him, as a rational being, a priori laws. (Kant 1959, 5)

Kant finally rejects empirical elements not only from ‘pure’ philosophy, but from philosophy per se. Philosophy must be altogether a priori in its underlying principles, or it does not deserve to be distinguished from empirical science. In order to assure that philosophy is worthy of its name, Kant concludes, it must have nothing whatsoever directly to do with empirical psychology, or, as he prefers to say, ‘anthropology’:

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

247

Philosophy which mixes pure principles with empirical ones does not deserve the name, for what distinguishes philosophy from common rational knowledge is its treatment in separate sciences of what is confusedly comprehended in such knowledge. Much less does it deserve the name of moral philosophy, since by this confusion it spoils the purity of morals themselves and works contrary to its own end. (Kant 1959, 6)

Here, then, is a Kantian template for anti-psychologism. We require of philosophy an objectivity and universality that is incompatible with a methodology that begins with or tries to justify its conclusions by virtue of their agreement with psychological factors. As Kant revealingly says, we do not want logical or moral principles that apply only to human beings, but to all rational thinkers, in principle if not in fact. Even if there are no angels, demons, gods, a monotheistic God, or rational extraterrestrials, the principles of philosophy should be chosen and established according to Kant by a distinctively philosophical method as if, counterfactually, there were nonhuman rational agents to which such principles would also need to apply. 3. Psychologism in Kant’s Anti-Psychologism

We do not need to pursue this line of reasoning very far in order to see that there is something disingenous about Kant’s rejection of psychologism. We should recall that Kant’s own project in the Critique of Pure Reason begins with what he calls the Transcendental Aesthetic, which takes its starting place in sensation, in the experience of space, time, and objects within the sensible manifold, and interprets space and time as pure forms of intuition (Kant 1965, A19-B73). Experience and intuition are obviously psychological concepts. So, it is not clear whether or in what sense Kant’s critical idealism is supposed to be nonpsychological or nonanthropological. A defender of Kant may want to say that Kant’s transcendental method is intended to uncover synthetic a priori principles that would apply, in the case of the Transcendental Aesthetic, to any perceiver and any rational agent, regardless of their human or nonhuman psychological nature. But, regardless of the aspirations of critical idealism, I think that in fact Kant cannot possibly have access to the existence, structure, or other attributes of sensation, or basis for checking on his conclusions, beyond his own subjective psychological experience, and what he can infer about the subjective experience of other human beings from their reports, filtered always through the prism of his own subjectivity (see Dorschel 1992; Bowie 1990; Hubbert 1993; Carr 1999). The same general point might be made about anti-psychologistic objections to psychologism generally. I shall have more to say later about psychologism in other areas of philosophy. But to give an indication of the

248

DALE JACQUETTE

direction I want to go, I propose to begin by casting a more critical light than I have so far attempted on Kant’s objections to psychologism in logic. Kant complains that logic would not be logic if it were in any way empirical, in the precise sense of trying to make ‘universal and necessary laws of thinking...rest upon grounds taken from experience’. Logic, Kant declares, must provide ‘a canon for understanding or reason which is valid for all thinking’. Yet again, do we need to emphasize that understanding, reason, and thinking are all psychological faculties or occurrences? Why do we want to systematize logical inference in the first place, if not as a set of standards for guiding reasoning in our thinking psychological lives? It is a commonplace that logic is not psychology, because logic is normative or prescriptive, providing in Kant’s words a valid ‘canon’ for reason or understanding, while psychology is descriptive of how thinkers actually reason. Yet even this hackneyed characterization of the distinction between logic and psychology will not bear close scrutiny. Do we not want to say that logic is at least descriptive of how some people sometimes think? If not, where could the normative standards in the logical canon possibly come from? A logic treatise or textbook in some ways can be seen as a description of a certain process of reasoning that its author has worked through, and recommends to others, at least in the activity of writing it, and in that sense documents an episode of the author’s psychology. How, then, aside from the fact that such pronouncements have become well engrained in philosophical consciousness, can we nod assent to Kant’s assertion in its full generality, that ‘Logic can have no empirical part’? If logic is the theory of how we ought to think, it is no less a theory of how we ought to think, based on or judged to be compatible with how, when we at our best, we actually do think. A normative or prescriptive conception of logic must minimally depend on assessments of how we are capable of thinking, determined empirically by experience of how we sometimes think, at least when we are teaching logic and writing logic textbooks. On what other conceivable foundation, and with what other imaginable data, could the study of logic conceivably be based? In ethics, too, where Kant proposes to eliminate empirical influences, and to establish moral philosophy without taint of psychology, anthropology, or facts about ‘the nature of man’, Kant formulates the categorical imperative in undisguisedly psychological terms. Despite the apparent variety of different versions of the principle, Kant writes that: “There is, therefore, only one categorical imperative. It is: Act only according to that maxim by which you can at the same time will that it should become a universal law” (Kant 1959, 39). Here, in Kant’s denunciation of psychology in ethics, we discover a clue to revisiting psychologism in all fields of philosophy. For, although Kant plainly wants to divorce moral philosophy from psychological factors involving pleasure, pain, happiness or unhappiness, and, in a word, inclination, he explicitly

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

249

includes reference to the psychological concept of will, and the maxims it is possible or not possible for a rational moral agent to will as universal law. Nor does it help to argue that Kant’s categorical imperative in referring to what an agent can or cannot will is speaking nonpsychologically about logical contradiction or noncontradiction. Anyone who has tried sympathetically to work through Kant’s four examples of applying the categorical imperative, in forbidding suicide, mandating the repayment of debt, developing one’s natural talents, and contributing charitably to the welfare of others in greater need, will appreciate the loose sense in which Kant must speak of contradiction in effecting to squeeze concrete implications out of the categorical imperative. Kant invariably falls back on a comparison of psychological acts of will whose objects are incompatible in the sense that they cannot be jointly realized, in order to argue that an agent cannot consistently will the universalization of a morally indefensible maxim (see Harrison 1957; Paton 1958; Swoyer 1983; Doore 1985). I have considered Kant’s anti-psychologism at length because of its historical influence, and because he neatly encapsulates the main objections to psychologism. As Martin Kusch, in Psychologism: A Case Study in the Sociology of Philosophical Knowledge, has demonstrated, Kant is the source for much of the anti-psychologistic furor that swept the German academy in the nineteenth and early twentieth centuries (Kusch 1995). Kant’s antipsychologism created an atmosphere of hostility toward psychological concepts in philosophy that also indirectly shaped contemporary AngloAmerican analytic thought via the logic of Frege and Wittgenstein. A similar response occurs in the anti-psychologistic wing of contemporary continental philosophy, through Edmund Husserl’s later transcendental phenomenology, especially in his objections to what he perceived as the latent psychologism in Brentano’s early immanent intentionality thesis in his 1874 Psychologie vom empirischen Standpunkt, and in the anti-psychologistic tendencies of Martin Heidegger’s phenomenological existentialism (see Rath 1994; Mohanty 1999, esp. 12-31, 79-109). I think it is significant that even Kant, despite his anti-psychologistic polemics, does not entirely free his own philosophy, even in its supposedly pure formulation, from psychological data, psychological concepts, and psychological principles. I am not trying to argue that Kant’s antipsychologism at the root of two main anti-psychologistic traditions is internally incoherent because it cannot consistently be put into practice. I think that Kant’s objections, like most carefully thought-out criticisms of psychologism, are cogent and make an important point about the limitations and disadvantages of at least certain types of psychologism. What I am prepared to argue instead is that Kant’s objections to psychologism, as representative of the suspicions of many critics of the psychological turn in philosophy, cannot be consistently interpreted as preaching and practicing the elimination of all psychological elements from philosophy, from what

250

DALE JACQUETTE

Kant refers to as the formal philosophical disciplines of logic and metaphysics, and especially from epistemology and ethics. Kant’s anti-psychologism, as I understand it, can only be consistently interpreted in the context of his philosophy as a whole, not as an objection to any and all psychological factors in philosophy, but as distinguishing between acceptable versus unacceptable psychological factors. Ultimately, I am inclined to say the same about any anti-psychologistic stance, whether or not it is directly inspired by Kant’s. There are, so to speak, good psychological considerations, that have a legitimate place in philosophy, and bad psychological considerations, that are unpromising and in specific ways inadequate as philosophical principles. We need to distinguish between good psychologism and bad psychologism; between psychology that is good for philosophical purposes, and psychology that is bad for philosophical purposes. In the case of Kant’s ethics, for example, inclination is a bad, harmful or theoretically useless psychological concept, while will is a good, helpful or useful psychological concept for the specific type of moral philosophy Kant is trying to develop.2 The same can be said, I believe, for anti-psychologistic criticisms of psychological factors in logic, metaphysics, and epistemology. Thus, I am ready at last to issue a manifesto for revisiting psychologism, sorting out good from bad psychological factors in all three of these philosophical subdisciplines. 4. Good and Bad Psychologism in Logic

I shall begin with logic. It is not hard to find bad psychologism in this field. A concise statement of ways in which psychology should probably not be combined with logic is given by Alexander Maslow in his valuable commentary, A Study in Wittgenstein’s Tractatus. Maslow writes: We must also understand that, while logic deals with symbols, it is not concerned with the actual psychological processes which may be going on in our minds when we deal with symbols. Identification of psychology with logic is due to the fallacy of psychologism: we find that in understanding certain logical principles we have to go through certain psychological experiences, and we conclude that the two are one and the same. For logic this coincidence is accidental and irrelevant. Many logicians were guilty of this fallacy; Boole held that logic dealt with ‘the nature and constitution of the human mind’. But for Wittgenstein logic ‘results from the essence of the notation’ (3.342), and not from the kind of material out of which the signs of our symbolism may happen to be made. Perhaps psychology may happen to have a special epistemological interest for us, but in a discussion of logic it has no privileged position. To give it a privileged position is a sheer metaphysical anthropomorphism. ‘Psychology is no nearer related to philosophy, than is any other natural science’ (4.1121). The fallacy of psychologism in logic is due to the confusion of the validity of thought with certain logically accidental experiences always associated with correct thinking. A genetic account of

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

251

logic is entirely irrelevant to the essence of logic. (Maslow 1961, 55)

The ‘fallacy’ of psychologism, according to Maslow, is the identification of logic with psychology or ‘certain psychological experiences’ that are irrelevant to the validity of inference, when logic is rightly interpreted as a system of symbols. This type of psychologism admittedly seems philosophically wrongheaded. But just because logic is not properly understood as identical with any and all actual thought processes does not mean that logic cannot be understood as refined and abstracted from actual thought processes in a philosophically more respectable type of psychologism. It is in this spirit that George Boole, whom Maslow associates with the psychologistic fallacy, in An Investigation of the Laws of Thought on Which Are Founded the Mathematical Theories of Logic and Probabilities, explains that: “The design of the following treatise is to investigate the fundamental laws of those operations of the mind by which reasoning is performed; to give expression to them in the symbolical language of a Calculus, and upon this foundation to establish the science of Logic and construct its method...” (Boole 1959, 1). To me, this does not sound so bad; it does not strike me as irredeemably bad psychologism. Bad, really bad, psychologism in logic in the end might be nothing more than a straw man. Boole does not propose to describe how any particular choice of thinkers actually think, and then try to pass this off as logic, as the exaggerated criticisms of anti-psychologists would often like us to believe. Even psychology as a general science is not limited to individual clinical histories, but strives to identify such laws as the law of association or the principle of negative reinforcement. What Boole states as his method is similarly to empirically study thinking, specifically with an eye to abstracting the logical laws of deductively valid reasoning and the theory of probability. Thus, he writes: “Like all other sciences, that of the intellectual operations must primarily rest upon observation, — the subject of such observation being the very operations and processes of which we desire to determine the laws” (Boole 1959, 3). If Boole’s method were simply to codify reasoning in psychological experience, however he happened to find it, and present that mishgamuh as a system of logic, then his efforts would certainly amount to bad psychologism. But that is not what he does, nor what he says he is going to do. It is not for nothing, after all, that Boole’s research has been a foundation stone for contemporary mathematical logic and probability theory. The system he presents does not look very different from other pioneering formalizations of logic in the mid-nineteenth century, including those of logicians like Frege who are more self-consciously antipsychologistic. Does Boole, then, simply misspeak himself in referring to his investigations as a study in the laws of thought? Is Boole a good logician, but a bad philosopher of logic? I would say, on the contrary, that

252

DALE JACQUETTE

Boole is right to describe his target as the laws of thought, provided that with proper emphasis we understand his project as one of looking into the laws of thought, and not the accidental facts of psychology. If Boole’s goal were merely to assimilate psychological facts irrespective of logically correct reasoning, then it would not make sense for Boole to maintain that, having derived the principles of logic from the study of thought, he will finally use his system of logic in order to better understand the facts of psychology. For, he says that it is part of his purpose, after his logic and theory of probability are in place, “to collect from the various elements of truth brought to view in the course of these inquiries some probable intimations concerning the nature and constitution of the human mind” (Boole 1959, 1; see Musgrave 1972; Jacquette 2002, esp. 77-91). Boole, who for too long has been the poster boy for bad psychologism, is hard to categorize as guilty of this offense. There may well be logicians who have espoused bad psychologism. Yet their mistakes do not detract from the good psychologism of theorists like Boole who are engaged in a philosophically respectable project of discerning as laws of thought whatever is general and essential to logically correct inference in the empirical study of reasoning. Where else can or should logic begin? To deny without qualification that logic has or should have anything to do with psychology, as in most popular anti-psychologistic rhetoric, surely goes too far. That is the sort of extreme position represented by formalism, which to a large extent has been motivated by the desire to avoid psychologism in the philosophy of logic. But formalism is a controversial interpretation of logic, insofar as it presupposes the gameslike rule-governed manipulation of meaningless symbols or uninterpreted syntax. It is unclear whether the concept of a meaningless symbol, a symbol that does not symbolize anything, is itself logically coherent, or whether tokens or types of marks on paper or similar objects could possibly qualify as syntax if or insofar as they are altogether uninterpreted.3 Even Wittgenstein in the Tractatus Logico-Philosophicus, whom Maslow portrays as vehemently anti-psychologistic, does not regard any and every structure of potentially designating facts as genuine syntax. Identity of logical form or mathematical multiplicity, according to the picture theory of meaning, is a necessary but not sufficient condition for fact-picturing syntax; otherwise, the world would as much be a picture of language as language is supposed to be a picture of the world. Wittgenstein rejects bad psychologism, as Maslow recognizes, perhaps nowhere more harshly than in his 1913 review of Peter Coffey’s book, The Science of Logic (Wittgenstein 1913). But Wittgenstein in Tractatus 3.11 also insists in what I consider to be his good anti-formalistic psychologism that: “We use the sensibly perceptible sign (sound or written sign, etc.) of the proposition as a projection of the possible state of affairs. The method of projection is the thinking of the sense of the proposition” (Wittgenstein 1922). If formalism

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

253

is diametrically opposed to any type of psychologism, including what I would call good psychologism, and if it is the only real alternative even to the philosophically least objectional psychologism, to psychologism on its best behavior, then I personally have no special difficulty discounting formalism as a challenge to psychologism on grounds of material inadequacy. The problem affects all efforts to eliminate psychological factors from philosophical semantics by eliminating propositions as abstract entities in favor of properties and relations. It may be possible to eliminate propositions in favor of properties and relations, so that the meaning of a sentence token or type or statement made by means of a sentence is interpreted as representing relations of inclusion in and exclusion from the extensions of predicates. But such a reduction will not have purged all psychological elements from semantics if, as I am prepared to argue, a sentence, word, or other meaningful unit of discourse can only be understood as the expression of a thought, concept, or idea. Nor, for obvious reasons, is the situation improved if propositions are understood merely as truth value bearing sentences. Semantics presupposes ontology, but is not reducible to any ontology that does not include psychological, and especially intentional, properties and relations (see Heil 1981, 334; similar themes about the semantic impotence of ‘lifeless syntax’ in the later Wittgenstein are developed by McGinn 1984, 17, 117-19). 5. Good and Bad Psychologism in Metaphysics

I have devoted so much of my discussion to the problems of psychologism in logic, because this is the application which anti-psychologists have found most objectionable. It is therefore the area in which anti-psychologists have been most active, where they seem to believe psychologism is most vulnerable, and in which it may be most important for psychologism to be defeated. But there is good and bad psychologism also in metaphysics, some conspicuous examples of which I want to examine, in order to explain how psychologism can go wrong, and what prospects there may be for a philosophically respectable partnership between psychology and metaphysics. An idealist metaphysics of the sort that G.W.F. Hegel championed in imagining that the world as spirit or self-consciousness is developing in distinct successive dialectical stages toward realization of the Absolute is perhaps as good an example as we need of the potential excesses of bad psychologism in ontology (Hegel 1968; see Hager 1974; Navickas 1976). George Berkeley’s anti-materialism is probably next in line, interesting as it is in its own way to professional philosophers. By defining all existence as psychological, in the slogan, Esse est percipi aut posse percipere, Berkeley

254

DALE JACQUETTE

is committed in what he evidently sees as an advantage of his theory over godless materialism, to the existence of God as an ‘infinite mind’ that continuously perceives all sensible things that are only intermittently perceived by such finite minds as ourselves. The infinite mind of God is supposed to guarantee objectivity for the world of sensible entities construed as congeries of ideas, ectypally in human minds, and archetypally in God’s infinite mind, in every dimension and from every perspective. Berkeley’s psychologistic metaphysics has never attracted a serious following. Yet his idealism is an inevitable consequence of his uncompromising empiricism, in the desire to limit knowledge to sense experience. Berkeley in this way is led to conclude that all of reality exists psychologically, idealistically, in our minds and in the infinite mind of God (Berkeley 1949). We might also think naturally in this context of Spinoza in the Ethics, first defining the concept and then proving the existence of God as coextensive with the universe (Spinoza 1914). Or, consider the teachings of the old Stoa, of Zeno of Elea and Chrysippus, arguing that godhead and the universe are identical, that the universe is a rational being with a rational purpose, and that ethics and politics can enable us to achieve tranquility by bringing our own desires and expectations into agreement with whatever events happen to transpire in the world as an expression of divine will (Zeno of Citium 1973, ‘Introduction’, 13, 41-5; see Schofield 1983). The Stoics supported this psychologistic conclusion in metaphysics, among other ways, by arguing in an inference also examined by Cicero in De Natura Deorum, that whatever is rational is greater than whatever is not rational, that nothing is greater than the universe, and that therefore the universe is rational (Cicero 1972, 201; Jacquette 1995). David Hume, in A Treatise of Human Nature, is as committed to empiricism or the experimental method in the moral sciences as Berkeley. But without sharing Berkeley’s theistic preoccupations, Hume reaches more skeptical conclusions about the impossibility of philosophically demonstrating the existence of an external reality independent of thought, or of what Hume calls perceptions, of impressions of sensation and reflection, and of their faint psychological traces that remain as ideas. Hume takes up the slack left by the limitations of an empiricist philosophy by turning from reason to the passions, and affirming that where reason leaves off the imagination is compelled to accept the existence of an external world. Here, also, insofar as Hume can be described as a metaphysician, he tenders a psychologistic metaphysics, based on the mind’s faculties of perception and imagination.4 Kant, again, in the Prolegomena to Any Future Metaphysics That Will be Able to Come Forward as Science, criticizes the rationalists as lacking a scientific method in metaphysics, and attacks empiricists like Berkeley and Hume as too skeptical, in the absence of collaborating sense experiences, about the existence of a mind-independent external world, the

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

255

necessity of causation, and the existence of the self as a unified subject of experience (Kant 1977, 2-8). Kant hopes to overcome these limitations by defending transcendental reasoning as a method to recapture some of the prized conclusions of the rationalists while satisfying the empiricist’s concern that pure reason divested of experiential content is inadequate in the search for truth (Kant 1977, 12-24, 69-90). And we have already seen that transcendental reasoning in Kant’s critical idealism is, in the most general sense, as I define the concept, despite Kant’s disclaimers, equally psychologistic (see Jansohn 1969; Oliveira 1973; Jerrius 1984; Hubbert 1993). Philosophers who are attracted to Hegel’s metaphysics, or to Berkeley’s, Spinoza’s, or Zeno’s, and recognize them as psychologistic, will naturally want to argue about whether their psychologisms are good or bad. For present purposes, I do not much care how they are evaluated, as long as the distinction between good and bad psychologism is taken seriously. I am offering my opinions, if you like, by way of illustration. But to continue in this vein, if Hegel’s and Berkeley’s philosophies exemplify bad psychologism, what sort of good psychologism, if any, might we say has a place in metaphysics? From what I have argued so far, perhaps Hume’s, perhaps Kant’s. I am not supposing that good or bad psychologism in metaphysics or the other branches of philosophy needs to take any single form. I shall not try to develop the idea very much further here, except to point to what I find very positive in Nicholas Rescher’s discussion of ontology in his recent study, Objectivity: The Obligations of Impersonal Reason. Rescher defines objectivity, not as mind-independence in any absolute sense, but as an abstraction from the cognitive limitations and evaluative interests inherent in any particular subjective experience of the world. Rescher argues that: “Objectivity is ‘impersonal’ and ‘impartial’ not by way of dehumanization but by way of abstracting from those idiosyncracies that stand in the way of someone’s doing what ‘any of us’ — any rational, reasonable, sensible, normal person — would do in one’s place” (Rescher 1997, 7-8). He continues: “To strive for objectivity is to seek to put things in such a way that not just kindred spirits but virtually anyone can see the sense of it. It is not so much a matter of being impersonal as impartial — of putting aside one’s idiosyncratic predilections and parochial preferences in forming one’s beliefs, evaluations, and choices” (Rescher 1997, 8). In applying this conception to metaphysics, Rescher acknowledges what I would call the good psychologistic implications of ontological objectivity, stating that: “In the end, our commitment to ontological objectivity rests on the idealistic foundation of a mind-projected postulation [‘from the presently available data base’ of limited finitely experienced properties of physical objects ‘to a completed or ideally perfected data base’] whose validation is, in the end, functional and pragmatic, validated in the first instance not by evidence but by

256

DALE JACQUETTE

considerations of utility” (Rescher 1997, 115). According to Rescher’s definition of objectivity, the commitment of an objective metaphysics to a mind-independent reality is itself idealistic and mind-projected. Indeed, I feel compelled to ask rhetorically in support of Rescher’s conclusion, how could it be otherwise? 6. Good and Bad Psychologism in Epistemology

My final example concerns good psychologism in epistemology. For comparison, an illustration of bad psychologism in the theory of knowledge is harder to find. Knowledge is a specially qualified state of mind, often construed as some type of justified true belief, so that psychology may be inseparable from epistemology. I suggest, however, that Descartes’s epistemology in Meditations on First Philosophy is an instance of bad psychologism. By this I mean that, together with its rationalist proof for the existence of God as no deceiver, Descartes errs in interpreting clear and distinct belief states as constituting epistemically certain knowledge by virtue of their psychological character as revealed by introspection. The bad psychologism in Descartes’s epistemology is not necessarily the only or even the worst defect from which his philosophical system suffers, but one that is relevant to the present discussion by virtue of trying to elevate the psychological determination of psychological properties of psychological states into a criterion of truth and knowledge in what Descartes refers to as the accurate representation of external reality (Descartes 1984, 2; see Caton 1973). By contrast, good psychologism in epistemology can be found in W.V.O. Quine’s landmark essay, “Epistemology Naturalized”. Quine makes epistemology a branch of psychology in this famous passage: Philosophers have rightly despaired of translating everything into observational and logico-mathematical terms. They have despaired of this even when they have not recognized, as the reason for this irreducibility, that the statements largely do not have their private bundles of empirical consequences. And some philosophers have seen in this irreducibility the bankruptcy of epistemology...But I think that at this point it may be more useful to say rather that epistemology still goes on, though in a new setting and a clarified status. Epistemology, or something like it, simply falls into place as a chapter of psychology and hence of natural science. (Quine 1969, 82; see Kornblith 1985; Shimony and Nails 1987; Duran 1994)

In somewhat the same way that Boole proposes to gather the principles of logic from psychology, and then use logic to shed light on psychology, Quine similarly suggests that when epistemology is considered a branch of psychology among the natural sciences, it can also be used to understand the methodology of science, including psychology. Quine remarks:

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

257

The old epistemology aspired to contain, in a sense, natural science; it would construct it somehow from sense data. Epistemology in its new setting, conversely, is contained in natural science, as a chapter of psychology. But the old containment remains valid too, in its way. We are studying how the human subject of our study posits bodies and projects his physics from his data, and we appreciate that our position in the world is just like his. Our very epistemological enterprise, therefore, and the psychology wherein psychology is a component book — all this is our own construction or projection from stimulations like those we were meting out to our epistemological subject. There is thus reciprocal containment, though containment in different senses: epistemology in natural science and natural science in epistemology. (Quine 1969, 83)

I shall not enter into greater detail in defense of good psychologism in Quine’s epistemology naturalized. That it is psychologism in the sense I have defined is obvious enough. What I mean by describing Quine as recommending a good psychologism is that I believe he is right to interpret knowledge-seeking as a partly psychological activity undertaken by psychological subjects, that can be studied as a psychological phenomenon and improved in its methods, by abstracting and refining epistemic principles from the knowledge validating procedures in which we engage. The situation as I see it in Quine’s advocacy of epistemology naturalized is rather like Boole’s attitude toward the empirical investigation of the principles of deductive inference and probability judgments in formal symbolic logic. We begin, as we must, in this formula for good psychologism, with an intellectual pursuit which we can investigate as occurring in first-person subjective psychological experience and from the reports of subjective psychological experiences of other persons, in their efforts to master effective reasoning and perfect their understanding in knowledge. The principles or even the laws of logic and epistemology that are discovered in this way are then abstracted and refined from particular experience ideally in the way that Rescher describes in metaphysics. They can be interpreted as a projection in which we try to identify what is essential and look beyond whatever is peculiar or idiosyncratic to any individual subjective psychology, or even to whatever is specifically human, in order to clarify a priori concepts and develop more fully generalized philosophical methods that transcend the limits of any particular psychology or particular type of psychology. 7. Toward New Paradigms of Psychologism

To begin with and end with the limitations of any particular psychology or particular type of psychology in philosophy is bad psychologism, whether in logic, metaphysics, or epistemology, ethics or aesthetics. But to conclude that because some psychologism is or can be bad, that therefore all

258

DALE JACQUETTE

psychologism is bad, or that psychology has no place whatsoever in philosophy, is bad metaphilosophy. The word ‘psychologism’ has acquired such negative connotations that I doubt whether many of the philosophers I have mentioned would much appreciate being classified as advocating any form psychologism, including what I have called ‘good psychologism’. Yet that, by my definition, albeit in different ways, is precisely what Rescher, Quine, and others, are doing. There are unlimitedly many opportunities for elaborating good psychologistic philosophies. I have argued that it is a fundamental mistake to suppose that we can or should try to do entirely without psychological considerations in philosophy. An anti-psychologism that is not merely a criticism of bad psychologism, but that takes issue with psychology in any form in philosophy, as I have tried to show, is misdirected and selfdefeating. I have here barely taken the first steps in discriminating between good and bad psychologisms, let alone indicating how a good psychologism ought to be developed in logic, metaphysics, and epistemology. I will be satisfied if I am perceived as having at least said something convincing by way of distinguishing between psychologism and bad psychologism, indicating the possibility of a good psychologism with virtues and advantages that are worth considering in light of the most virulent generalized antipsychologisms that persist in philosophy. If my conclusions are correct, then no one who is conscientious about the proper conduct of philosophy can afford blithely to turn the page on psychologism, and assume, as many of its detractors have done, that psychologism is dead, buried, and never going to be resurrected. I have suggested, on the contrary, that psychologism needs to be revisited in more searching and rigorous ways.5 Department of Philosophy The Pennsylvania State University, USA

NOTES 1 2

See, inter alia, Philipse (1989), 58-74. Kant (1959), 4: “In this way there arises the idea of a twofold metaphysics — a metaphysics of nature and a metaphysics of morals. Physics, therefore, will have an empirical and also a rational part, and ethics likewise. In ethics, however, the empirical part may be called more specifically practical anthropology; the rational part, morals proper.” 3 I offer a more sustained attack on the philosophical underpinnings of formalism in philosophy of logic in Jacquette (1991); (1994), 76-89; 331-33. I criticize Frege’s efforts to eliminate reference to ‘associated idea’ (Bild, Vorstellung) from scientific semantics in Jacquette (1998), 63-79. 4 Hume (1978), 63-4: “’Twill probably be said, that my reasoning makes nothing to the matter in hand, and that I explain only the manner in which objects affect the senses, without endeavouring to account for their real nature and operations ...I answer this objection, by

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

259

pleading guilty, and by confessing that my intention never was to penetrate into the nature of bodies, or explain the secret causes of their operations. For besides that this belongs not to my present purpose, I am afraid, that such an enterprize is beyond the reach of human understanding, and that we can never pretend to know body otherwise than by those external properties, which discover themselves to the senses.” See Yolton (1984), Chapter VIII, ‘Hume on Single and Double Existence’, 147-64. Baier (1991), Chapter 5, ‘The Simple Supposition of Continued Existence’, 101-28. A controversy concerning the interpretation of Hume’s realism versus pyrrhonic skepticism about perception of the external world appears in the many references to some of the voluminous secondary literature on this topic in Wilson (1989), 49-73. Livingston (1991), 281-90. Wilson (1991), 291-96. An important earlier work in understanding this exchange is Price (1940). Also Popkin (1966), 53-98. 5 This essay was presented as an invited feature paper at a symposium on ‘Psychologism: The Current State of the Debate’ at the American Philosophical Association, Pacific Division, Albuquerque, NM, April 5-8, 2000. I am grateful to my co-panelists Michael Jubien and John H. Dreher, and to audience participants, especially the session organizer, Dallas Willard, for stimulating discussion.

REFERENCES

Baier, Annette. A Progress of Sentiments: Reflections on Hume’s Treatise. Cambridge: Harvard University Press, 1991. Beck, Lewis White. A Commentary on Kant’s Critique of Practical Reason. Chicago: University of Chicago Press, 1960. Berkeley, George. The Works of George Berkeley Bishop of Cloyne, edited by A.A. Luce and T.E. Jessop. London: Thomas Nelson and Sons Ltd., 1949. 9 volumes. Boole, George. An Investigation of the Laws of Thought on Which Are Founded the Mathematical Theories of Logic and Probabilities. New York: Dover Publications, Inc., 1959. Bowie, Andrew. Aesthetics and Subjectivity: From Kant to Nietzsche. Manchester: Manchester University Press, 1990. Carr, David. The Paradox of Subjectivity: The Self in the Transcendental Tradition. New York: Oxford University Press, 1999. Caton, Hiran. The Origin of Subjectivity: An Essay on Descartes. New Haven: Yale University Press, 1973. Cicero, Marcus Tullius. De Natura Deorum, translated by Horace C.P. McGregor, as Cicero, The Nature of the Gods. London: Penguin Books, 1972. Descartes, René. Meditations on First Philosophy in which are demonstrated the existence of God and the distinction between the human soul and the body, Meditation 1, in The Philosophical Writings of Descartes, translated by John Cottingham, Robert Stoothoff, and Dugald Murdoch. Cambridge: Cambridge University Press, 1984. 3 volumes. Doore, Gary. “Contradiction in the Will”, Kant-Studien, 76, 1985. Dorschel, Andreas. Die idealistische Kritik des Willens: Versuch über die

260

DALE JACQUETTE

Theorie der praktischen Subjektivität bei Kant und Hegel. Hamburg: Felix Meiner Verlag, 1992. Duran, Jane. Knowledge in Context: Naturalized Epistemology and Sociolinguistics. Lanham: Rowman & Littlefield, 1994. Hager, Achim. Subjektivität und Sein: das Hegelsche System als ein geschichtliches Stadium der Durchsicht auf Sein. Freiburg: Alber Verlag, 1974. Harrison, Jonathan. “Kant's Examples of the First Formulation of the Categorical Imperative”, The Philosophical Quarterly, 7, 1957. Hegel, G.W.F. Gesammelte Werke, herausgegeben im Auftrag der Deutschen Forschungsgemeinschaft. Hamburg: Felix Meiner Verlag, 1968. 21 volumes. Heil, John. “Does Cognitive Psychology Rest on a Mistake?”, Mind, 90, 1981. Hubbert, Joachim. Tranzendentale und empirische Subjektivität in der Erfahrung bei Kant, Cohen, Natorp und Cassirer. New York, Frankfurt am Main: Peter Lang, 1993. Hubbert, Joachim. Tranzendentale und empirische Subjektivität in der Erfahrung bei Kant, Cohen, Natorp und Cassirer. New York, Frankfurt am Main: Peter Lang, 1993. Hume, David. A Treatise of Human Nature [1739-40], edited by SelbyBigge, second edition revised by Nidditch. Oxford: The Clarendon Press, 1978. Jacquette, Dale. “The Myth of Pure Syntax”. In Topics in Philosophy and Artificial Intelligence, edited by Liliana Albertazzi and Roberto Poli, 114. Bolzano: Instituto Mitteleuropeo di Cultura, 1991. Jacquette, Dale. “Intentionality and the Myth of Pure Syntax”, Protosoziologie, 6, 1994. Jacquette, Dale. “Zeno of Citium on the Divinity of the Cosmos”, Studies in Religion / Sciences Religieuses, 24, 1995. Jacquette, Dale. “Intentionality on the Instalment Plan”, Philosophy, 73, 1998. Jacquette, Dale. On Boole. Belmont: Wadsworth Publishing Co., 2002. Jansohn, Heinz. Kants Lehre von der Subjektivität: eine systematische Analyse des Verhältnisses von tranzendentaler und empirischer Subjektivität in seiner theoretischen Philosophie, Mainzer philosophische Forschung, Band 2. Bonn: H. Bouvier, 1969. Jerius, Holger. Subjektive Allgemeinheit: Untersuchungen im an Kant. Freiburg, München: Alber Verlag, 1984. Kant, Immanuel. Foundations of the Metaphysics of Morals, translated with an introduction by Lewis White Beck. New York: Macmillan Publishing Company, 1959. Kant, Immanuel. Critique of Pure Reason. Translated by Norman Kemp Smith. New York: St. Martin’s Press, 1965.

PSYCHOLOGISM IN LOGIC, METAPHYSICS, EPISTEMOLOGY

261

Kant, Immanuel. Prolegomena to Any Future Metaphysics That Will be Able to Come Forward as Science. Translated by Paul Carus, revised by James W. Ellington. Indianapolis: Hackett Publishing Company, 1977. Kornblith, Hilary (editor). Naturalizing Epistemology. Cambridge: The MIT (Bradford Books) Press, 1985. Kusch, Martin. Psychologism: A Case Study in the Sociology of Philosophical Knowledge. London: Routledge, 1995. Livingston, Donald. “A Sellarsian Hume?”, Journal of the History of Philosophy, 29, 1991. Maslow, Alexander. A Study in Wittgenstein’s Tractatus. Berkeley and Los Angeles: University of California Press, 1961. McGinn, Colin. Wittgenstein on Meaning. Oxford: Basil Blackwell, 1984. Mohanty, J.N. Logic, Truth and the Modalities from a Phenomenological Perspective. Dordrecht, Boston, London: Kluwer Academic Publishers, 1999. Musgrave, Alan. “George Boole and Psychologism”, Scientia, 107, 1972. Navickas, Joseph L. Consciousness and Reality: Hegel’s Philosophy of Subjectivity. The Hague: Martinus Nijhoff, 1976. Notturno, Mark A. (editor). Perspectives on Psychologism. Leiden: E.J. Brill, 1989. Oliveira, Manfredo Araújo de. Subjektivität und Vermittlung: Studien zur Entwicklung der tranzendentalen Denkens bei Kant, E. Husserl u. H. Wagner. München: Fink, 1973. Paton, H.J. “The Aim and Structure of Kant’s Grundlegung”, The Philosophical Quarterly, 8, 1958. Philipse, Herman. “Psychologism and the Prescriptive Function of Logic”. In Notturno, editor, Perspectives on Psychologism. Popkin, Richard H. “David Hume: His Pyrrhonism and his Critique of Pyrrhonism”. In V.C. Chappel, editor, Hume. Garden City: Doubleday and Co., Anchor Books, 1966. Price, H.H. Hume’s Theory of the External World. Oxford: The Clarendon Press, 1940. Quine, W.V.O. “Epistemology Naturalized”. In Quine Ontological Relativity and Other Essays. New York and London: Columbia University Press, 1969. Rath, Mathias. Der Psychologismusstreit in der deutschen Philosophie. Freiburg: Alber Verlag, 1994. Rescher, Nicholas. Objectivity: The Obligations of Impersonal Reason. Notre Dame: Notre Dame University Press, 1997. Schofield, Malcolm. “The Syllogisms of Zeno of Citium”, Phronesis, 28, 1983. Shimony, Abner and Nails, Debra (editors). Naturalistic Epistemology: A Symposium of Two Decades. Dordrecht, Boston, Norwell: D. Reidel Publishing Company, Kluwer Academic Publishing, 1987.

262

DALE JACQUETTE

Spinoza, Benedictus (Baruch) de. Benedicti de Spinoza Opera quotquot reperta sunt, edited by J. Van Vloten and J.P.N. Land. The Hague: Martinus Nijhoff, 1914. 4 volumes. Swoyer, Chris. “Kantian Derivations”, Canadian Journal of Philosophy, 13, 1983. Wilson, Fred. “Hume’s Critical Realism: A Reply to Livingston”, Journal of the History of Philosophy, 29, 1991. Wilson, Fred. “Is Hume a Sceptic with Regard to the Senses?”, Journal of the History of Philosophy, 27, 1989. Wittgenstein, Ludwig. “Review of Peter Coffey, The Science of Logic”, Cambridge Review, 34, 1913. Wittgenstein, Ludwig. Tractatus Logico-Philosophicus. Edited by C.K. Ogden. London: Routledge & Kegan Paul, 1922. Yolton, John W. Perceptual Acquaintance from Descartes to Reid. Minneapolis: University of Minnesota Press, 1984. Zeno of Citium, The Fragments of Zeno and Cleanthes, A.C. Pearson, editor. New York: Arno Press, 1973.

PAUL A. ROTH

WHY THERE IS NOTHING RATHER THAN SOMETHING: QUINE ON BEHAVIORISM, MEANING, AND INDETERMINACY

I consider myself as behavioristic as anyone in his right mind could be. — W.V. Quine, Words and Objections

INTRODUCTION

Quine’s behaviorism belongs to his notion of evidence for scientific explanations, an account which places a premium on observability. He prizes observability because it features crucially in his project of “externalizing” empiricism. In contrast with classical empiricism, which takes what appears to appear to individuals, e.g., sense impressions, as explanatorily basic, empiricism externalized explains meaningful behavior by taking the publicly available as where explanation must begin.1 This externalization of basic evidence allows Quine to “naturalize” epistemology. For externalization positions epistemology so that it can utilize science both to explicate and to analyze the justificatory relations between evidence and beliefs, an exploration which, if successful, would yield as one result how we develop these very sciences.2 Empiricism itself is an intra-theoretic assumption endorsed by science as Quine understands it; the argument for empiricism resides in the explanatory successes of the sciences.3 Conversely, whatever plays no legitimate role as an explanatory posit has no claim to reality, to belong to the realm of facts.4 A naturalized epistemology — science self-applied — also examines and reveals where and how humans bridge the many gaps between input and output.5 Linguistic behavior provides the most striking and important case of such an “input-output” gap. Just here Quine emphasizes that language learning, to be explained at all, must be explained behaviorally. “I hold further that the behaviorist approach is mandatory. In psychology one may or may not be a behaviorist, but in linguistics one has no choice.”6 But just 263 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 263-287. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

264

PAUL A. ROTH

how Quine’s behaviorism fits with his account of scientific explanation prompts perplexities. Human linguistic abilities represent a crucial unsolved puzzle, a standing challenge to the explanatory aspirations of current science. On the one hand, Noam Chomsky famously identifies the failure of Skinnerian behaviorism as an explanation of linguistic ability as its Achilles‘ heel.7 But, on the other hand, when he similarly indicts Quinean behaviorism, Quine remains unpersuaded.8 Yet how could Quine fail to find Chomsky’s criticisms relevant? Why presume that an externalized empiricism drives a principled wedge between empirical science (a notion Quine construes quite broadly)9 and the explanatory utility or scientific respectability of meanings in his pejorative sense? Chomsky’s views provide a natural foil for those of Quine’s. For they share a common view of what needs to be explained. Both Quine and Chomsky emphasize the interesting, important, and deeply puzzling explanatory challenge that language use presents. Yet Chomsky insists upon the scientific respectability of just the sort of theoretical posits — the reality and so causal efficacy of mental states, the necessity of conceptual structures — that Quine finds entirely bereft of explanatory merit. Moreover, while Chomsky insists that Quine’s behaviorism signals a vast disagreement between their positions, Quine finds nothing important that divides them with regard to the characterizations of the problem and the empirical bases for answers. This essay explicates Quine’s behaviorism and its place in his philosophy. Attention to Quine’s behaviorism serves, somewhat surprisingly, to make perspicuous the most contested thesis of Quine’s philosophy, the indeterminacy of translation of theoretical sentences. Quine’s thesis of the indeterminacy of translation — that there is “no fact of the matter” to disputes about meaning — develops from arguments that conceptual structures and mental states cannot contribute to explanation. Section I clarifies in what Quine’s behaviorism consists and how this constrains his account of what legitimates posits for purposes of explanation. Section II further develops Quine’s notion of what constrains explanation by connecting the constraints sketched in the previous section with Quine’s critique of the notion of linguistic convention. Section III shows how the points regarding posits, conventions, and explanation developed in the first two sections combine to provide a compelling argument for the indeterminacy of translation. This argument shows why indeterminacy does not hinge on what future science may hope to discover. Finally, this examination of Quine’s behaviorism, its place in his general epistemology, and how these intersect with Quine’s views on mind and meaning point to what constitutes a post-Quinean philosophical problematic.

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

265

I People speak. On this everyone agrees. But what must be the case for a language to be the case? Here disputes arise. In particular, must any explanation of a capacity for language involve conceptual or intentional elements, i.e., attribute necessity to some meanings and causal efficacy (reality) to some mental states? Quine thinks not. Such posits have no place, he insists, in any scientifically respectable strategy of explanation. What particularly needs explaining is how humans manage the transition from, in Quine’s words, “meager input to torrential output,” a study of how collectively we move from “stimulus to science.” The question of how people acquire language and knowledge holds intellectual center stage for both Chomsky and Quine. As Alexander George rightly emphasizes, “Chomsky’s explananda do not differ from Quine’s...Like Quine, Chomsky considers his work a response to ‘the problem of “causation of behavior”.’”10 The striking and obvious fact for both concerns how our accepted pronouncements far outrun the perceptually available evidence. This fact, in turn, sets the explanatory agenda for their competing accounts. Yet their very agreements deepen the mystery of how exactly to account for what divides them. As George observes, “If the disagreement is not over the sort of data to be explained, the need for innate mechanism, or the commitments of the far-future fundamental theory, then whence does it stem?”11 Quine and Chomsky diverge as well on a basic consequence of accepting the inadequacy of behaviorist/empiricist explanations in linguistics. Quine declares, “If Chomsky’s antiempiricism or antibehaviorism says merely that conditioning is insufficient to explain language-learning, then the doctrine is of a piece with my doctrine of the indeterminacy of translation.”12 Chomsky disagrees. Regarding indeterminacy of translation in particular, Chomsky maintains, “Quine’s thesis of indeterminacy of radical translation amounts to an implausible and quite unsubstantiated empirical claim about what the mind brings to the problem of acquisition of language...as an innate property.”13 Construed epistemologically, Chomsky finds no distinction between the alleged indeterminacy of meaning and garden variety forms of underdetermination. “Quine’s thesis is simply a version of familiar skeptical arguments which can be applied as well to physics, to the problem of veridical perception or, for that matter, to his ‘genuine hypotheses’.”14 Chomsky finds the underdetermination of linguistic theory to be true but trivial, but the alleged indeterminacy of meaning unsupported by argument.15 Goaded perhaps by Quine’s self-characterized behaviorism, Chomsky further complains that it is “not at all obvious that the potential concepts of ordinary language are characterizable in terms of simple physical dimensions of the kind Quine appears to presuppose or, conversely, that concepts characterizable in terms of such properties are potential concepts

266

PAUL A. ROTH

of ordinary language.”16 As a psychological theory, Chomsky insists, Quine’s theory simply refuses to recognize the mind’s role in structuring the deliverances of experience. Understood epistemologically, Chomsky puzzles how linguistics could be worse off than physics, for each explains the observed by the unobserved. In the end, Chomsky insists, Quine’s account “receives no support from what is known about language learning, or from human or comparative psychology.”17 Quine’s response here signals the depth of the divide between his views and Chomsky’s. For, on the one hand, Quine remarks that “Chomsky’s remarks leave me with feelings at once of reassurance and frustration. What I find reassuring is that he nowhere clearly disagrees with my position. What I find frustrating is that he expresses much disagreement with what he thinks to be my position.”18 Yet, on the other hand, regarding Chomsky’s dismissal of indeterminacy as a special problem, Quine responds, “Chomsky did not dismiss my point. He missed it.”19 What Chomsky missed, I suggest, are the parameters of a Quine’s account of explanation, and how his behaviorism fits into that account. Behaviorism constitutes Quine’s criterion for ascertaining what there is by way of evidence for shared meaning. Behaviorism naturalizes, in Quine’s sense of the term, the study of meaning. With Dewey I hold that knowledge, mind, and meaning are part of the same world that they have to do with, and that they are to be studied in the same empirical spirit that animates natural science... When a naturalistic philosopher addresses himself to the philosophy of mind, he is apt to talk of language. Meanings are, first and foremost, meanings of language. Language is a social art which we all acquire on the evidence solely of other people’s overt behavior under publicly recognizable circumstances. Meanings, therefore, those very models of mental entities, end 20 up as grist for the behaviorist’s mill.

Taken in its (for Quine) pejorative non-behavioral/mentalist sense, meaning is indeterminate. There is “no fact of the matter,” no truth-maker, which could possibly decide the issue between incompatible attributions of meanings — different translations.21 Quine’s own characterizations of his behaviorism tend to be terse, concerned more to distance himself from standard characterizations of behaviorism than positive expositions of how he intends the notion.22 Positive characterization proves somewhat more difficult to come by. Looking to work by Wilfrid Sellars helps here, for Quine strongly endorses and embraces an exposition of behaviorism which Sellars provides.23 Sellarsian behaviorism gets its philosophical grip precisely when one focuses on cases of initial language acquisition, i.e., infant languagelearning. Both Sellars and Quine consider this to be the philosophically significant case. For an infant just is someone who must break into discourse

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

267

with nothing by way of prior training or background to help. Accounting for an infant’s success in acquiring language represents a philosophical natural experiment, a genuine case of “radical translation.” In asking how an infant qua radical translator might manage, Sellars parallels the acquisition of concepts used to characterize mental states with those used to speak of observable physical objects. Just as objectification involves mastering terms others use to characterize what goes on around us, “subjectification” similarly involves coming to apply terms acquired in the public domain to what goes on within. That is, Sellarsian behaviorism takes introspection as “a conceptual response to psychological states and the concepts included in this response are common sense psychological concepts and, as such, no more adequate to an understanding of what is really going on than are common sense concepts pertain to the middle sized physical objects of everyday experience.”24 Sellars emphasizes that notions so acquired have no special claim to scientific standing. Sellars observes that some sort of reasoning must be involved in an infant’s learning to speak. But how in turn to account for this ability? Is reasoning learned or part of one’s natural endowments? He astutely observes that these alternatives are neither as simple nor as clear as might first appear. The former alternative seemingly endows infants with too few resources, the latter with too many. For the latter case particularly imagines infants qua radical translators as, ab initio, practicing scientists, relentlessly applying the method of hypothesis and test to connect newly heard words to their newly found world. Infants so conceived are not merely potentially rational, but actually so, and to a high degree. §14. According to this model the infant — confronted by salient linguistic configurations of sounds, formulates increasingly subtle hypotheses about them and the contexts in which they occur; and accepts, rejects or modifies these hypotheses. Each new occasion provides more grist for the child’s inductive mill. §15. Now it leaps to the eye that this kind of explanation assumes that the child is no merely potential rational animal. Its rationality is full fledged. It operates with concepts and logical forms which have a high degree of sophistication... §16. If the question is now raised — How does the infant acquire the conceptual abilities mobilized by his inductive reasonings? One answer is clearly precluded: “in the course of learning a language.” Nor could these abilities be acquired by a process of inquiry or reasoning. The pressure towards an innatist account of the child’s logical powers would be almost 25 irresistible.

The latter alternative, in short, creates an explanatory need for a very rich conception of a “language of thought.” But Sellars enters a critical cautionary note on just this point. The caution is this. The primary argument for innateness turns on claims that key elements of reasoning and conceptualization cannot be acquired

268

PAUL A. ROTH

through experience, i.e., be learned. Explaining language learning, on this view, must begin by identifying what cannot be learned, and so what is necessarily innate. These capacities must always already be there. But what, in general, distinguishes what is learnable from what is not? Sellarsian behaviorism expresses just this caveat: beware positing preexisting mental capacities, for this may create more mysteries than insights. Needed is an account of just how the unlearned makes learning possible; invoking mental states threatens instead only obscurum per obscurius. §25 ...But from the standpoint of methodology the binding principle was to be: Don’t simply borrow concepts and principles from the framework of introspective knowledge. Use all the analogical and suggestive power of Mentalistic concepts and principles, but be sure that the concepts and principles you introduce have no more Mentalistic structure than can be justified in terms of their ability to explain observable behavior phenomena. §26. As I see, this was — and remains — the methodological stance of a sophisticated behaviorism26.

He advocates instead a behaviorism that, while recognizing no a priori restriction on the character of “intervening variables,” also represents a principled refusal to countenance any more mental structure than required to turn the learning trick. How much to attribute to innate capacities awaits, then, a more general theory of learning. Precisely the wrong, indeed unjustifiable, methodological move here is to posit innate structures in the absence of some prior general account of what can or cannot be learned. “Invoking ‘Innateness’ only postpones the problem of learning; it does not solve it. Until we understand the strategies which make general learning possible — and vague talk of ‘classes of hypotheses’ — ...no discussion of the limits of learning can even begin.”27 In light of how little of behavior can be made sense of by appeal to rules, time has only underscored the wisdom of this behaviorist caution.28 In this regard, Quine’s identification of the Sellarsian account as his own puts his demands on justifying posits in proper perspective. What work must posits do to earn their theoretical keep?29 Quine’s behaviorism recognizes two strategies for justifying theoretical posits. First, a posit could be held to be constitutive. Meeting this requirement involves proving, for example, a need to assume some such ability as a condition of utilizing behavioral or other empirical evidence in the first place. Quine’s argument for quality spaces provides a model here. “A standard of similarity is in some sense innate. This point is not against empiricism; it is a commonplace of behavioral psychology...Without some such prior spacing of qualities, we could never acquire a habit; all stimuli would be equally alike and equally different...Needed as they are for all learning, these distinctive spacing cannot themselves all be learned; some must be

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

269

innate.”30 A theoretical posit — e.g., a genetic/innate disposition to notice color differences — explains something — a set of stable responses — that cannot be explained except by a prior ability to make certain discriminations in perception. The stability of responses across individuals and across times marks the same as the same. A biologically rooted and evolutionarily acquired ability to make certain perceptual discriminations provides a mechanism. Second, in the context of a naturalized epistemology, proposed explanatory posits may “pay their way” by providing a means to independently mark out what they purport to characterize. This demand sets a special hurdle, or so I argue, for posits claimed explanatory of “meaningful” behavior. For much behavior taken to be meaningful is so only on the prior assumption that it is intentional or expressive of some conceptual ability. But precisely this assumption puts the theoretical cart before the horse. Behavior may be seen as flirtatious, insincere, etc., but what makes it such depends on nothing that objectively identifies the behavior as being as the type it is. Unlike the stabilities of behavior which underwrite the imputations of quality spaces, there exists no consensus across individuals and times regarding how to interpret specific behaviors. What explanatory advantage accrues to science by positing a kind and then being able only to identify willy-nilly behaviors that supposedly fit it? How otherwise to distinguish between kinds actually needed for explanation and those that merely free-ride and do no explanatory work? Quine’s commitment to behaviorism asks only for evidence that the categories invoked are about something rather than nothing.31 Regarding the interpretation of behavior, there are no facts objectively determining if things are what putative explanations say they are. But in such cases, what then gets explained? Wheels turn, but nothing moves. This breaks the parallel between conflicts of interpretation and garden-variety underdetermination. For the issue does not concern conflicting but empirically equivalent theories, but how initially to ascertain what the explanations explain.32

II What would the discovery of meaning be the discovery of? How would one “know the meaning” if, so to speak, one stared it in the face? The previous section brings this question to the fore. The friends of meaning owe a story of how to mark by any jointly observable, stable, and so objectively determinable feature that which the theorized entity picks out and supposedly explains. Quine only asks how to identify what the evidence allegedly evinces. He makes no prior judgment regarding what may serve to verify.

270

PAUL A. ROTH

If, for example, a speaker avers that ‘Gavagai’ pick out a whole rabbit and not an undetached rabbit part, what makes the utterance evidence for something over and above a disposition to affirm some sentences and deny others? Nothing in the world remains unexplained by the dispositional account, an account itself explicable in terms of learning.33 Thus what may first appear to be a base physicalist prejudice is not. The foregoing suggest why Quine believes that the “problem of evidence for a linguistic universal is insufficiently appreciated.”34 What, he puzzles, does a claim to have discovered such a universal “amount to, pending some standard of faithfulness and objectivity of translation?”35 The very posing of evidence as evidence for universals “presupposes some prior behavioral standard of what, in general, to aspire to include.”36 At issue here is imputations of sameness of meaning. Quine only asks for an empirical mark by which the “same” can be, in point of fact, shown to be such. But what would the discovery of “same meaning ” be the discovery of?37 Quine distinguishes between what is evidentially fundamental, naturally fundamental, and conceptually fundamental.38 What is conceptually fundamental concerns terms and descriptions whose use we learn first. Fluency here qualifies us for membership in our linguistic community. By contrast, what is either evidentially or naturally fundamental concerns evolving scientific explanations of what there is. Quine concedes that the intentional idiom might just be conceptually fundamental; what he denies is that this idiom finds a place in our best explanatory scheme of things.39 Issues surrounding when posits explain and when they do not can be clarified by paralleling Quine’s worries about the mental with claims that appeal to conventions sheds light on the nature of mathematical truths.40 As in other cases, discussion of the mathematical or logical contexts proves instructive for more ambiguous contexts, e.g., ordinary language. For standards of clarity or explicitness lacking in the former case invariably obtain and to a greater degree in the latter. Quine suggests two criteria which any satisfactory definition or a term or sign must satisfy. On the one hand, there is a formal requirement of “unambiguous eliminability.” Definitions must show how the definiendum can be replaced by other, presumably more primitive and well-defined terms, e.g., truth tables definitions of logical connectives. On the other hand, a proposed definition “must also conform to the traditional usage in question. For such conformity it is necessary and sufficient that every context of the sign which was true and every context which was false under traditional usage be construed by the definition as an abbreviation of some other statement which is correspondingly true or false under the established meaning of its signs.”41 Without this substitutivity, I take it, the proposed definition simply becomes stipulative, just as the truth table definition of ‘and’ respects some ordinary language contexts but not others.

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

271

Moreover, definitions should discriminate between statements true by accident of circumstance from those which are logically true. “A definition which fails in this latter respect is no less Pickwickian than one which fails in the former; in either case nothing is achieved beyond the transient pleasure of a verbal recreation.”42 Nothing inherent to the process of assigning conventions itself restricts us, including consistency. Conformity to antecedent use alone guides us in assessing whether the translation comes out “right.”43 Yet, conventions per se fail to explain the esteem given some truths over others because the putative strategy of explanation applies equally as well to subject matters not so esteemed. Conventions provide a method one can use to circumscribe certain types of truths — ones of mathematics or logic, say — but nothing about this method distinguishes these truths from, e.g., those of empirical science.44 If in describing logic and mathematics as true by convention which is meant is that the primitives can be conventionally circumscribed in such fashion as to generate all and only the accepted truths of logic and mathematics, the characterization is empty; our last considerations show that the same might be said of any other body of doctrine as well. If on the other hand it is meant merely that the speaker adopts such conventions for those fields but not for others, the characterization is uninteresting; while if it is meant that it is a general practice to adopt such conventions explicitly for those fields but not for others, the first part of the characterization is false.

Rather, we tailor conventions to fit established usage. But the use of antecedent behavior as a guide precludes taking the resulting conventions as explanatory of behavior. “Still, there is the apparent contrast between logico-mathematical truths and others that the former are a priori, the latter a posteriori...View behavioristically and without reference to a metaphysical system, this contrast retains reality as a contrast between more and less firmly accepted statements; and it obtains antecedently to any post facto fashioning of conventions.”46 That is, the assignment of conventions merely formalizes what prior behavior already reveals. [I]t is not clear wherein an adoption of the conventions, antecedently to their formulation, consists; such behavior is difficult to distinguish from that in which conventions are disregarded. When we first agree to understand ‘Cambridge’ as referring to Cambridge in England, failing a suffix to the contrary, and then discourse accordingly, the role of linguistic convention is intelligible; but when a convention is incapable of being communicated until after its adoption, its role is not so clear. In dropping the attributes of deliberateness and explicitness from the notion of linguistic convention we risk depriving the latter of any explanatory force and reducing it to an idle label. We may wonder what one adds to the bare statement that the truths of logic and mathematics are a priori, or the still barer behavioristic statement that they are firmly accepted, when he characterizes them as true by convention in such a sense.47

272

PAUL A. ROTH

Quine’s well-known conclusion to the argument here is that if the goal is to justify intuitions that truths of logic or mathematics differ in kind from those of the natural sciences, appeal to conventions supply no such explanation or justification. Appeal to conventions represents no advance (for purposes of explanation) over the claim that some truths are behaviorally well-confirmed. Quine’s denial that meanings have any explanatory role to play parallel his arguments that “truth by convention” fails to explains how mathematical truth differ in kind from those of empirical science. Work by Thomas Ricketts helps makes this point and indicates how Quine’s concerns about explanation connect to indeterminacy.48 The starting point here, as with so much else with Quine, involves Quine’s disputes with Carnap. Carnap believed that the notion of shared linguistic frame-works had explanatory utility, specifically in regard to accounting for the character of scientific rationality. The attribution of a “shared framework,” Carnap claims, explains in what the rationality of scientific investigators consists. Constitutive of scientific rationality are those beliefs/statements not open to revision by experience. For Carnap, the notion of statements that are rationally unrevisable — analytic statements — provides a supposed key to explicating scientific rationality extrascientifically. Quine, as argued above, finds nothing that distinguishes socalled analytic sentences from the rest — “It is sometimes cold on Thursdays.” Quine’s challenge to Carnap asks for a justification of the explanatory “value added” by positing a distinction in kind between truths. The question now becomes how to identify those rules constitutive of scientific rationality — rules it would be irrational to give up. Quine presses Carnap to show in what non-arbitrary feature the alleged difference between the two sorts of statements consists. Carnap attempts to meet the challenge by appeal to behavioral criteria. In the end, Carnap does recognize Quine to be asking after the ground for attributing linguistic frameworks. Carnap attempts to meet this challenge by presenting a “behavioristic, operational procedure” for identifying the analytic sentences of a person’s language by reference to the person’s speech dispositions...[T]hese dispositions are those which mark a sentence as rationally unrevisable. Description of these dispositions is, however, couched in concrete, more or less behavioral terms. Thus it, unlike the previously described criterion, avoids illicitly presupposing the availability of a criterion of analyticity. Carnap hypothesizes that the procedure for attributing linguistic frameworks could be cast into the form of a handbook, a manual.49

Carnap’s behavioral criterion promises an empirical basis for distinguishing the “merely” empirical from the rationally unrevisable. Such a behavioral standard would vindicate his distinction between truths constitutive of rationality and other accepted beliefs.50

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

273

But how can behavioral evidence do that? For nothing in behavior marks the behaviors as cases of following one rule rather than another. The problem is not physicalism. The issue, rather, concerns what marks behavior as evidence for one kind of rule and not another? As Ricketts observes, “the thesis of indeterminacy of translation is Quine’s response to Carnap. Quine‘s examination of the role independently describable speech dispositions play in guiding and constraining translation show Carnap’s criterion to be a counterfeit.”51 It is not (just) that behaviors might be compatible with different rules. Lacking is an argument that the behaviors can only result as a consequence of some specific rule. Fitting behaviors to rules does not show that the behavior results from a rule. As in the case of the assignment of conventions, rules just recapitulate rather than explain what behavior first reveals. Ricketts analysis helpfully identifies the key role played in the debate by Quine’s behaviorism. On his account, what Quine develops is a behavioristic analog for terms of positivist epistemology that cannot be justified as the positivists sought — so Quine‘s behavioral definition of an observation sentence — and dismisses those for which he can find no suitable behavioral analog, e.g., notions of analyticity and synonymy.52 Distinctions between kinds of truths fails Quine‘s behavioral tests. Experience reveals some truths to be more stubbornly held than others, but what goes for “7+5=12” goes for “It is sometimes cold on Thursdays.” Ricketts remarks also prove enlightening regarding the core differences between Quine and Chomsky. On this reading, Chomsky is just one more case of explanation by attribution of shared frameworks. Quine could take Chomsky to be arguing with him, one scientist to another. In particular, Quine could construe Chomsky as offering a rival approach to Quine‘s behaviorism...To dismiss Chomsky’s challenge here, Quine need only, from the vantage point of the physical theory of which psychology is a component part, express his justified confidence that every instance of speech behavior admits of a physical explanation. The failure of behaviorism shows only that we will not have very much to say by way of systematic explanation of verbal behavior until neurophysiology is far more advanced than it is today.53

What Quine questions is not whether Chomsky has the framework right, but what appeals to a “shared something” actually explain. This brings together Quine’s doubts about the illumination provided by explanation via conventions and his worries about the unlearned as explanatory of the learned. One the last move remains for friends of the meaning. Advocates could insist that there exists a class of explananda left unexplained by Quine’s naturalized epistemology. This is the class characterized as meaningful speech, and not “mere” behavior.

274

PAUL A. ROTH To explain an emission of noise is not, however, to explain an utterance of an English sentence, let alone an assertion that it is raining outside...On the basis of this explanatory deficiency, Chomsky challenges the explanatory omnicompetence of Quine’s science.54

At this point, Ricketts worries, the debate reaches a logical impasse. But now, from Quine’s perspective, we see Chomsky as begging the question, or rather as failing to raise any question at all...The disagreement between Quine and Chomsky is not a clash of rival theories addressed to the same data but a difference over what counts as data, over the terms we take for granted to represent data.55

Since their rival theories cannot agree on what needs explaining, a choice between them “will then opt for Quine’s account of language or Chomsky’s depending on whose convictions about what there is to be explained persuade us.”56 But this conclusion is premature. Claims that some set of innate rules explains verbal behavior would need to first establish that certain rules guide, and not merely fit, behavior.57 “Guiding” is a causal notion. Syntactical models at best fit the data. No account presently exists explaining how individuals manage to be guided by syntactic structures in the relevant causal sense. Moreover, it is difficult to imagine just how such an account might help, for adhering to syntactic forms is neither necessary nor sufficient for meaningful speech. But the steeper obstacle concerns how to provide an objective determinant of the supposed products of mental causes. For behavioral criteria, as already noted, cannot suffice for this purpose. Nothing in the world marks, e.g., states, meanings, or concepts as “the same.” We do not know what the discovery of “same meaning” would be the discovery of. What marks out mental kinds as members of those kinds cannot be given a defining observable correlate. Quine endorses Davidson’s anomalous monism for this very reason.58 Why assume with regard to meaning or the mental that something exists which needs explaining? No shared facts distinguish behaviors said to have an intentional property from those that lack it. Assigning mental states, on this view, is like formulating conventions. We tailor assignments to fit behaviors, hence the assignments cannot explain the behaviors. The distinction between the intentional and the non-intentional goes the way of the distinction between the analytic and synthetic. Nothing objective allows us to determine what such distinctions supposedly distinguish.

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

275

III The argument for indeterminacy, as I reconstruct it, does not depend upon any a priori philosophical assumption regarding what must be excluded from scientific explanation. Indeterminacy obtains because meaning in the pejorative sense has no explanatory use. It has none because it is not required to explain how evidence is possible (and so play a constitutive role) or to designate properties which objectively define a thing as being of a certain kind, and so a scientific or explanatory role. The overall structure of the argument may be represented as follows: Pr. 1: For posits to have claim to reality, they must be justified as part of an explanatory theory within a naturalized epistemology;59 Pr: 2: In order to be a justified part of an explanatory theory, the posits must either be necessary for (constitutive of) stimulations being evidence (as in quality spaces), or must provide an observable mark marking them as the things they are, i.e., there must be an objective basis for ascertaining that some behavior has been rightly or wrongly categorized a behavior of a certain type60; Pr. 3: But mental states and concepts have yet to be shown to be necessary for or even robustly explanatory of observed stabilities in behaviors; in addition, nothing marks a behavior as objectively categorized correctly or incorrectly61; C: Meanings and mental states have no claim to reality (because no claim to any genuine explanatory function). Whatever has no place in our best scientific scheme of explanations is not real. Whatever is not real does not belong to the realm of facts, i.e., constitute a fact of the matter. The first premise strikes me as self-evident. Why call something real if it has no role in explanation? If causally idle or redundant with regard to impacting how everything else in the world goes on, in what would its “reality” consist? The second premise states what legitimates a posit. Posits can be constitutive of the possibility of stimulations being evidence, or they can be kinds made objective by accepted theoretical explanations. Substantiating the third premise has been the burden of each of the first two sections. Together, they establish that meaning in the pejorative sense fails to be legitimated by either standard — it has no constitutive role and it lacks objective markers. The failure of both constitutive and objective/ verifcationist rationales for meaning is why there is nothing rather than something.

276

PAUL A. ROTH

As Quine comes to concede, the notion of “sameness” needed to account for agreement by speakers cannot be cashed out in terms of some comparison of nerve endings of those involved. No doubt information comes to us perceptually; but equally not in doubt is that, at this perceptual level, individuals exhibit more differences than not.62 Quine comes to speak, late in his writings, of a “preestablished harmony”, an evolutionarily developed disposition for people to respond alike under certain conditions. But, in this case, behavior provides only a symptom of a physiological condition. Chomsky rightly notes my penchant for innate ideas. Rightly, anyway, if we construe ‘innate ideas’ in terms of innate dispositions to overt behavior ...[T]his penchant is one I share with behaviorists generally...Language aptitude is innate; language learning, on the other hand, in which that aptitude is put to work, turn on intersubjectively observable features of human behavior and its environing circumstances, there being o innate language and to telepathy. The linguist has little choice but to be a behaviorist at least qua linguist; and, like any behaviorist, he is but to lay great weight upon innate endowments.63

What explains, on Quine’s view from early to late, are accounts cashed out in terms of how bodies respond and work.64 The presence of stable behavioral criteria provides a unified explananda. “Another distinctive point about the indeterminacy of translation is that it clearly has nothing to do with inaccessible facts and human limitations. Dispositions to observation behavior are all there is for semantics to be right or wrong about.”65 As Quine emphasizes, taking mental states as causes simply introduces causal Doppelgänger into a theory. If evolutionary explanations can account for the apparent “preestablished harmony,” then there is no need to explain things twice, once by a biological mechanism, a second by a conceptual one. Føllesdal nicely characterizes the relation of what he terms Quine’s “evidential behaviorism” to Quine’s general argument for indeterminacy.66 The problem, as Føllesdal develops it, is that in attempting to infer from purported evidence for mental states back to those states the evidence allegedly evinces, there are only two options. The states can be identified directly, by comparison of one to the other, or indirectly, via corresponding physical states.67 This harkens back to the Sellarsian point so strongly endorsed by Quine that in appropriating the vocabulary of a folk psychology, we literally do not know what it is we appropriate. In claiming that the verbal behavior is evidence for anything at all, the friend of mentalism needs a way of objectively discriminating the objects the evidence picks out from others — a fact of the matter to the claim that something is of one kind rather than another. But physical states, including behavior, provide no such criterion. Different anatomies do not match up. Moreover, to identify any two bodily states as states of the “same meaning,’ we would already have to know what to count as such.68 But, of course, what counts here is precisely what is unknown.

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

277

In “Two Dogmas of Empiricism,” when discussing interchangeability salva veritate as a criterion of analyticity, Quine notes that one could discern which contexts were analytic and which not provided one understood a term such as ‘necessarily.’ But understanding this term, he goes on to complain, presupposes rather than explicates the meaning of ‘analytic.’ Such an account, he observes, “is not flatly circular, but something like it. It has the form, figuratively speaking, of a closed curve in space.”69 A related metaphor figures as well in his complaints regarding taking evidence to be evidence for one linguistic universal rather than another. “We are looking for a criterion of what to count as the real or proper grammar, as over against an extensionally equivalent counterfeit...And now the test suggested is that we ask the native the very question which we do not understand ourselves: the very question for which we ourselves are seeking a test. We are moving in an oddly warped circle.”70 The suggested criteria — a prior understanding of ‘necessarily’, a prior mode of discriminating behaviorally — prove unsatisfactory because they turn out to presuppose exactly what we seek to know. Just what should be looked for? What would the discovery of meaning be the discovery of? The metaphysics of meaning collapses under the weight of these questions. In this respect, Quine anticipates and responds to what Warren Goldfarb terms the “scientific objection” to critiques of the possible role of mental states and meanings in the scheme of explanation.71 The objection goes as follows: a philosopher who denies explanatory efficacy of mental states (Wittgenstein, Quine) “is simply making a bet on the future course of science or else he is engaged in a priori anti-science, denying a priori that certain projects could bring results, and hence they ought not even be investigated empirically.”72 But if the former, the Quinean claim loses its principled cast; if the latter, it represents just the sort of philosophic prescribing to science Quine otherwise abjures. Goldfarb examines this objection in relation to Wittgenstein’s procedures, but suggests (quite correctly on the account developed here) that the considerations apply as well to Quine.73 On Goldfarb’s reading, Wittgenstein shares Quine’s skepticism about the explanatory efficacy of the mental, and for similar reasons: such terms cannot be shown to mark out kinds needed for or needing explanation. “Wittgenstein’s treatment...of the cognitive or intentional mental notions are evidently meant to persuade us that, in some sense,...[they] are not particular or definite states or processes”.74 But this appears to provide an opening for the “scientific objection.” Why, such an objector may ask, “isn’t Wittgenstein here usurping the place of empirical inquiry? Is it not possible that empirical science — neurophysiology, in particular — will find specific states and processes that will fill the bill, as far as understanding, believing, remembering, etc. are concerned?”75 Clearly, a Quinean or Wittgenstein owes an answer to this objection.

278

PAUL A. ROTH

Goldfarb’s analysis here brings together the Sellarsian and Quinean points about learning and behaviorism in a perspicuous fashion. For the core of a Wittgensteinian response to the scientific objection resides in consideration of examples. The examples work to show that characterizations of “same behavior” exemplify no defining attributes. Wittgenstein asks us to look in detail at the range of our practices relevant to an ascription of understanding. We find an enormous variety of considerations that can enter, a dependence on context that is impossible to describe accurately by any general rules, a lack of uniformity in mental accompaniments. In individual cases [of ascription of mental states such as understanding] we have stories to tell to justify our ascriptions, but there is no uniform feature that pulls the various cases together...If it is at such a juncture that the scientific objection is voiced — if the objector is saying, even though our ascriptive practices now show no uniform features, ‘that’s only because we have too little acquaintance with what goes on in the brain and the nervous system’ ([PI] § 158) — then the roots of the objection in a priori demands that something unitary and definite must ground the ascription are evident.76

Considerations such as these coincide precisely with Quine’s complaints. Nothing in the world marks the ascription of mental kinds as being the kinds they supposedly are.77 In this regard, the scientific objection gets things backwards.78 For it is the friend of a science of the mental who demands science respect certain categories even while lacking any independent evidence of what might possibly make things members of this category. “[T]he scientific objection is undercut for its attractive empirical stance — its appearance of making only the modest claim that things might turn out either way, we might or we might not discover an appropriate neural state — is seen to be a pretence. The objection loses its scientific cachet...It becomes just a restatement of a picture of what has to be there.”79 Absent an account of what behaviors objectively exemplify when exemplifying one mental state rather than another, the scientific objection merely manifests a philosophical disease, not its cure.80 Quine charts what falls by the way in an advancing scientific picture of what there is. His view of epistemology as science self-applied, and his corresponding conclusion about the explanatory utility of meanings, has interesting parallels here to Kantian themes. For at least one important link that connects a tradition that runs from Kant to Quine studies how a human minds come to constitute a shared and mutually intelligible world. Kant, in this regard, assigns the inquiring mind a strongly constitutive role. But how concepts and percepts come together to form understanding remains an unsolved problem, a problem which finds its modern incarnation in questions of finding rules that guide behavior. Quine despairs of the task of using philosophical analysis to find such rules. He proposes, instead, to turn that job over to empirical psychology. How an account of the place of mind

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

279

in the world plays out in the context of Quinean naturalism, however, is a story for another day.81 Department of Philosophy University of Missouri, Saint Louis, USA

NOTES Quine (1980b), 259. Quine does not limit epistemology, in this regard, to the so-called “doxastic assumption,” i.e., the view that only beliefs can be used to justify other beliefs. Causation and implication are, for him, just two interrelated strategies for the justification of beliefs. “The business of naturalized epistemology, for me, is an improved understanding of the chains of causation and implication that connect the bombardment of our surfaces, at the one extreme, with our scientific output at the other. The first link is causal: The bombardment of the exteroceptors causes a neural intake. The next link connects the neural intake with language: observation sentences become associated with ranges of perceptually similar neural intakes, at first by conditioning. The subject learns to assent to the observation sentence on the occasion of any global neural intake in the associated range.” Quine (1995c), 349. See also his discussion in (1990), 41-4. I explicate and defend Quine’s account in Roth (1999), 87-109. A sympathetic treatment may also be found in Koppelberg (2000), 101-19. 3 Quine (1990), 20-1: “Even telepathy and clairvoyance are scientific options, however moribund. It would take some extraordinary evidence to enliven them, but, if that were to happen, then empiricism itself — the crowning norm, we saw, of naturalized epistemology — would go by the board. For remember that that norm, and naturalized epistemology itself, are integral to science, and science is fallible and corrigible.” 4 See, e.g., Quine (1960), 22-3, also 264. 5 The focus centers on the theory-evidence relation, one which, as Quine remarks (1969d), 83: “...we are prompted to study for somewhat the same reasons that always prompted epistemology; namely, in order to see how evidence relates to theory, and in what ways one’s theory of nature transcends any available evidence.” 6 Quine (1990), 37-8. Quine emphasizes the role of his behaviorism as a theory of explanation in many places throughout the course of his writings. It is strongly implicit in remarks such as: “Language is a social art. In acquiring it we have to depend entirely on intersubjectively available cues as what to say and when. Hence there is no justification for collating linguistic meanings, unless in terms of men’s dispositions to respond overtly to socially observable stimulations.” Quine (1960), “Preface”, ix. It is explicit in remarks such as: “The more special point is that verbal behavior is determined by what people can observe of one another’s responses to what people can observe of one another’s verbal stimulations. In learning language, all of us, from babyhood up, are amateur students of behavior, and, simultaneously, subjects of amateur studies of behavior...For empiricism as thus reoriented, the focus of understanding is outside us...This is behaviorism.” Quine (1970), 4: “In all we may distinguish three levels of purported explanation, three degrees of depth: the mental, the behavioural, and the physiological. The mental is the most superficial of these, scarcely deserving the name of explanation. The physiological is the deepest and most ambitious, and it is the place for causal explanations. The behavioural level, in between, is what we must settle for in our descriptions of language, in our formulations of language rules, and in our explications of semantical terms. It is here, if anywhere, that we must give our account of the understanding of an expression, and our account of the equivalence that holds between an expression and its translation or paraphrase.” Quine (1975), 87. 7 The locus classicus of this criticism is Chomsky (1959), 26-58. 1 2

280 8

PAUL A. ROTH

The debate has many iterations, representative being Chomsky (1969) and Quine (1969e). Quine remains unswayed from this view. Quine (1990), 37-8: “Critics have said that the thesis [of the indeterminacy of translation] is a consequence of my behaviorism...I agree...Each of us learns his language by observing other people’s verbal behavior and having his own faltering verbal behavior observed and reinforced or corrected by others. We depend strictly on overt behavior in observable situations. As long as our command of our language fits all external check-points, where our utterance or our reaction to someone’s utterance can be appraised in the light of some shared situation, so long all is well. Our mental life between check-points is indifferent to our rating as a master of the language. There is nothing in linguistic meaning beyond what is to be gleaned from overt behavior in observable circumstances.” 9 See, e.g., Quine (1997b), 571-2; and (1995b), 251-61. See also my discussion of this topic in Roth (1999). 10 George (1986), 489-99, esp. 491-2. 11 Ibid., 492; see also 494. 12 Quine (1980b), 258-9. 13 Chomsky (1969), 66. 14 Ibid. 15 See, e.g., Chomsky (1978), 12-24. 16 Chomsky (1969), 63. 17 Ibid. 18 Quine (1969e) 302 19 Ibid., 304. Pinker’s work indicates that nothing has changed over the intervening 30 years since the exchanges before, in, and after Words and Objections with regard to the inadequacies that Chomskians takes to be inherent in Quine’s position. Pinker structures the dispute as a disjunctive syllogism: either explain language acquisition and use in terms of local/cultural conditioning (what he also terms the Standard Social Science Model — SSSM — aka the Sapir/Whorf view), or explain it by postulating innate, universally shared concepts existing prior to (and as necessary to account for) language acquisition. Quine is among those identified as holding the SSSM. Sufficient to convict Quine of this view is that Quine insists, “The most notable norm of naturalized epistemology...[and] the watchword of empiricism: nihil in mente quod non prius in sensu. This is a prime specimen of naturalized epistemology, for it is a finding of natural science itself, however fallible, that our information about the world comes only through impacts on our sensory receptors.” Quine (1990), 19. Yet, Pinker responds, by this very precept empiricism stands refuted: “Grammar offers a clear refutation of the empiricist doctrine that there is nothing in the mind that was not first in the senses...[The] paraphernalia of syntax are colorless, odorless and tasteless, but they, or something like them, must be a part of our unconscious mental life.” (Pinker (1994), 117) There is more in mind, on Pinker’s view, than can be dreamed of in Quine’s philosophy. See Pinker (1994) e.g., 11, 72-3, and esp. 419-48. This characterization of Quine’s views on language acquisition retains wide currency. Note the easy attribution of this view in Gosselin (2000), 57-77. For Quine’s skeptical take on this view, see, e.g., (1970), 15. 20 Quine (1969b), 26. 21 The point to emphasize, in order to achieve the intended contrast between the indeterminacy of translation and the more general underdetermination of (any) physical theory is that, once one settles on a physical theory, one has a basis for attributing reality to objects and truth to sentences. But, Quine insists, settling on a choice of physical theory fails to provide suitable justification for ascribing reality to mental states or for differentiating among truths by marking out some as necessary and some not. “Though linguistics is of course a part of the theory of nature, the indeterminacy of translation is not just inherited as a special case of the underdetermination of our theory of nature. It is parallel but additional....The point about indeterminacy of translation is that it withstands even all this truth, the whole truth about nature...[W]here indeterminacy of translation applies, there is no

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

281

real question of right choice; there is no fact of the matter even to within the acknowledged underdetermination of a theory of nature.” Quine (1969e), 303. 22 A sympathetic effort to exposit Quine’s notion of behaviorism and distinguish it from other senses that the term has in the literature can be found in Rappaport (1978), 164-83. 23 Quine (1980a) 26-30; see esp. 26, 29. The exposition by Sellars Quine refers to is Sellars (1980), 3-25. 24 Sellars (1980), 4, §9. 25 Ibid., 5. 26 Ibid., 6-7. 27 Putnam (1980), 249. The context of the quote is worth noting as well. “Just how impressed should we be by the failure of current learning theories to account for complex learning processes such as those involved in the learning of language? If innateness were a general solution, perhaps we should be impressed. But the I.H. cannot, by its very nature, be generalized to handle all complex learning processes ...the theorems of mathematics, the solutions to puzzles, etc., cannot on any theory be individually ‘innate’; what must be ‘innate’ are heuristics, i.e., learning strategies. In the absence of any knowledge of what general multipurpose learning strategies might even look like, the assertion that such strategies (which absolutely must exist and be employed by all humans) cannot account for this or that learning process, that the answer or an answer schema must be ‘innate’, is utterly unfounded.” Quine (1969e), 307, cites with strong approval exactly this criticism. 28 Kripke’s (1982) work on Wittgenstein remains exemplary here. I discuss the futility of hypothesizing rules as explanatory of behavior in Roth, “Mistakes,” unpublished. 29 Quine, (1980a), 26: “His [Sellars] moderate behaviorism is exactly to my taste. Mentalistic predicates can be tolerated in the manner of theoretical predicates of physics, e.g. electron spin, or even electron. For them there is no observational criterion, except as these predicates contributes to the coherence and simplicity of an inclusive theory for which there is observational support as a whole. In a word, we can admit them as hidden variables. Among hidden variables, however, there are better and worse. There is a premium on any links to observation, however partial and indirect; the less partial the better, and the more direct the better. The importance of behaviorism is its insistence on shoring up mentalistic terms, where possible, by forging substantial links with observation. For a deep causal explanation of mental states and events, on the other hand, we must look not just to behavior but to neurology. For this reason there is a premium not only on substantial connections between our hidden variables and observation, but also on the amenability of these hidden variables to explanatory hypotheses in neurology. Their value lies in fostering causal explanations.” 30 Quine, (1969c), 123. 31 Quine (1990), 44: “Unlike Davidson, I still locate the stimulations at the subject’s surface, and private stimulus meaning with them. But they maybe as idiosyncratic, for all I care, as the subject’s internal wiring itself. What floats in the open air is our common language, which each of us is free to internalize in his peculiar neural way. Language is where intersubjectivity sets in. Communication is well named.” Ibid., 43-4: “What is utterly factual is just the fluency of conversation and the effectiveness of negotiation that one or another manual of translation serves to induce...Such was my parable of the trimmed bushes, alike in outward form but wildly unlike in their inward twigs and branches. The outward uniformity is imposed by society, in inculcating language and pressing for smooth communication.” 32 This I take to be a key point in a critical but difficult passage in which Quine summarizes his disagreements with Chomsky. Quine (1969e), 304: “A conviction persists, often unacknowledged, that our sentences express ideas, and express these ideas rather than those, even when behavioral criteria can never say which. There is the stubborn notion that we can tell intuitively which idea someone’s sentence expresses, our sentence anyway, even when the intuition is irreducible to behavioral criteria. This is why one thinks that one’s question ‘What did the native say?’ has a right answer independent of choices among mutually incompatible manuals of translation.” On my account, what is “irreducible to behavioral criteria” just is the claim that most categorizations of so-called meaningful behaviors sort into kinds in the way

282

PAUL A. ROTH

that behavioral discriminations of color do. 33 As will become apparent, although I endorse Kripke’s observation that Quine’s route to indeterminacy goes through a skepticism about meaning based on behavioral behavior while Wittgenstein’s skepticism goes through introspection, he underrates the full force of Quine’s argument as developed below. Kripke (1982), 14-5, 55-8. Both Quine and Wittgenstein present fundamental challenges to the folk psychological assumption that the terms of folk psychology mark out “kinds” distinguishable by some common behavioral or introspectible property. See esp. discussion of Goldfarb in Section III. 34 Quine (1972), 446. 35 Ibid. 36 ibid., 445. 37 Quine is looking for something in the public realm that marks the same as the same. Here it is possible to see Quine and Wittgenstein converging, as Kripke suggests, inasmuch as Wittgenstein is read as skeptical that anything in introspection reveals a mark marking things as the same, even when the “sameness” at issue is the rule one allegedly follows in doing what one does. More on this below. 38 Quine (1976c), 233-41. 39 See esp. Quine (1960), 221. 40 Quine (1976b). 41 Ibid., 72 42 Ibid., 77 43 Ibid., 89-90: “Let us now consider the protest...that our freedom in assigning truth by convention is subject to restrictions imposed by the requirement of consistency. Under the fiction, implicit in an earlier stage of our discussion, that we check off our truths one by one in an exhaustive list of expressions, consistency in the assignment of truth is nothing more than a special case of conformity to usage...It is only the objective of ending up with our mother tongue that dissuades us from marking both ‘—‘ and ‘~ ‘ [as true], and this objective would dissuade us also from marking ‘it is always cold on Thursday’...As theoretical restrictions upon our freedom in the conventional assignment of truth, requirements of consistency thus disappear. Preconceived usage may lead us to stack the cards, but does not enter the rules of the game.” 44 See, e.g., Ibid., 93 45 Ibid., 95 46 Ibid. 47 98-9, emphases mine 48 Ricketts(1982), 117-36. 49 Ibid., 125 50 Ibid., 126: “Analyticity is then revealed by speakers’ reflective responses to suitably designed contrary-to-fact questionnaires. So, Carnap would have us attribute to a scientist a linguistic framework whose logical-consequence relations mirrors the scientist’s intuitions of rational unrevisability.” 51 Ibid., 127 52 Ibid., 132. 53 Ibid., 135 54 Ibid., 135-6. 55 Ibid. 56 Ibid., 136. 57 See Quine (1972), 442-54, esp. 442-5. 58 Quine (1990), 71: “The point of anomalous monism is just that our mentalistic predicates imposes on bodily states and events a grouping that cannot be defined in the special vocabulary of physiology. Each of those individual states and events is physiologically describable, we presume, given all pertinent information.” Quine (1995a), 88. Likewise, Quine maintains, “What are irreducibly mental are ways of grouping them [physical, e.g.,

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

283

neural events]: grouping a lot of respectably physical perceptions as perception that p. I acquiesce in what Davidson calls anomalous monism, also known as token physicalism: there is no mental substance, but there are irreducibly mental ways of grouping physical states and events. The keynote of the mental is not the mind; it is the content-clause syntax, the idiom ‘that p’.” 59 Quine (1960), 22 60 Consider the following example. A substance may or may not be salt. Categorizing something as a salt, and taking NaCl as its chemical structure are all, let us say for the sake of argument, products of a theory. But once that theory is in hand, whether or not something is salt can be determined by looking at its chemical structure. The chemical structure marks the substance as belonging or not to a certain kind, whether or not that kind and its markers are artifacts of a contingently held theory. In claiming that behavior is intentional, in contrast, no objective markers can be given as in the case of chemical structure. 61 Quine (1960), 216 ff. 62 As I argued in chapter 2 of Roth (1987), overt behavioral provides the only plausible criterion for Quine has for determining such key notions such as, e.g., sameness of stimulus meaning. Indeed, in later writings, Quine gives up the notion of “same stimulus meaning” altogether in favor of either basically behavioral or assumed genetic accounts. See, e.g., Quine (1996), 159-63, esp. 162-3 and (1997a), 575-7, esp. 576. 63 Quine (1969e), 306. 64 The issue of the possible explanatory efficacy of appeal to mental states is explicit in Quine (1975), 86-7. 65 Quine (1990), 101 66 Føllesdal (1990), 98: “Another version of behaviorism, which I call ‘evidential’ behaviorism, and which I regard as that of Quine, is a position concerning evidence: the only evidence we can build on in our study of man, as in any other scientific study, is empirical evidence, in particular the observation of behavior. A Behaviorist of this kind is open to accepting mental states to the extent that there is evidence for them.” 67 Ibid, 106-7: “Traditionally, philosophers of mind have tended to identify mental states by help of language. Two speakers are in the same mental state if and only if their states are appropriately expressed by the same or synonymous expressions...But thereby we are back full circle. We asked what it means for two mental states to be the same, we were told that they are the same if they are expressed by synonyms. And we are now told that for two expressions to be synonyms, they must express the same mental state. One might try to avoid the circularity by finding other ways of correlating different people’s mental states, without going by way of language...[O]ne might perhaps hold that two mental states are the same if their physical aspects are the same, or very similar. However, given the individual differences between human brains and also the holistic nature of the mental, such a proposal, if anybody were to make it, would seem highly implausible.” 68 Quine (1980a), 29: “Anyway, this question of uniformity of neural hookup is not relevant to the problem of explicating meaning, or likeness of meaning. We are already assuming this relation, after all, when we speculate on the neural mechanism that make for like meanings.’ 69 Quine (1961b), 30. 70 Quine (1972), 448. 71 Goldfarb (1992), 109-22. 72 Ibid., 109-10. 73 Ibid., 118: “The note Wittgenstein is here sounding has a Quinean ring. In fact, I do see a confluence between the two at this point...[T]here exists little reason to think there will be a scientifically discoverable state that will group together just the right things under the concept of understanding...You have a mere metaphysical insistence rather than anything that comes from a scientific worldview.” 74 Ibid., 109. 75 Ibid.

284 76 77

PAUL A. ROTH

Ibid., 113. Ibid., 121: “[T]he inclination to think that science might discover an appropriate state rests on an accession to a picture of the mental apparatus, of definite states and processes of some sort; given those states and processes, questions can then be framed of whether they can be tied down physically. If, on the contrary, the only apparatus assumed is physical, then upon scrutiny it becomes clear how little conception we have of what is supposed to be discovered.” Ibid., 118. Or, again, “[O]ne has to undermine a picture of the mental apparatus at the start, and emphasize the intricacy, connectedness, and nonuniformity of our practices of ascription, so as to make clear how different this is from the case of mechanisms in the physical world. The scientific objector always has examples form the physical sciences in mind. We discovered the internal constitution of ammonia, after all...To this we ought respond: understanding is not like ammonia. Wittgenstein’s investigations are meant to bring us to see how different the cases are.” 78 Ibid.: “[W]e shall lose the idea of the specific particular process or state that is the understanding, along with the idea that we need such a thing. That is, Wittgenstein’s investigations are to have the effect of making entirely obscure what could be scientifically discoverable that would be the mental state-or-process of understanding — since, if you will, there is nothing for it to be; there is no role for it to play.” 79 Ibid. 80 Ibid., 112: “For Wittgenstein, it is characteristic of the notions that figure in philosophical problems — prominently, mental concepts and linguistic concepts like meaning — that a structure is imposed on them, without grounding in the ordinary use of these notions and without being noticed, when they are taken to be amenable to certain explanatory projects...Hence, only through clarification of what the legitimate questions are can proper sense be made of the applicability of science. A scientistic viewpoint ignores this need for clarification. As a result, for Wittgestein scientism is just as misguidedly metaphysical as traditional, more transparently a prioristic, approaches.” Goldfarb expands on this point in a powerful and illuminating way, one which makes clear the overlap I suggest between the Quinean and Wittgensteinian strategies at this point. Ibid., 116-7: “The variegation and lack of uniformity make clear how far any unitary state or process is from what we observe in the cases of understanding, and thus emphasizes the leap made n postulating such a ‘hidden’ state or process. Any of the features that play a role in an ascription of understanding would have to be construed as evidence for this hidden state. But our grasp on the state-evidence model is cast into doubt by the variability of what and when any particular feature functions as evidence. For we lack any conception of how the state is connected to just the particular things that matter in a particular context. If we say that in each particular case there is some particular connection, then all we are saying is that the particular circumstances justify me in my ascription; the one state drops out. But what conception can we have of a general connection that issues in such variability?...[T]his is a notion of internal standard gone wild. It also look as if the state has to be able to take account of all possible circumstances. But that is not a well-defined totality...If, however, we return to the scientific objection, with its envisaged brain states and processes and its recourse to physical causation as linking state and manifestation, the upshot of that range and variegation is fairly straightforward. They show that we have little grasp f how to identify what is at issue in the brain...If the objector is saying that we could in principle discover that which biochemically causes us to behave in ways that a person who understands does to this we can now point out that the latter is simply not definite enough: there is no surveyable checklist of such behaviors...”. 81 For an interesting and important discussion of potential issues developing from how Quinean themes play out against a Kantian back-drop in contemporary philosophical discussion, see Friedman (1996), 427-67.

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

285

REFERENCES

Barrett, R. and Gibson, R. (editors). Perspectives on Quine. Oxford: Blackwell, 1990. Chomsky, Noam. “A Review of B.F. Skinner’s Verbal Behavior,” Language, 35, 1959. Chomsky, Noam. “Quine’s Empirical Assumptions”, in Davidson and Hintikka, Words and Objections, 1969. Chomsky, Noam. Rules and Representation. New York: Columbia University Press, 1978. Davidson, Donald and Hintiikka, Jaakko (editors).Words and Objections. Boston: Reidel, 1969. Decock, L. (editor). Quine: Naturalized Epistemology, Perceptual Knowledge and Ontology. Amsterdam: Rodopi, 2000. Føllesdal, Dagfinn. “Indeterminacy and Mental States,” in Barrett and Gibson, Perspectives on Quine, 1990. Friedman, Michael. “Exorcising the Philosophical Tradition: Comments on John McDowell’s Mind and World,” The Philosophical Review, 105, 1996. George, Alexander. “Whence and Whither the Debate between Quine and Chomsky?”, The Journal of Philosophy, 83, 1986. Goldfarb, Warren. “Wittgenstein on Understanding,” in Midwest Studies in Philosophy, XVII: The Wittgensteinian Legacy, P. French et al., eds. Notre Dame: University of Notre Dame Press, 1992. Gosselin, Mia. “Quine’s Hypothetical theory of Language Learning,” in L. Decock, ed., Quine: Naturalized Epistemology, Perceptual Knowledge and Ontology. Koppelberg, Dirk. “Quine and Davidson on the Structure of Empirical Knowledge” in L. Decock, Quine: Naturalized Epistemology, Perceptual Knowledge and Ontology. Kripke, Saul A. Wittgenstein on Rules and Private Language. Cambridge: Harvard University Press, 1982. Leonardi, Paolo. On Quine: New Essays. New York: Cambridge University Press, 1995. Morick, H. (editor). Challenges to Empiricism. Indianapolis: Hackett, 1980. Pinker, Steven. The Language Instinct. New York: HarperCollins, 1994. Putnam, H. “The ‘Innateness Hypothesis’ and Explanatory Models in Linguistics” in Morick, ed., Challenges to Empiricism, 1980. Quine, W.V. Word and Object. Cambridge: The MIT Press, 1960. Quine, W.V. From A Logical Point of View , 2nd Ed., Rev. New York: Harper & Row, 196la. Quine, W.V. “Two Dogmas of Empiricism,“ in From A Logical Point of View, 1961b.

286

PAUL A. ROTH

Quine, W.V. Ontological Relativity and Other Essays. New York: Columbia University Press, 1969a. Quine, W.V. “Ontological Relativity,” in Quine, Ontological Relativity and Other Essays, 1969b. Quine, W.V. “Natural Kinds,” in Ontological Relativity and Other Essays, 1969c. Quine, W.V. “Epistemology Naturalized,” in Quine, Ontological Relativity and Other Essays, 1969d. Quine, W.V. “Reply to Chomsky” in Davidson and Hintikka, ed., Words and Objections, 1969e. Quine, W.V. “Philosophical Progress in Language Theory,” Metaphilosophy,1, 1970. Quine, W.V. “Methodological Reflections on Current Linguistic Theory,” in D. Davidson and G. Harman, eds., The Semantics of Natural Language, 1972. Quine, W.V. “Mind and Verbal Dispositions,” in S. Guttenplan, ed. Mind and Language. New York: Oxford University Press, 1975. Quine, W.V. The Ways of Paradox and Other Essays. Rev. and enlarged edition. Cambridge: Harvard University Press, 1976a. Quine, W.V. “Truth by Convention,” in Quine, Ways of Paradox, 1976b. Quine, W.V. “Posits and Reality,” in Quine, Ways of Paradox, 1976c. Quine, W.V. “Sellars on Behaviorism, Language and Meaning,” Pacific Philosophical Quarterly, 61, 1980a. Quine, W.V. “Linguistics and Philosophy,” in Morick, ed., Challenges to Empiricism, 1980b. Quine, W.V. Pursuit of Truth. Cambridge: Harvard University Press, 1990. Quine, W.V. From Stimulus to Science. Cambridge: Harvard University Press, 1995a. Quine, W.V. “Naturalism; Or, Living Within One’s Means,” Dialectica, 49, 1995b. Quine, W.V. “Reactions”, in Leonardi, ed., On Quine: New Essays, 1995c. Quine, W.V. “Progress on Two Fronts,” The Journal of Philosophy, 93, 1996. Quine, W.V. “Reply to Lewis and Holdcroft” Revue Internationale de Philosophie, 51, 1997a. Quine, W.V. “Response to Haack,” Revue Internationale de Philosophie, 51, 1997b. Rappaport, Steven. “Quine’s Behaviorism,” Philosophical Research Archives, 4, 1978. Ricketts, Thomas. ”Rationality, Translation, and Epistemology Naturalized,” The Journal of Philosophy, 79, 1982. Roth, Paul A. Meaning and Methods in the Social Sciences. Ithaca: Cornell University Press, 1987. Roth, Paul A. “The Epistemology of ‘Epistemology Naturalized’,”

QUINE ON BEHAVIORISM, MEANING, INDETERMINACY

287

Dialectica, 53, 1999. Sellars, W. “Behaviorism, Language and Meaning,” Pacific Philosophical Quarterly, 61, 1980.

This page intentionally left blank

SELMER BRINGSJORD AND YINGRUI YANG

COGNITIVE ILLUSIONS AND THE WELCOME PSYCHOLOGISM OF LOGICIST ARTIFICIAL INTELLIGENCE

We begin by using Johnson-Laird’s ingenious cognitive illusions (in which it seems that certain propositions can be deduced from given information, but really can’t) to raise the spectre of a naive brand of psychologism. This brand of psychologism would seem to be unacceptable, but an attempt to refine it in keeping with some suggestive comments from Jacquette eventuates in a welcome version of psychologism — one that can fuel logicist (or logic-based) AI, and thereby lead to a new theory of contextindependent reasoning (mental metalogic) grounded in human psychology, but one poised for unprecedented exploitation by machines.

1. Introduction The overall goal of this chapter is two-fold: (i) sketch a perfectly acceptable brand of psychologism that can function as a productive foundation for logicist (= logic-based) artificial intelligence (AI); (ii) introduce a new theory of human context-independent reasoning as the cornerstone of this foundation. To reach our goal, we begin by using Johnson-Laird’s ingenious cognitive illusions (in which it seems — to most, anyway — that certain propositions can be deduced from given information, but really can’t) to raise the spectre of an unpalatable brand of psychologism. Once this brand of psychologism is refined, the promised welcome version of psychologism is produced (and, in fact, in part displayed in action), as is our new theory of human reasoning: mental metalogic, or just ‘MML,’ for short. We conclude by briefly presenting the context of our belief that though MML is firmly grounded in human reasoning and empirical psychology, it’s a theory poised for unprecedented exploitation by machines.

289 D. Jacquette (ed.), Philosophy, Psychology, and Psychologism, 289-312. © 2003 Kluwer Academic Publishers. Printed in the Netherlands.

290

SELMER BRINGSJORD AND YINGRUI YANG

2. Logical Illusions and a Naïve Brand of Psychologism Due in no small part to ingenious experimentation carried out by JohnsonLaird, one of the dominant themes in current psychology of reasoning is the notion of a cognitive or logical illusion. In a visual illusion of the sort with which you’re doubtless familiar, one seems to see something that, as a matter of objective fact, simply isn’t there. (In the Sahara, what seems to be a lovely pool of water is just the same dry-as-a-bone expanse of sand — and so on.) In illusions of the sort that Johnson-Laird has brought to our attention, what seems to hold on the basis of inference doesn’t. Here’s a first specimen:1 Illusion 1 (1) If there is a king in the hand, then there is an ace, or else if there isn’t a king in the hand, then there is an ace. (2) There is a king in the hand. Given these premises, what can you infer? Johnson-Laird has recently reported that: Only one person among the many distinguished cognitive scientists to whom we have given [Illusion 1] got the right answer; and we have observed it in public lectures — several hundred individuals from Stockholm to Seattle have drawn it, and no one has ever offered any other conclusion. (Johnson-Laird 1997b, 430)

The conclusion that nearly everyone draws (including, perhaps, you) is that there is an ace in the hand. Bringsjord (and, doubtless, many others) has time and time again, in public lectures, replicated Johnson-Laird’s numbers — presented in (Johnson-Laird & Savary, 1995) — among those not formally trained in logic. (The reason for the italicized text will become clear later.) This is wrong because ‘or else’ is to be understood as exclusive disjunction,2 so (using obvious symbolization) the two premises become:

It follows from the right (main) conjunct in (1' ) by DeMorgan’s Laws that either or is false. But by the truth-table3 for in which a conditional is false only when its antecedent is true while it’s consequent is false, it then follows that either way A is false, i.e., there is not an ace in the hand. Technical philosophers and logicians, in our experience, invariably succeed in carrying out this deduction. In fact, when

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

291

Bringsjord presented this problem at the 2000 annual International Computing and Philosophy Conference, two logicians rapidly not only produced the proof in question, but commented that “Of course, this assumes that you’re dealing with an exclusive disjunction.” Predictably, both these thinkers symbolized the problem in first-order logic (FOL). From the standpoint of discussing psychologism — put roughly for now, the view that correct reasoning is reasoning in accord not with abstract formalisms, but rather in accord with how humans in fact reason — Illusion 1 is uneventful. (On the other hand, from the standpoint of those who know a thing or two about logic, the fact that so many cognitive scientists went awry on this problem is remarkable.) But now consider a second logical illusion, one given again by Johnson-Laird (1997b, 430): Illusion 2 (*) Only one of the following assertions is true: (3) Albert is here or Betty is here, or both. (4) Charlie is here or Betty is here, or both. (5) This assertion is definitely true: Albert isn’t here and Charlie isn’t here. Given these premises, what can you infer? Johnson-Laird says: “These premises [in Illusion 2] yield the illusion that Betty is here” (1997b, 431). But upon closer inspection this pronouncement is peculiar. To see this, using A and B as obvious abbreviations, (5) becomes We know from (*) that (3) is true, or (4) is, but not both. Suppose that (3) is true; then by disjunctive syllogism on (3) and (A we obtain that Betty is here (B). Suppose instead that (4) is true; then by disjunctive syllogism on (4) and we obtain that Betty is here (B). Either way, contra Johnson-Laird, Betty is here.4 Bringsjord showed this simple proof (and other similar ones in connection with other “illusions” offered by Johnson-Laird) to JohnsonLaird himself; Bringsjord’s assumption was that Johnson-Laird would say that an error on his part had been made. That assumption turned out to be wrong. Johnson-Laird replied in personal communication that people don’t make inferences as sophisticated as the ones Bringsjord had made in dissecting Illusion 2, and that the illusion stands. Incredulous, Bringsjord pointed out that surely a logical illusion I is classified as such because the premises in I are taken by most subjects to entail while in fact cannot be deduced from using laws of logic. Johnson-Laird replied that logic is

292

SELMER BRINGSJORD AND YINGRUI YANG

not the arbiter in illusions; the arbiter by his lights is what most logically untrained subjects believe about what must be true given the truth of the propositions in As we see it, this position, and the attitude toward logic that accompanies it, is about as good a candidate for an instance of (a presumably objectionable) psychologism as can be found anywhere. But let’s slow down. Is Johnson-Laird’s position even coherent? Suppose that there is some purported illusion I with premises and that all subjects declare that can be deduced from If in fact all subjects make this declaration, then we have here a powerful illusion. (After all, that Illusion 2 is powerful is something Johnson-Laird attempts to substantiate by pointing out that audience after audience has declared that Betty is here can be inferred.) But if all subjects make this declaration, then by JohnsonLaird’s psychologistic stance it follows that it’s valid to infer from But since by hypothesis we’re dealing with an illusion, by hypothesis the inference isn’t valid! We thus have a contradiction, incoherence.5

3. Toward a Palatable Brand of Psychologism However, there’s a reading of Johnson-Laird’s response that provides him with a possible escape, and points the way toward a welcome version of psychologism. To begin, we need to be more precise about what an illusion is. Let denote declarative sentences in English (actually, any natural language will do, but that needn’t detain us). (So for example in Elusion 1 [the sentence] Only one of the following two assertions is true.6 could be Let Q stand for some question in English. Following the notation we used above, we can let capital Greek letters range over sets of formulas in some logical system (e.g., first-order logic, or just ‘FOL’), and lowercase Greek letters refer to individual formulas. Next, notice that there is some variation in the questions. Sometimes the question isn’t open-ended, as it is in Illusions 1 and 2. Instead, sometimes it is a query about whether a particular proposition (or state of affairs) is “possible.” Here’s an example:7 Illusion 3

(6) Only one of the following premises is true about a particular hand of cards. (7) There is a king in the hand or there is an ace, or both. (8) There is a queen in the hand or there is an ace, or both. (9) There is a jack in the hand or there is a ten, or both. Is it possible that there is an ace in the hand?

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

293

Now obviously, at least at this point, we don’t want to call upon some of the recherché machinery of symbolic logic (e.g., modal logic) to formally analyze the question here as one involving a possibility operator. Fortunately, the question is clearly intended to be a question about consistency in the standard, first-order, extensional sense.8 That is, the question here is whether the declarative sentence There is an ace in the hand is consistent with (6)-(9). (Is it?)9 More generally, we will say that the question Q can be one of three types: open-ended (it can ask what can be inferred); a question about whether (a) given inferential relation(s) hold(s) between the or a question about whether (a) given inferential relation(s) hold(s) between the and another sentence (such as There is an ace in the hand in Illusion 3). Because the are concerned with an inferential relation, it would seem to make sense to stipulate that one of the arguments in this relation is the “single turnstyle,” which traditionally denotes provability; that is, indicates that can be proved from Putting together what we have so far gives us the following scheme for cognitive illusions:

But how can we describe the structure of the illusions themselves, rather than just the presentation of them? Well, first, to ease exposition, but without loss of generality, hereafter suppose that we have an illusion specifically of the variety; and suppose, specifically, that the inferential relation in question is Assuming this framework, we might venture that an illusion obtains when two conditions are met:

294

SELMER BRINGSJORD AND YINGRUI YANG

A cognizer c succumbs to a logical illusion

if and only if: 1. c believes that 2. NOT

can be inferred from

i.e.,

Unfortunately, this account doesn’t help Johnson-Laird: on it, Illusion 2 isn’t a cognitive illusion, since Betty is here can in fact be derived from the relevant knowledge (as we’ve shown).10 However, some additional elements can change the story dramatically — elements that reflect sensitivity to Johnson-Laird’s seemingly unpalatable notion that how reasoners reason should determine what’s “valid” and “invalid.” Specifically, here’s what we have in mind: We can turn to a more sophisticated version of our first definition, by including some indication of what system S underlies

A cognizer c succumbs to a logical illusion

if and only if: 1. c believes that 2. NOT

can be inferred from

in S, i.e.,

For example, given we can speak of deduction in accordance with a full, standard set of natural deduction rules — a set that includes the rules of inference we

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

295

used in our proof that Betty is here. In such a set, for example in the elegantly presented set that’s part of the natural deduction system in Barwise and Etchemendy’s Language, Proof, and Logic (1999), there is always a rule for introducing a logical connective, and a rule for eliminating such a connective. Specifically, then, there is a rule for eliminating and a rule for introducing The first of these corresponds to so-called constructive dilemma, and it’s a rule that is used in our proof that Betty is here. To see this, a formal version of the proof in is shown in Figure 2. Where is here composed of the two Givens, this proof establishes that:

But now we are finally in position to see how to make respectable Johnson-Laird’s response to this result: We can read him to be presupposing that the S in is simply indexed to the population in question. For example, suppose that a particular implication can only be proved if rule r is employed; to completely fix the situation, suppose that this

296

SELMER BRINGSJORD AND YINGRUI YANG

implication goes through only if

is available. In this case,

but the subjects in question may nonetheless believe that B is provable in the relevant manner. It seems possible that what Johnson-Laird may have in mind is that untrained reasoners at least generally believe that Betty is here can be inferred in a system without constructive dilemma, and that this belief may be incorrect. The general notion that untrained reasoners lack an understanding of highly specific patterns of primitive inference is not unprecedented in psychology. Indeed, apparently the view that such reasoners lack an understanding of reductio ad absurdum is widespread among psychologists of reasoning. Johnson-Laird, in personal communication, has told us that that is in fact his view; many other eminent psychologists of reasoning are of the same view (e.g., Braine). And more concretely, there is the example of what Lance Rips (1994) seems to have done in devising his PSYCOP system, for this system includes (e.g.) conditional proof, set out on page 116 of (Rips, 1994) as

but not the rule which sanctions passing from the denial of a conditional to the truth of this conditional’s antecedent (a rule that will be central to our concluding illusion, Illusion 4). For example, Rips (1994, 125-26) says that this inference cannot be made by PSYCOP: NOT (IF Calvin passes history THEN Calvin will graduate). Calvin passes history. At this point it may be safe to say that we have sketched a formal account of logical illusions on which Johnson-Laird’s response to the provability of B in Illusion 2 is coherent. But how does this account relate to the promised welcome version of psychologism? We answer this question now.

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

297

4. Toward a Welcome Brand of Psychologism Jacquette anticipates an approximation of the brand of psychologism we have in mind when he writes: In engineering one learns how to build a bridge ‘correctly’ so that it will bear its load properly, withstand wind shear and other kinds of stress, minimize mental fatigue and the like; similarly, in logic one learns how to reason ‘correctly’ to draw sound inferences and avoid fallacious ones. When we learn to build bridges or perform heart surgery or reason correctly, we also learn what not to do relative to presumed practical purposes. We want the bridge to span that gorge safely and not collapse, the patient to recover with improved health and not die on the operating table, and our reasoning to expand our knowledge and improve our decision-making ability, and not lead us from truth to falsehood. Now, I think that most theorists would be reluctant to conclude that engineering and medicine are not reducible to physics and biology, but rather that their respective practices combine these sciences with an assumption about the goals they should help their practioners to achieve. Why should things be different with respect to logic as an applied discipline grounded in the psychology of reasoning? (Jacquette 1997, 324-25)

The version of psychologism that we have in mind involves a relationship not between what Jacquette calls “applied logic” and psychology of reasoning, but rather between a field that marks the marriage of engineering and logic, and the psychology of reasoning; the field is: logicist (= logic-based) AI. Logicist AI, or just ‘LAI’ for short, is devoted to engineering machines capable of simulating intelligent behaviors of various sorts, where these machines carry out their relevant work via formal reasoning. (A full introduction to LAI can be found in Bringsjord & Ferrucci (1998). For a book-length treatment of LAI as an approach to machine creativity see Bringsjord & Ferrucci, 2000.) Though LAI isn’t exactly what Jacquette ascribed the term ‘applied logic’ to, this term is certainly an appropriate one for the kind of AI we’re talking about. We believe that LAI is, or at least ought to be, to use Jacquette’s phrase, “grounded in the psychology of reasoning.” On this view, the process of building the intelligent machines in question involves looking to human reasoning for guidance; as this building proceeds, new knowledge of value to psychologists of reasoning is produced; and the process iterates, back and forth. Now, in order to see this looping reciprocity in action here, let’s suppose that we wish to build a computing machine capable of solving English-expressed logic problems, including logical illusions. A bit more precisely, we want to construct an artificial intelligent agent A that takes some such problem as input, and produces for output an answer A that constitutes a solution.11 We would also like to be able to modify A so that it succumbs to logical illusions, and thereby offers a mechanical instantiation of the human reasoning that takes place when a logical illusion “tricks”

298

SELMER BRINGSJORD AND YINGRUI YANG

some human. What we have learned so far about the underlying structure of logical illusions should be helpful in working toward such an intelligent agent, but is it? Not very; for only a little reflection from the perspective of LAI and the goal of building A reveals that is inadequate. Why? There are many reasons; here’s a fatal one. When you aim at building A you immediately see that we have swept under the rug the thorny issue of how to go from the English to (in this case) FOL. Our artificial agent is going to scan the English in, parse it, and represent the English in FOL.12 But we have completely left this process out of and Suppose (and notice now that we’re moving back for guidance to the human sphere) that a human looks at Illusion 1 and represents English sentences (1) and (2) as P and Q, respectively; and suppose that she represents There is an ace in the hand as A. Suppose as well that the set of inference rules in question, S, is a standard Fitch-style set at the propositional level (such as prepositional Clearly, not in this set; this implies that the second clause of is satisfied. But now suppose that this human also happens to believe that A can be derived from these premises, for (irrational) reasons having to do with the price of tea in China. At this point, both clauses in the definition are satisfied — but no one would want to say that this human has fallen prey to the illusion in question. Ergo, can’t be right, and should presumably be supplanted with: A cognizer succumbs to a logical illusion

if and only if: la. represents the as and the “target” English sentence in the query as 1. believes that can be inferred from in S, i.e., 2. NOT Do we have here an acceptable account? Unfortunately, no; counterexamples are still easy to come by: Suppose that Jones “correctly” represents the premises in Illusion 1 as believes that A, where S is appropriately instantiated — but: suppose that this belief, as before, has an

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

299

irrational justification, namely that because the price of tea in China has reached exorbitant levels, the implication in question is correct. Certainly we wouldn’t want to say that Jones has been affected by Illusion 1. What’s missing is that Jones doesn’t have in mind the right sort of justification for believing A. It would seem that those who succumb to Illusion 1 believe that A follows from the premises on the strength of a relevant argument — it’s just that the argument isn’t sound. For example, in our experience, when interviewed subsequent to giving their response in Illusion 1 that the presence of the ace in the hand can be deduced, most subjects give such rationales as: “Well look, either there is a king in the hand or there isn’t, but we know that either way there is an ace.” It would seem that subjects tripped up because they have in mind an inchoate, enthymematic argument establishing as a true equation:

Let’s take stock of where we are, then, from the standpoint of LAI. Apparently there are two tasks to consider: the representation of the English in an illusion in formal form, and the judgment as to whether this formal knowledge can be used to (given, as you’ll recall, that we are limiting ourselves to illusions of form) establish Given this, there are three possibilities, in general:13

Obviously, our definition needs to take account of these possibilities via suitable disjunctions. Where stands for a correct representation and an incorrect one, we have:14 A cognizer succumbs to a logical illusion

300

SELMER BRINGSJORD AND YINGRUI YANG

if and only if either: 1a', c represents the as and the “target” English sentence in the query as 1'. c believes, on the basis of a relevant invalid argument that can be inferred from in S, i.e., 2. NOT or

1a. c represents the as and the “target” English sentence in the query as 1". c believes, on the basis of a relevant valid argument that can be inferred from in S, i.e., 2. NOT

or 1a', c represents the as and the “target” English sentence in the query as 1'. c believes, on the basis of a relevant invalid argument that can be inferred from in S, i.e., 2. NOT This account is perhaps approaching respectability, but it’s doing only that: approaching: lots of work remains to be done. We can perhaps assume that standard natural language processing techniques will secure for A from the But what about the incorrect representation of the English (denoted by How is this representation produced, specifically? And what about the notion of a “relevant argument”? What does this amount to, exactly? Without answers to these questions, our agent A will not be attainable. Though we haven’t the space to provide the answers, we can say a few words, with help from our core examples. Take Illusion 1 first. In this case, as we’ve already indicated,

and and the “relevant valid argument” is shown in Figure 3. Now, what about Illusion 2? Can we analyze it in terms of

Yes, as

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

301

follows. First, there are two possibilities: either Johnson-Laird didn’t err, and our charitable reading of him on this illusion is correct, or he did in fact simply slip up. Either way we can make sense of Illusion 2. Suppose first that Johnson-Laird did make a mistake in writing that Illusion 2 is an illusion despite the fact that B is provable in elementary logic. Then (with a standard set of inference rules) what he apparently wanted was English that when represented would yield and Our hypothetical cognizers could for example then produce the incorrect representation: along with There would then be a straightforward valid proof of B. On the other hand, if Johnson-Laird didn’t err (because he had in mind the subtle variations on the variable S in then B can’t be obtained 15 because constructive dilemma is unavailable. But would our cognizer have a correct or incorrect representation? She can’t have a correct representation, and at the same time believe that

she can obtain B. The reason is that this belief would be based on an argument that itself deploys constructive dilemma. Hence, if Johnson-Laird didn’t make a mistake, cognizers must have on hand some from which you can get B in this restricted set of inference rules. At this point, perhaps can be said to be at least somewhat promising. But surely it’s still incomplete, if for no other reason than that it simply doesn’t present all the permutations that arise from the formal structure we have set out. For example, it should be obvious that we would have an illusion if we simply “flip the switches” in first disjunction in straightforward ways. To be more specific, we would have an illusion of the variety if subjects were tricked into thinking that some proposition can’t be derived from the correct representation of the premises, when in

SELMER BRINGSJORD AND YINGRUI YANG

302

fact

This would be the relevant variation on

disjunction:

1a'. represents the as and the “target” English sentence in the query as 1'. believes, on the basis of a relevant invalid argument that cannot be inferred from in S, i.e., 2'. Now, what about generating illusions? If is any good, we should be able to use it to construct an artificial agent that generates bona fide illusions before any data is gathered. This agent should yield some rather tricky illusions, given that we have now identified six variations to instantiate (two for each of the disjuncts in and that individual illusions can actually instantiate at least two of these variations simultaneously. Of course, without a specification of the relationship between correct and incorrect representations, and of justifying arguments, we concede that we can’t literally construct the agents we seek, but we can certainly at this point “hand simulate” the processes involved. Accordingly, here’s a new illusion Bringsjord derived from the analysis we’ve produced to this point:

Illusion 4

(10)

The following three assertions are either all true or all false: If Billy helped, Doreen helped. If Doreen helped, Frank did as well. If Frank helped, so did Emma.

(11)

The following assertion is definitely true: Billy helped.

Can it be inferred from (10) and (11) that Emma helped? (If you’re interested in pondering this illusion without knowing the answer, pause before reading the rest of this paragraph.) Sure enough, in our preliminary results, most students, and even most professors, often including, in both groups, even those who have had a course in logic, answer with “No.” The rationale they give is that, “Well, if all three of the if-thens are true, then you can chain through them, starting with ‘Billy helped,’ to prove that Harry helped. But you can’t do that if the three are false, and we’re told they might be.” But this is an illusion. In the case of reports like these, the premises are represented correctly, but there is an invalid argument for the belief that E can’t be deduced from the conjoined negated

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

303

conditionals (while they do have in mind a valid argument from the true conditionals and Billy helped to Emma helped). This argument is invalid because if the conditionals are negated, then we have:

The first two of these yields by propositional logic the contradiction Given this, if we assume that Emma failed to help, that is, we can conclude by reductio ad absurdum that she did help, that is, E.16 Notice that this solution makes use of both the rule that Rips says humans don’t have (a negated conditional to the truth of its antecedent) and the rule that most psychologists of reasoning say humans don’t have (reductio). In light of this, it’s noteworthy that a number of those to whom we presented Illusion 4 did produce the full solution we just explained. We next draw your attention to an interesting feature of Illusion 4, and use it as a stepping stone to our theory of human reasoning: mental metalogic.

5. Mental MetaLogic: A Glimpse Illusion 4 instantiates a variation in which cognizers conceive of disproofs. More specifically, a cognizer fooled by Illusion 4 imagines an argument for the view that there is no way to derive Emma helped from the negated trio of conditionals and Bill helped. Such arguments cannot be expressed in And when — in keeping with the version of psychologism that is guiding us — we look to psychology of reasoning for help in making sense of such arguments we draw a blank. The reason for this is that psychology of reasoning is dominated by two competing models of human reasoning (especially deductive reasoning), and neither one allows for the expression of disproofs. The two theories are mental logic (ML) and mental models (MM).

304

SELMER BRINGSJORD AND YINGRUI YANG

ML is championed by Braine (1998) and Rips (1994), who hold that human reasoning is based on syntactic rules like modus ponens (recall yet again the rule we earlier pulled from Rips (1994) and displayed). MM is championed by Johnson-Laird (1983), who holds that humans reason on the basis not of sequential inferences over sentence-like objects, but rather over possibilities or scenarios they in some sense imagine. As many readers will know, the meta-theoretical component of modern symbolic logic covers formal properties (e.g., soundness and completeness) that bridge the syntactic and semantic components of logical systems. So, in selecting from either but not both of the syntactic or semantic components of logical systems, psychologists of reasoning have, in a sense, broken the bridges built by logicians and mathematicians. Mental logic theory and mental model theory are incompatible from the perspective of symbolic logic: the former is explicitly and uncompromisingly syntactic, while the latter is explicitly and uncompromisingly semantic. Our theory, mental metalogic (or just ‘MML’), bridges between these two theories (see Figure 5). More concretely, MML provides a mechanism which our agent A will need if it is to simulate human cognizers who have beliefs about provability on the basis of valid and invalid disproofs. In this mechanism, step-by-step proofs (including, specifically, disproofs) can at once be syntactic and semantic, because situations can enter directly into line-by-line proofs. The HYPERPROOF system of Barwise and Etchemendy (1994) can be conveniently viewed as a simplistic instantiation of part of MML.17 In HYPERPROOF, one can prove

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

305

such things as that NOT Accordingly, let’s suppose that in Illusion 4, a “tricked” cognizer moves from a correct representation of the premises when (10)’s conditionals are all true along with (11), to an incorrect representation when the conditionals in (10) are false. Suppose, specifically, that the negated conditionals give rise to a situation, envisaged by the cognizer, in which four people (objects represented as cubes) b, d, f, and e are present, the sentence Billy is happy is explicitly represented by a corresponding formula, and the issue is whether it follows from this given information that Emma is happy. This situation is shown in Figure 6. (We have moved from helping to happiness because HYPERPROOF has ‘Happy’ as a built-in predicate.) Notice that the full logical import of the negated conditionals is nowhere to be found in this figure. Next, given this starting situation, a disproof in HYPERPROOF is shown in Figure 7. Notice that a new, more detailed situation has been constructed, one which is consistent with the original given info (hence the CTA rule), and in which it is directly observed that Emma isn’t happy. This demonstrates that Emma’s being happy can’t be deduced from the original given information. Note that such demonstrations can be created and processed by computer programs in general, and hence can be created and processed by our agent A. We will stop at this point on the road toward A. Hopefully the reader now has a sense for how the kind of psychologism we are promoting can be constructive from the standpoint logicist AI. We end with a brief discussion of our promised final point.

306

SELMER BRINGSJORD AND YINGRUI YANG

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

307

6. Mental MetaLogic and Machine Reasoning:Future Work for Our Brand of Psychologism How productive can our version of psychologism be? What grander things might it enable in LAI? Well, our hope is that it can make “Simon’s Dream” a reality. What is Simon’s Dream? It is a dream that Herb Simon, one of the visionary founders of AI (and, for that matter, of LAI, specifically) expressed in the Summer before he died: viz., to build an AI system able to produce conjectures, theorems, and proofs of these theorems as intricate, interesting, subtle, robust, profound, etc.as those produced by human mathematicians and logicians. In this talk Simon pointed out something that everybody in the relevant disciplines readily concedes: machine reasoning, when stacked against professional human reasoning, is laughably primitive. For example, John Pollock, before delivering a recent talk at RPI (Rensselaer Polytechnic Institute) on his system for automated reasoning, OSCAR, conceded that all theorem provers, including his own, are indeed primitive when stacked against expert human reasoners, and he expressed a hunch that one of the reasons is that such provers don’t use diagrammatic, pictorial, or spatial reasoning: they can all be viewed as machine incarnations of mental logic. Simon’s “condemnation” of today’s machine reasoning has also been carefully expressed in specific connection with Gödel’s incompleteness theorems: Bringsjord (1998) has explained why the well-known and highly regarded prover OTTER, despite claims to the contrary, doesn’t really prove Gödel’s incompleteness results in the least. We take this quote from Jacquette very seriously: “Logic can be understood as descriptive of how some reasoning occurs, at the very least the reasoning of certain logicians.” (Jacquette 1997, 324) We specifically submit that mental metalogic, used in conjunction with our Jacquetteinspired variety of psychologism, is AI’s best bet for reaching Simon’s Dream. As we have said, expert reasoners often explicitly work within a system that is purely syntactic (and hence within a system that relates only to mental logic) — but it’s also undeniable that such reasoners often work on the semantic side. Roger Penrose has recently provided us with an interesting example of semantic reasoning: He gives the case of the mathematician who is able to see via images of the sort that are shown in Figures 8 and 9 that adding together successive hexagonal numbers, starting with 1, will always yield a cube. MML (Yang and Bringsjord, in press; Rinella, Bringsjord and Yang 2001), as we’ve said, is a theory of reasoning that draws from the proof theoretic side of symbolic logic, the semantic (and therefore diagrammatic) side, and the content in between: metatheory. Furthermore, while theories of reasoning in psychology and cognitive science have to this point been restricted to elementary reasoning (e.g., the prepositional and predicate calculi), MML includes psychological correlates to all the different forms of

308

SELMER BRINGSJORD AND YINGRUI YANG

deductive reasoning that we find in relevant human experts; for example, modal, temporal, deontic, conditional, reasoning. Accordingly, MML can serve as the basis of an automated reasoning system that realizes Simon’s Dream, and we will attempt to demonstrate this — not now, but in tomorrow’s work arising from the psychologism we have described today. Department of Philosophy, Psychology & Cognitive Science Department of Computer Science Rensselaer Polytechnic Institute, USA

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

309

310

SELMER BRINGSJORD AND YINGRUI YANG NOTES

1 Illusion 1 is from Johnson-Laird & Savary (1995). Variations are presented and discussed in Johnson-Laird (1997a). 2 Even when you make the exclusive disjunction explicit, the results are the same. E.g., you still have an illusion if you use Illusion 1' (1) If there is a king in the hand then there is an ace, or if there isn’t a king in the hand then there is an ace, but not both. (2) There is a king in the hand. Given these premises, what can you infer? 3 Of course, a fully syntactic proof is easy enough to give; indeed we give and discuss it below (see Figure 1). 4 In a recent presentation of Illusion 2 in a lecture by Bringsjord, the reasoning just given was provided by a high school student. Everyone else in the audience declared that Betty is here, “Obviously” — but they provided, at best, exceedingly murky arguments. The same student went on to demonstrate, on the basis of the given information, that Betty isn’t here. His reasoning this time involved indirect proof: The premises are inconsistent, as you may wish to verify. In addition, this clever student declared that he could prove that the moon is made of green cheese, given the premises in question (reductio again, of course). Figure 2 shows a proof in HYPERPROOF, the fourth line of which confirms that a contradiction is a logical consequence of the given info. 5 Another route to reach apparent incoherence is to simply point out that Illusion 1 is ruled an illusion by Johnson-Laird because the reasoning we gave above yields while subjects don’t apprehend this, and the reasoning is based on formal logic, not on what is going on in the minds of subjects. 6 While Johnson-Laird apparently likes to use the colon, if we use a period we lose nothing, and gain the fact that we’re dealing with clearly demarcated sentences. 7 From page 1053 of (Yang & Johnson-Laird, 2000). 8 And the idiosyncratic use of ‘possible’ may stem from what might be called “mental model mindset,” in which subjects are presupposed to be conceiving of possibilities in the mental model sense. Syntactically, a set of formulas is inconsistent if and only if a contradiction can be derived from Semantically, a set Í of first-order formulas is inconsistent just in case there is no interpretation on which all of are true. The syntactic and semantic senses of inconsistency are equivalent in first-order logic. 9 If you’re letting yourself function as a subject in connection with this illusion, don’t read any further in this footnote. The answer is “No.” 10 It’s interesting to note that the account just given, with modifications of the sort we momentarily express in subsequent definitions, apparently captures the general phenomena that have long driven psychology of reasoning. For example, puzzles like the Wason selection task can be precisely represented in the machinery we have introduced and crystallized in successors to Indeed, it would seem that apparently all the puzzles, problems, and tasks at the heart of psychology of reasoning can be formalized along the lines set out in the present chapter. See Stanovich & West (2000) for an excellent catalogue of these puzzles and problems. 11 Such an agent would have lots of practical value, but such value isn’t our concern herein. For example, a great many standardized tests consist in giving humans logic problems; the Graduate Record Exam (GRE) and Law School Admission Test (LSAT) are two examples. An agent able to solve these problems would presumably point the way to agents able to automatically generate the problems in question, which would presumably save many institutions lots of money. 12 For coverage of how this would work, in general, see Russell & Norvig (1994).

PSYCHOLOGISM OF ARTIFICIAL INTELLIGENCE

311

13 We say “in general” because of such finer-grained possibilities as that the subject could correctly represent the premises but incorrectly represent the proposition the query concerns, and so on. In the fourth possibility, Possibility 3 in our table, the cognizer isn’t tricked, so this possibility has no bearing on the definition we’re seeking. 14 For ease of exposition and to work within the space constraints of this chapter, we leave aside certain niceties. For example, to be more precise, we would need to distinguish between representation of one premise versus another, between the premises versus the proposition implicit in Q, etc. 15 There are of course other ways to obtain B. For example, ignoring every fixed prooftheoretic system S, one could produce a “semantic” proof, one based, say, on a truth table. But we are assuming in accordance with that we are restricting the space of proofs in such a way that these kinds of proofs are inadmissible. 16 A fully formal proof in is shown in Figure 4. The predicate letter H is of course used here for ‘helped.’ 17 We have our own, more sophisticated instantiations of MML, but HYPERPROOF is simpler, is widely used in teaching philosophy and computer science, and in the present context is perfectly sufficient.

REFERENCES

Barwise, J. & Etchemendy, J. Hyperproof, CSLI. Stanford, CA, 1994. Barwise, J. & Etchemendy, J. Language, Proof, and Logic. New York: Seven Bridges, 1999. Braine, M. (1998), “Steps Toward a Mental Predicate-Logic”, in M.Braine & D. O’Brien, eds, Mental Logic, 273-331. Mahwah: Lawrence Erlbaum Associates, 1998. Bringsjord, S. “Is Gödelian Model-Based Deductive Reasoning Computational?”, Philosophica, 61, 1998, 51-76. Bringsjord, S. & Ferrucci, D. “Logic and artificial intelligence: Divorced, Still Married, Separated...?”, Minds and Machines, 8, 1998, 273-308. Bringsjord, S. & Ferrucci, D. Artificial Intelligence and Literary Creativity: Inside the Mind of Brutus, a Storytelling Machine. Mahwah: Lawrence Erlbaum, 2000. Jacquette, D. “Psychology the Philosophical Shibboleth”, Philosophy and Rhetoric, 30, 1997, 312-31 (and this volume, 1-17). Johnson-Laird, P.N. Mental Models. Cambridge: Harvard University Press, 1983. Cambridge, MA. Johnson-Laird, P.“Rules and Ilusions: A Criticial Study of Rips’s The Psychology of Proof”, Minds and Machines, 7, 1997a, 387-407. Johnson-Laird, P.N. “An End to the Controversy? A Reply to Rips”, Minds and Machines, 7, 1997b, 425-32. Johnson-Laird, P. & Savary, F. “How to Make the Impossible Seem Probable”, in Proceedings of the 17th Annual Conference of the Cognitive Science Society, 381-84. Hillsdale: Lawrence Erlbaum Associates, 1995. Rinella, K., Bringsjord, S. & Yang, Y. “Efficacious Logic Instruction:

312

SELMER BRINGSJORD AND YINGRUI YANG

People Are Not Irremediably Poor Deductive Reasoners”, in J.D. Moore & K. Stenning, eds, Proceedings of the Twenty-Third Annual Conference of the Cognitive Science Society, 851-56. Mahwah: Lawrence Erlbaum Associates, 2001. Rips, L. The Psychology of Proof, Cambridge: The MIT Press, 1994. Russell, S. & Norvig, P. Artificial Intelligence: A Modern Approach, Saddle River: Prentice Hall, 1994. Stanovich, K.E. & West, R.F. “Individual Differences in Reasoning: Implications for the Rationality Debate”, Behavioral and Brain Sciences, 23, 2000, 645-65. Yang, Y. & Bringsjord, S. “Mental Metalogic: A New Paradigm for Psychology of Reasoning”, in The Proceedings of the Third International Conference on Cognitive Science, 199-204. Hefei: Press of the University of Science and Technology of China, 2001. Yang, Y. & Johnson-Laird, P. N. “How to Eliminate Illusions in Quantified Reasoning”, Memory & Cognition, 28, 2000, 1050-59.

INDEX

aesthetic, 2, 7, 79, 175, 247; see esthetics aggregate, 56 AI, 289, 297, 305, 308; see artificial intelligence algebra, 100, 111, 154 analysis, 29, 41, 49, 54, 57, 62, 69, 82, 87, 90, 100, 101, 108110, 148, 153, 173, 190, 196, 198, 200, 201, 203, 206, 207, 273, 278, 279, 302; analytic, 2, 3, 6-7, 14, 16, 19, 29, 40, 45, 54, 57, 63, 82, 92, 94, 117, 121, 145, 147, 202, 203, 205209, 233-235, 238, 240, 241, 249, 272, 274, 277; analyticsynthetic distinction, 3; analyticity, 3, 57, 76, 205-211, 233, 272, 273, 276, 282; analytic of principles, 63; analytic philosophy, 2, 3, 6, 8, 19, 145, 202 Anderson, Alan Ross, 96, 97,109; Anderson-Scheme, 96 Anderson, Douglas, 177 Anerkennen, 83 antecedent, 33, 37, 43, 109, 1221, 169, 171, 188, 204, 217, 271, 290, 296, 303 anthropology, 34, 163, 243, 246, 248, 258; anthropologism, 85, 116, 119, 120, 214; anthropological-psychological, 85;

A a posteriori, 11, 117, 271 aprioricity, 11,230 apriorism, 206, 207 a priori, 6, 10, 11, 34, 35, 63, 74, 77, 115, 117, 120, 128, 133, 203, 210, 211, 229-231, 233, 234, 236, 237, 239, 242, 243, 246, 247, 257, 268, 271, 272, 275, 277, 278, 284 Abbagnano, Nicola, 177 ability, 7, 12, 24, 34, 37, 54, 60, 62, 74, 98, 99, 132, 222, 264, 267-269, 297; see canpossibility absolute necessity, 94, 246 absolute truth, 121 abstract, 7, 15, 30, 32, 37, 39, 41, 43-46, 52, 60, 61, 67, 68, 71, 73, 77, 86, 87, 126, 164, 171, 173-175, 206, 253, 291; abstracted, 73, 251, 257; abstracting, 61, 72, 73, 251, 255, 257; abstraction, 69, 73, 103, 172, 179, 196, 210, 255; abstractedness, 89 absurdity, 121, 139 act-based theory of meaning, 124, 125; acts and operations, 196 Adams, R.M., 76, 78, 212 Adler, Pierre, 121, 129 Adler, Max, 134-136, 142, 146,

147 313

314

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

anthropological logic, 85, 86, 88 anthropomorphism, 171, 250 anti-descriptivism, 54; anti-descriptivist theory of reference, 52, 64 anti-Leibnizian, 59, 67 anti-objectivistic, 83, 85 anti-realism, 229, 234, 243 antinomy, antinomies, 63, 65, 68, 76 antipsychologism, 2-9, 13, 15-18, 83, 114, 115, 117-119, 122, 142, 143, 149, 195, 197-200, 202, 203, 210; anti-psychologist, 4, 16, 200, 203; anti-psychologistic, 1, 3, 4, 6, 15, 82, 84, 113, 117-119, 124, 131, 141, 145, 159, 196, 247, 249252; anti-psychologists, 1-5, 9, 16, 142 antisociologism, 131, 146 antithesis, 16, 63, 65 Apel, Karel-Otto, 177 apodictic judgment, evidence, 90, 92, 120, 133 apophantic logic, 124 applied logic, 34, 48, 84, 89, 90, 100,101,108, 151, 297 archetypal idealism, 254; see Berkeley; ectypal idealism argument, 5, 9-15, 23, 27, 29, 35, 36, 63, 68, 121, 129, 131, 158, 166, 167, 175, 176, 198-201, 203, 209-211, 216, 229, 231, 238-241, 243, 244, 246, 263265, 267, 268, 272, 273, 275, 276, 282, 283, 299-303 arguments, 2, 5, 7-8, 15, 26-27, 31, 33, 37, 48, 55, 65-66, 99, 103-104, 106, 108, 112-113,

115-117, 127, 138, 141, 171, 173, 176, 178, 189, 207, 209, 232-233, 238, 258, 266-267, 273 Aristotle, 23, 24, 26, 28, 32, 44, 45, 49, 76, 153; Aristotelian, 28, 88, 107, 108 arithmetic, 18, 41, 90, 120, 129, 195, 210, 230, 231; arithmetic relations, 230; Arithmetik, 15, 18, 109, 110, 114, 129, 148, 149 Arnauld, Antoine, 22, 23, 25–28, 30, 44, 47

Arndt, H.W., 31, 45, 49 ars inveniendi, 27, 33, 45; ars iudicandi, 27, 33 artificial, 22, 25, 26, 34, 219, 260, 289, 297, 292, 302, 312, 313; artificial intelligence, 260, 289, 312, 313; see AI ascription, 278, 284 Ashworth, Jennifer, 45, 47 assertibilism, 60, 62, 63, 65–68, 77; assertibility theory, 52, 63, 65; assertoric, 90–92, 234; assertoric discourse, 234 association, 24, 27, 41, 43, 70–72, 142, 193, 219, 220, 251, 259; associationist, 27, 161 assumption, 2, 5, 12, 13, 65, 66, 77, 97, 120, 139, 189, 192, 263, 269, 275, 279, 282, 291, 297 attention, 5, 18, 21–23, 33, 34, 45, 70, 74, 102, 125, 159, 166, 167, 173, 177, 211, 233, 264, 290, 303 Ayer, A.J., 145, 234

INDEX

B Bacon, Francis, 21, 28, 29, 44, 45, 47,110,149 bad psychologism, 250–253, 256–

258 Baier, Annette, 259 Bain, Alexander, 75, 79, 157, 174 Baker, G.P., 5, 6, 18, 204, 211 Baldwin, James Mark, 160 Barrett, R., 285 Barth, H., 18, 47, 138, 147–149 Barwise, Jon, 295, 305, 312 Baumgarten, Alexander Gottlieb, 45

Bayle, Pierre, 32 Bazhanov, Valentine A., 108, 109 Bealer, George, 216, 217, 221, 226, 227 Beck, Lewis White, 78, 154, 246, 259, 260 behavior, 13, 14, 94, 192, 205, 253, 263, 265, 266, 268, 269, 271, 273–276, 278–283, 285; behavioral, 266, 268, 270, 272–274, 276, 281, 283, 313; behavioral criteria, 272, 274, 276, 281 behaviorism, 7, 11, 245, 263–269, 273, 276, 278–281, 283, 286, 287; behavioristic, 206, 263, 272, 273 belief, 3, 15, 69, 70, 78, 126, 134, 157, 158, 169, 174, 182, 190, 215, 221–224, 229, 233, 237, 238, 240, 241, 256, 289, 296, 299, 301, 303; believing, 83, 207, 209, 216, 222, 223, 226, 278, 299 Belnap, Nuel, 109

315

Benacerraf, Paul, 217, 219, 220, 226, 227; Benacerraf’s dilemma, 217, 219, 220 Beneke, Friedrich Eduard, 37, 38, 43, 47, 85–87, 109, 147, 160, 161 Berg, Ian, 46, 47 Berger, P.L., 141, 147 Berkeley, George, 6, 18, 47, 129, 130, 174, 213, 253–255, 259, 261; Berkeleyan, 76, 232 biology, 10, 12, 136, 185, 193, 297; biological, 135, 136, 143, 185, 276; biological and psychological mechanisms and processes, 185; biochemical, 284 bivalence, bivalent logic, 197, 201–203, 207, 209, 211 Bloor, David, 146, 147, 211 Bochen,’ ski, Innocent Maria (Josef), 110, 148 Bohnert, Herbert G., 206, 211, 213 Bolzano, Bernard, 7, 17, 18, 21, 26, 33, 36, 39–49, 83, 108–110, 148, 149, 260 Boole, George, 17–19, 101, 103, 107, 109, 148, 250–252, 256, 257, 259–261; Boolean, 13 Boolos, George, 226, 227 Born, Friedrich Gottlob, 35, 47, 148 Bosanquet, Bernard, 48, 151 Bowie, Andrew, 247, 259 Bradie, Michael, 181, 184, 185, 193 brain, 18, 177, 278, 284, 313 Braine, M., 296, 304, 312 Brent, Joseph, 155, 177

316

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

Brentano, Franz, 1, 3, 18, 122, 128, 129, 249 Bringsjord, Selmer, 290, 291, 297, 302, 308, 311–313 Brockhaus, R., 148 Brouwer, L.E..J., 63, 78 Brown, Thomas, 161 Buchler, Justus, 177 C Caesar, Julius, 56, 57, 59, 76, 166 calculus, 17, 100, 105, 107, 109, 148, 251 Calker, Johann Friedrich, 33, 47, 148 Campbell, Donald T., 48, 151, 184–186, 189, 193 can-possibility, 98; capability, 99; see ability Carl, Wolfgang, 46, 47, 79, 110, 148, 149, 178 Carnap, Rudolf, 18, 48, 145, 181, 182, 187, 203–213, 232–234, 272, 273, 282; Carnapian, 35, 197 Carr, David, 212, 247, 259 Cartesian, 6, 182, 183, 188, 189, 192, 212, 245; Cartesian dream, 183, 188, 189, 192; Cartesian dualism, 247; Cartesianism, 6; see Descartes Cassirer, Ernst, 7, 129, 148, 196, 260 categorical imperative, 248, 249, 260 category, 63, 102, 103, 115, 278; categories, 3, 8, 10, 63, 64, 74, 102, 104, 113, 114, 117, 135, 160, 161, 176, 200, 230, 269, 278; category of “quantity”, 63; categorial framework, 74

Caton, Hiran, 256, 259 cause, 9, 24, 40, 63, 78, 94–96, 135, 185, 207, 208, 230, 231, 233; causal explanation, 279, 281; causal influence, 120; causal order of nature, 95; causal relations, 93, 115, 243; causality, 32, 62, 72, 77, 82, 93, 117; causation, 24, 120, 255, 265, 279, 284 Cerebus, 171 certainty, 17, 29, 65, 66, 91, 92, 123, 161, 175, 186–188, 196, 208 chemistry, 10, 14 Chisholm, Roderick M., 128, 129, 145 Chomsky, Noam, 264–266, 273, 274, 276, 280, 281, 285, 286 Chrysippus, 254 Church, Alonzo, 10, 17, 18 Churchland, Patricia Smith, 6, 17,

18 Cicero, Marcus Tullius, 45, 254, 259; Ciceronian, 27 class, 54, 55, 58, 61, 93, 99, 103, 106, 107, 137, 139, 199, 273, 274; classes, 69, 104, 106, 107, 227, 269; see sets class struggle, 139 classical logic, 16, 65, 66, 68, 77, 82, 101, 107, 108, 202; see logic; traditional logic Coffa, Alberto, 48, 203, 210, 212 cognition, 32, 33, 40, 55, 59, 62, 66–68, 73, 77, 89, 94, 114, 117, 120, 123, 126, 127, 129, 175, 196, 198, 305; cognitionbased semantics, 68; cognitive, 6–8,10,11,51,52,55,59–65,

INDEX

69, 72, 75, 76, 121, 131, 141, 164, 184, 185, 195, 198, 199, 207, 212, 245, 255, 260, 277, 289–291, 293, 294, 308, 309, 312, 313; cognitive illusions, 289, 293; cognitive science, 7, 10, 121, 131, 212, 245, 308, 309, 312, 313; cognitive semantics, 64; cognitive state, 55, 75; cognitive structures, 184; cognitivism, 52, 234 Cohen, H., 4, 19, 118, 132, 148, 260 coinstantiation, 218, 219, 222 Colapietro, Vincent, 155, 177 colloquial language, 90 communication, 281, 291, 296 complex predicate, 58, 227 Comte, Auguste, 135 conceive, 22, 39, 53, 132, 172, 174, 176, 230, 303; conceiving, 22, 53, 70, 310 concept, 1, 6, 11, 34, 36, 39, 48, 49, 53–61, 64, 67, 68, 70–74, 76, 77, 101–107, 109, 113, 115, 116, 118, 121, 123, 125, 138, 191, 214, 233, 243, 249, 250, 252–255, 283; concepts, 11, 31, 32, 36–38, 42, 44, 53–57, 59, 60, 62, 64, 65, 67, 70–74, 76, 77, 86, 100–107, 119, 123, 124, 132, 140, 144, 161, 169, 191, 229, 245–247, 249, 257, 265, 267, 268, 274, 275, 278, 280, 284; conceptual, 54, 56, 58, 62, 63, 90, 93, 127, 182, 187, 193, 194, 200, 206, 207, 213, 230, 235, 246, 264, 265, 267, 269, 276; conceptualism, 7, 52–56, 58, 59;

317

conceptualist, 53–57, 67, 75; conceptualization, 267 Condillac, Étienne Bonnot de, 33, 43,45 conditional normative psychologism, 83, 85, 88 conditional, 16, 69, 71, 83, 85, 88, 92, 93, 184, 290, 296, 303, 309; conditional proof, 296 conscious, 15, 71, 120, 161, 168, 200, 217; consciousness, 21, 35, 42, 44, 87, 92, 114, 115, 120, 123, 127, 169–171, 177, 179, 200, 232, 248, 253, 261; consciousnesses, 123 consequent, 2, 37, 76, 109, 290 consistency, 36, 55, 66, 69, 170, 173, 271, 282, 293; constructive dilemma, 295, 296, 301 constructive, constructivism, 7, 211, 295, 296, 301, 305 content, 6–8, 33, 34, 37, 58, 81, 83, 89–93, 96, 100–103, 116, 118, 120, 137, 138, 141–143, 195, 196, 198, 233, 234, 255, 283, 308; content and extension, 102; content-logical foundation, 102; content-logical ideas, 102 context, 34, 83, 93, 98, 106, 109, 137, 144, 159, 163, 167, 168, 170, 175, 210, 236–238, 250, 254, 260, 269, 270, 278, 279, 281, 284, 289, 312; contexts of concept acquisition, 11 continental philosophy, 249 contingency, 70, 73–76, 200, 205 contradiction, 35, 42, 62, 70, 76, 82, 88, 90, 120, 121, 249, 259, 292, 303, 311; contradiction in

318

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

will, 62; contradictory, 35, 36, 42, 54, 76, 77, 106, 120, 145, 201–203; contradictory opposites, 76, 77 convention, conventional, 222, 264,271,272,282,286 Copp, David, 227 Cranston, Maurice William, 45, 47 criterion of truth, 240, 256 critical idealism, 229, 247, 253; critical philosophy, 34, 76, 77, 79, 157, 232; critical undertaking, 160; critical-teleological, 2 Crusoe, Robinson, 146 crypto-Platonist, 146 culture, cultural, 25, 27, 30, 44, 47–49, 135, 141, 148, 151, 153, 230, 234, 235, 237–239; cultural context, 238; cultural relativism, 229, 235, 238–243 Curtius, E. Robert, 138, 140, 141, 148 Cussins, Adrian, 114, 129 D d’Alembert, Jean le Rond, 25 Darwin, Charles, 184, 185, 192, 193; Darwinian biology, 185 Davidson, Donald, 186, 193, 194, 211, 212, 231, 243, 274, 281, 283, 285, 286 de cogitatione, 197, 210; de dicto, 197; de facto, 160, 163; de jure, 160, 163;de re, 197, 210, 222 de Staël, Madame, 36 DeBoer, Thomas, 114, 129 Decock, L., 193, 194, 285

deductive inference, 71, 78, 257; deductive logic, 70, 175, 208; deductive reasoning, 109, 148, 303, 308, 312 deep causal explanation of mental states, 281 Delaney, C.F., 177 deliberation, 72, 169 DeMorgan’s Laws, 290 Denken, 44, 47, 83, 84, 86, 110, 122, 151; Denknotwendigkeit, 83 Dennett, Daniel C. 165, 171 deontic logic, 13, 109; see modal logic Derrida, Jacques, 210, 212 Descartes, René, 28, 45, 47, 48, 160, 187, 256, 259, 262 descriptivism, 54; descriptive epistemologies, 185, 187; descriptive psychology, 123, 128, 129 determinism, 98, 144, 231; determination, 76, 77, 98–100, 103, 104, 107, 256; determinand, 103, 104; determinator, 103, 104 Dewey, John, 62, 78, 186, 193, 209, 213, 366 diagram, 105, 293, 295, 299, 301, 304, 305, 307, 310; diagrammatic, 308 dialectic, 7, 16, 63; dialectical opposition, 76 Diderot, Denis, 25, 26, 45, 47 Dilthey, Wilhelm, 134; Diltheyan, 197 Ding an sich, 115 disposition, 76, 99, 160, 241, 269, 270, 276; dispositional, 7, 99, 270; dispositional predicates, 99

INDEX

Doore, Gary, 249, 259 Doppelgänger, 276 Dorschel, Andreas, 247, 259 doxastic assumption, 279 Dreher, John H., 227, 229, 259 Dreyfus, Hubert L., 201, 210, 212 dualism, 121, 134, 137, 144, 146, 182, 241, 242, 245; see Descartes; philosophy of mind Duhem, Pierre-Maurice-Marie, 182, 235 Dummett, Michael, 5, 6, 18, 52, 63, 65, 78, 114, 124, 125, 129, 145, 197, 199, 201–203, 205, 207, 209, 213 Dunkmann, K., 148 Duran, Jane, 256, 259 Durkheim, Emile, 134 E Earl of Clarendon, 30 economism, 135 ectypal idealism, 254; see Berkeley; archetypal idealism Eddington, Arthur, 231 Einstein, Albert, 231 Eisler, Rudolf, 4, 18 eliminative psychologism, 43 Elias, N., 148 Ellis, Brian, 5 Elsenhans, Theodore, 132, 148 emotivism, 7 empirical evidence, 127, 238, 242, 243, 268, 283; empirical psychology, 2, 6, 31, 34, 41, 72, 86, 87, 100, 183, 229, 230, 234, 246, 279, 289; empirical science, 11, 34, 65, 70, 86, 101, 119, 182, 187, 230, 232, 234, 239, 240, 246, 264, 271, 272, 277; empiricism, 123, 177, 182, 183, 190,

319

191, 197, 198, 206–209, 213, 233, 254, 263–265, 268, 277, 279, 280, 285, 286; empiricist, 2, 6, 86, 114, 121, 127, 160, 161, 182, 206, 230, 233, 254, 255, 265, 280 empty classes, 106, 107 engineering, 12, 184, 190, 297 Enlightenment, the, 22, 26, 237 enthymeme, entheymematic reasoning, 48, 55 entity, entities, 5, 14, 15, 21, 22, 27, 57, 59, 76, 99, 115–117, 119–121, 124, 126, 132, 133, 187, 215, 221, 223, 226, 241, 253, 266, 269 environment, 158, 184, 185 epistemology, 1, 5–8, 19, 29, 38, 39, 42, 45, 71, 75, 108, 117, 118, 130, 136, 137, 139–143, 145–147, 153, 181–194, 197, 198, 203–205, 208, 213, 245, 250, 256–258, 260, 261, 263, 264, 269, 273–275, 278–280, 285–287; epistemologists, 45, 183, 187–189; epistemological, 2, 8, 27, 41, 45, 70, 72, 73, 114, 115–117, 118, 127, 128, 139, 140, 142, 144, 181, 183, 188, 190, 195, 209, 250, 257; epistemological idealism, 114; epistemological theory of experience, 127; epistemology naturalized, 6, 19, 153, 187, 204, 205, 213, 256, 257, 261, 286, 287; epistemic, 8, 17, 55, 64–66, 73, 77, 83, 85, 87, 88, 90, 91, 107, 158, 186, 190, 193, 196–198, 204, 207, 256, 257; epistemic pessimism, 65, 77;

320

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

epistemic task, 55; epistemicpragmatic, 85 epoché, 128; see Husserl Eppstein, Paul, 144, 148 Erdmann, Benno, 43, 47, 85, 87, 88, 90, 91, 108, 109, 119, 122, 129, 131, 132,139, 141, 148 Erfassen, 83 Erlebnisse, 197 Esposito, Joseph, 177 Esser, Wilhelm, 32, 43 esthetics, 158, 164, 173 Etchemendy, John, 295, 305, 312 ethics, ethical, 2, 62, 72, 88, 89, 141, 158, 160, 163, 164, 169, 171, 173, 229, 233–236, 238, 242, 243, 245, 246, 248–250, 254, 257, 258; ethics of thinking, 88, 89 Euclidean geometry, 230, 231 evidence, 27, 62–64, 69, 73, 117, 120, 123, 124, 126, 127, 132, 133, 135, 173, 182, 183, 190, 199, 208, 217, 237, 238, 242, 243, 255, 263, 265, 266, 268–270, 273, 275–279, 283, 284; evident, 15, 29, 70, 120, 174,246,275, 278 evolution, 158, 183, 194; evolutionary epistemology, 184, 185, 193; evolutionary theory, 182; see Darwin; Darwinian biology excluded middle, 57, 58, 65, 66, 77, 82, 197, 199, 202, 210 exclusive disjunction, 290, 291, 311 experience, 10, 11, 17, 30, 31, 36, 59, 63, 66, 71, 72, 74, 75, 86, 118, 123, 126–128, 132, 133, 135, 142, 158, 163, 169, 171, 174, 175, 179, 184, 195, 207,

230, 233, 235, 236, 239, 240, 245–248, 250, 251, 254, 255, 257, 266–268, 272, 273, 290, 299; experiences, 15, 61, 66, 71–74, 116, 120, 175, 184, 233, 235, 250, 251, 254, 257; experiential, 158, 174, 255; experiential element in logic, 158 experiment, 11, 13, 176, 177, 231, 241, 242, 267; experimental method, 10, 254; experimental psychology, 82, 134, 143, 158, 160, 170 explanation, 2, 7, 10, 74, 94, 95, 110, 143, 149, 192, 194, 197, 233, 263–267, 269, 271–273, 275, 277, 279, 281; explanatory, 14, 132, 263–265, 267, 269–275, 277, 278, 281, 283–285; explanatory efficacy, 277, 283; explanatory theory, 275; explanatory utility, 264, 272, 278 extensional, 13, 101, 106, 197, 199, 201, 293; extensionalism, 195; extensionalized, 197 external modalities, 93; external necessity, 93 external world, 181, 182, 187, 188, 232, 254, 258, 261; externalizing, 263 F faculties, 28, 36, 41, 67, 73, 77, 117, 124, 248, 254; faculty of reason, 72; faculty psychology, 72 fallacy, fallacious, 12, 26, 190, 227, 250, 251, 297 falsehood, 12, 46, 226, 232, 297 Feigl, Herbert, 3, 18 Ferrier, James, 45 Fisch, Max H., 163, 173, 175, 178

INDEX

Fitch, Greg, 227, 297 Fleck, Ludwik, 145, 146, 148 Fodor, Jerry A. 6, 17, 18 folk psychology, 158, 276, 281 formal logic, 78, 81, 82, 85, 86, 89, 107, 108, 311; formalization, 101, 251; formalized languages, 90, 213 Foucault, Michel, 211–212 foundation for arithmetic, 90 foundation of logic, 70, 85, 86, 121 foundational psychologism, 87–88 Fraser, Alexander Campbell, 44, 45, 48, 151 Frederick William (the Great), 31 freedom, free will, 76, 98, 282 Frege, Gottlob, 3–8, 15–18, 37, 41, 43, 46, 47, 81, 83–85, 87–93, 99, 108–110, 113–117, 119, 121, 122, 124–126, 129–132, 141, 145, 148, 149, 195–201, 203–212, 214, 215, 232, 249, 251, 258 Friedman, Michael, 285 Fries, Jakob, 43, 45, 47, 85, 86, 110, 113,149,161 Frisby, D., 149 Frischeisen-Köhler, M., 136, 149 full-blown cultural relativism, 238– 242 Føllesdal, Dagfinn, 117, 129, 130, 276, 283, 285 G Gadamer, Hans-Georg, 200, 212 Gallie, W.B., 178 gambler’s fallacy, 189 Gavagai, 270; see Quine; indeterminacy of (radical) translation

321

Geach, P.T., 18, 145, 234 Gedanken, 7, 43, 49,122,153 Geltung, 83 generality, 17, 55, 91, 97, 108, 248, 293; general concepts, 54, 56, 57; general empirical conception, 60, 61; general logic, 34, 35, 58, 101 genesis, genetic, 38, 81, 83, 86, 87, 137, 143, 144, 183, 184, 250, 269, 283; genetic-lively (genetisch-lebendig) logic, 86, 87 geometry, 13, 41, 230, 231; geometrical laws, 125 George, Alexander, 266, 285 George, Rolf, 18, 46–48, 83 Gesetzlichkeit, 118 Gibson, Roger, 182, 193, 285 God, 23–25, 52, 59, 64, 66, 77, 93, 179, 189, 233, 247, 253, 254, 256, 259; godhead, 254; gods, 247, 259 Goethe, Johann Wolfgang von, 26 Goldfarb, Warren, 277, 278, 282– 285 Goldman, Alvin I., 204, 212 good psychologism, 250, 252, 253, 255–258 Goodman, Nelson, 145 Gosselin, Mia, 280, 285 Goudge, Thomas, 178 Gödel, Kurt, 10, 17, 19, 308, 312 grammar, 124, 162, 163, 277, 280 grasping, 57, 73, 83, 86, 87, 89, 102, 125, 126, 195; grasping of truth, 83 Grünwald, Ernst, 139–141, 145, 149 H Haack, Susan, 17, 19, 178, 178, 286

322

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

Haaparanta, Leila, 108, 110, 129, 130, 149 Hacker, P.M.S., 2, 5, 6, 18, 19, 204, 211 haecceity, 57, 58, 64, 66 Hager, Achim, 253, 259 Hahn, Lewis Edward, 184, 191, 211–213 Hamilton, William, 51–59, 69, 70, 73, 75, 76, 78, 79, 161, 164; Hamiltonian, 57, 69, 70 Hanna, R., 115, 116, 126, 129 happiness, 248, 305 Hare, R.M., 234 Harman, Gilbert, 186, 193, 286 Harrison, Jonathan, 249, 259 Hartley, David, 161 Hausman, Carl, 178 Hegel, G.W.F., 26, 43, 48, 113, 122, 129, 153, 253, 255, 260, 261; Hegelian, 113, 114, 160 Heidegger, Martin, 144, 149, 249 Heil, John, 253, 260 Heim, K., 133, 149 Heimsoeth, Hans, 28, 45, 48 Heisenberg, Werner, 231 Hempel, Carl G., 94, 108, 110, 149 Hendel, Charles W., 44 Herbart, Johann Friedrich, 36–38, 43, 48, 82–86, 89, 108, 110, 149; Herbartian, 87 heuristic, 15, 33, 137, 143, 159, 176 Heymans, G., 132, 133, 149 historical, 2, 4, 9, 33, 69, 127, 131, 136, 137, 140, 143, 147, 161, 173, 200, 249; historicism, 146, 154; historism, 135, 213 history of logic, 33, 44, 82 Hobbes, Thomas, 23 holism, 182

Hookway, Christopher, 178, 187, 189–191, 193 Horkheimer, M., 138, 139, 144, 149 Hubbert, Joachim, 247, 255, 260 Hull, David, 186, 193 humanism, 52, 60, 67, 69; humanize, 59, 66, 68, 75; humanized, 51, 52, 62, 66, 68 Hume, David, 6, 24–27, 31–33, 44, 45, 48, 76, 187, 189, 239, 240, 254, 255, 258–262; Humean, 25, 26, 189; Humean predicament, 189 Husserl, Edmund, 3–7, 10, 16, 21, 41, 43, 48, 82, 90, 101, 108, 110–134, 139, 141, 143–145, 149–151, 156, 161, 195–197, 200, 201, 205, 207, 209–212, 232, 249, 261; Husserlian, 130, 144, 196, 199, 201, 204, 207–210 HYPERPROOF, 305, 311, 312 hypostatic, 172, 178 hypotheses, 10, 11, 182–184, 190, 191, 208, 265, 267, 268, 281; hypothetical judgment, 94, 97, 100, 109; hypothetico-deductive method, 181, 182, 191; hypothetical necessity, 95 I idealism, 5, 14, 15, 64, 113, 114, 150, 210, 229, 230, 232, 234, 247, 254, 255; ideal meanings, 118, 119, 123, 124, 128 ideas of reason, 60, 62 identity, 42, 57, 58, 88, 100, 105–107, 163, 209, 216, 226, 227, 246, 252; identification, 116, 158, 161, 164, 226, 250,

INDEX

268; identity calculus, 100; identity conditions, 216, 226, 227; identity of indiscernibles, 57, 209; see sameness ideology, ideological, 1, 22, 44, 81, 108, 136–139, 141, 154, 247 ignorance, 98, 99 illative, 166, 172, 177 illusion, 25, 77, 220, 290–294, 296, 298–303, 305, 307, 311; illusions, 35, 189, 289–293, 296–299, 302, 313 imagination, 41, 53, 60–63, 70, 72, 77, 79, 83, 84, 87, 167, 254; imaginative, 61, 62, 72, 237 imperative, 12, 131, 160, 175, 248, 249, 260 implication, 15, 96, 279, 295, 296, 299; implications, 5, 6, 15, 159, 163, 165, 171, 181, 233, 249, 255, 313 inclination, 24, 84, 248, 250, 284 incoherence, 119, 121, 292, 311 indeterminacy of (radical) translation, 204, 264–266, 272, 273, 275, 276, 280–282, 285; indetermination, 98 individual concept, 56–59, 76; individual substance, 56, 59 induction, inductive, 17, 28, 36, 69, 70, 78, 120, 127, 165, 184, 187, 190, 208, 209, 267; inductive generalization, 70, 127; inductive inference, 69; inductive logic, 208, 209; inductive reasoning, 78 inference, 51, 53–55, 59, 71, 75, 78, 118, 159, 160, 162, 164, 167–169, 172, 176, 177, 203, 248, 251, 252, 254, 257, 290, 292, 294, 296, 298, 301; infer-

323

ences, 12, 27, 36, 38, 132, 165, 168, 203, 232, 291, 297, 304; inferential, 22, 34, 35, 43, 71, 159, 163, 167, 293, 298, 300, 301; inference rules, 294, 296, 297 infima, infimum species, 56, 60 infinite, infinity, 59, 63, 68, 76, 77, 253, 255; infinite totality, 68; infinitely complex individual concept, 78 information, 11, 45, 54, 57, 58, 61, 67, 77, 174, 182, 183, 190, 192, 276, 280, 283, 289, 305, 311; information sciences, 11 innate, 184, 185, 190, 265, 268, 269, 274, 276, 280, 281; innateness, 268, 269, 281, 285 instrumental normativity, 189 intellect, 55, 67, 68, 73, 77, 84, 165, 170; intellectual, 2, 7, 49, 66–68, 73, 76, 77, 141, 155, 165, 174, 212, 251, 257, 265; intellectual intuition, 78; intelligible world, 66, 67, 78, 278 intensional, intensionality, 13, 57, 58, 62, 216; intensional logics, 13; intensional theory, 58; intensional truth theory, 62 intention, intentionality, 75, 89, 117, 122, 135, 142, 158, 169, 172, 197, 209, 212, 220–222, 225, 249, 258, 260; intentional, 125, 142, 215, 220–222, 224, 225, 253, 265, 269, 270, 274, 277, 283, 281; intentionalism, 7 interchangeability salva veritate, 276 internal and external modalities, 93; internal necessity, 93; internal realism, 75, 208 intersubjectivity, 141, 279

324

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

introspection, 70, 164, 228, 254, 265, 279, 280; introspective, 72, 73, 227, 228, 266 intuition, 7, 16, 55, 59, 63–68, 74, 76–78, 106, 118, 128–130, 195, 200, 208, 215, 228, 245, 269, 279, 280; intuitionism, 7, 65, 195; intuitionistic, 63, 65, 78, 200; intuitionistic logic, 65, 78; intuitionistic mathematics, 63 invalid, invalidity, 12, 65, 70, 74, 160, 290, 295–298, 300 J Jacobi, Maria Charlotta, 45, 46 Jacquette, Dale, 1, 227, 245, 252– 254, 258, 260, 289, 297, 308, 312 Jakob, Ludwig Heinrich, 35, 46, 48 James, William, 174, 175, 187 Jansohn, Heinz, 255, 260 Jäsche, Gottlob Benjamin 32, 33, 46, 78, 130 Jerius, Holger, 260 Jerusalem, Wilhelm, 5, 132, 134, 142, 143, 146, 150 Johnson-Laird, P.N., 289–292, 294– 296, 301, 304, 311–313 Jonas, H., 139, 151 Jubien, Michael 215, 226, 227, 259 judge, 22, 24, 44, 54, 92, 131, 160, 162; judgement, 133, 137, 141, 143, 144, 226; judging, 22, 27, 44, 53, 54, 67, 70, 83, 93, 118, 120, 125, 127, 226; judgment, 24, 36, 39, 44, 51, 53–58, 60, 62–64, 69, 71, 74–79, 82, 83, 88, 90–94, 96, 98, 99, 101, 105, 109, 117, 118, 120, 123, 158, 165, 171, 175, 176, 242, 269, 299; judgments, 10, 27, 36, 40, 44,

54–58, 60, 62, 64–67, 71, 76, 83, 85, 87, 89–95, 100, 103, 109, 111, 117, 163, 243, 257 justification, 3, 7, 10, 11, 15, 63, 70–73, 108, 118, 142, 160, 185, 187, 190, 191, 272, 279, 280, 299; justify, 63, 66, 70, 164, 165, 185, 237, 247, 272, 278, 279, 284 K Kant, Immanuel, 3, 4, 29–35, 38, 43, 46–49, 52, 53, 59–68, 71–79, 85, 117, 122, 124, 127, 129, 130, 136, 152, 159, 160, 162, 175, 196, 212, 230–232, 234, 245–250, 254, 255, 258–261; Kantian, 3, 33, 35, 36, 40, 69, 83, 85, 91, 92, 108, 114, 115, 117, 118, 162, 205, 230–232, 234, 236, 239, 247, 262, 278, 284; Kantian idealism, 230, 234; Kantians, 33, 34, 36, 132, 134, 141, 144, 229; see Neo-Kantians Katz, J., 145 Kauppi, 76, 78 Kent, Beverley, 178 Kettler, D., 151, 152 Kiesewetter, J.G.C.C., 33–35, 37, 42,46,48,151 Kim, Jaegwon, 186, 194 Kirkland, Frank, 113, 130 Kitcher, P., 102, 117, 130, 145 Kneale, Martha, 82, 110, 151 Kneale, William, 82, 110, 151 knowledge, 1, 6, 8, 12, 18, 19, 23, 26, 30, 31, 33, 38–41, 45, 48, 54, 55, 59, 60, 63, 64, 67, 72, 74–77, 79, 89, 91, 94, 100, 101, 114, 117, 118, 123, 127, 131–149, 151, 152, 154, 160–162, 169,

INDEX

174, 176, 178, 182–190, 192–197, 206–208, 211, 213, 221, 229, 230, 232–235, 238, 245–249, 254, 256, 257, 260, 261, 265, 266, 268, 281, 285, 294, 297, 299 Koppelberg, Dirk, 188, 194, 279, 285 Kornblith, Hilary, 153, 186, 194, 256, 261 Kripke, Saul A., 216, 281, 282, 285 Kuhn, Thomas S., 145–147, 151, 200, 212 Kusch, Martin, 1, 2, 4, 5, 7, 17, 19, 108, 110, 128, 130, 134, 141, 145, 151, 211, 213, 249, 261 L

LAI, 297–299, 308; see logicist artificial intelligence Lakatos, Imre, 19, 146, 147, 151, 153, 154, 164, 178, 179

language, 1, 3, 7–9, 13, 17, 52, 73, 82, 90, 93, 108, 125, 168, 176, 177, 181, 187, 194, 197, 198, 200, 202, 205, 208, 212, 215, 224, 225, 233, 234, 238, 251, 252, 263–268, 270–272, 274, 276, 279–281, 283, 285–287, 292, 295, 300, 312; language learning, 263, 266, 268, 276, 285; language of science, 90, 205 Lapp, A., 133,151 law, 18, 35, 42, 43, 58, 62, 65, 72, 77, 94, 100, 105–109, 111, 118, 120, 123, 177, 199, 210, 246, 248, 249, 251, 311; law of contradiction, 35, 42, 43; law of double negation, 105, 106; law of identity, 42; laws, 4, 10, 12, 17, 18, 22, 27, 32–35, 37, 42, 55, 58,

325

67, 70–72, 74, 76, 82–91, 101, 108, 110, 118, 120, 121, 123–125, 128, 129, 132, 133, 148, 161, 163, 164, 173, 176, 195, 198–201, 204–211, 246, 248, 251, 252, 257, 259, 290, 291; laws of association, 27, 70; laws of identity, 43; laws of logic, 34, 35, 55, 58, 74, 84, 101, 120, 128, 133, 173, 199, 257, 291; laws of thinking, 89, 246, 248; laws of thought, 4, 18, 22, 70, 110, 132, 148, 195, 199–201, 204–211, 251, 252; law of the excluded middle, 77; lawlike, 12 Leibniz, G.W., 21, 24, 45, 46, 48, 52, 56–59, 62–64, 74–79, 103, 151, 210, 213; Leibnizian, 52, 57–59, 61–64, 66, 67, 76, 77, 206, 210 Leonardi, Paolo, 193, 194, 285, 286 Levinas, Emmanuel, 115, 117, 130 Lewalter, E., 140, 151 Lewis, David, 226, 227, 286 Lévy-Bruhl, Lucien, 134 linguistic, 2, 75, 76, 121, 122, 125, 141, 160, 200, 205–207, 216, 224, 225, 263–265, 267, 270–272, 277, 279, 280, 282, 284, 286; linguistic behavior, 263; linguistic critique, 121; linguistic theory, 265, 286; linguistic-pragmatic, 121 Lipps, Theodore, 5, 85, 88, 89, 108, 110, 131, 132, 151 Livingston, Donald, 259, 261, 262 Lobachevskian geometry, 13 Locke, John, 6, 23–33, 43–45, 47, 48, 76, 151, 176; Lockean, 25 Loemker, L.E., 79

326

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

logic, 1, 2, 4–19, 21–49, 51–53, 55–60, 65–71, 74–76, 78, 79, 81–93, 95, 96, 100–103, 106–111, 113–125, 128–130, 132, 133, 142, 145, 148, 149, 151–154, 157–176, 178, 181, 184, 196–210, 214, 223, 232–234, 243, 245, 246, 248–253, 256–259, 261, 262, 271, 272, 289–293, 295, 297, 301–304, 306, 308, 311–313; logic of inquiry, 158; logic of modalities, 82, 91; logicassemeiotic, 151, 164; logicbased, 289, 297 logica docens, 22 logica utens, 22 logical, 2, 3, 5–8, 11, 13–16, 18, 19, 21, 22, 24, 27, 28, 30, 31, 34–38, 40–43, 45, 49, 51–53, 57, 58, 65, 66, 68, 70, 71, 75, 77, 81–94, 100–105, 107, 108, 114–116, 118, 120–130, 132, 133, 142, 144, 157, 160–172, 174–176, 192, 196–199, 204–210, 212, 213, 216, 218, 219, 222, 231, 243, 247–252, 267, 270, 274, 282, 286, 290–292, 294–299, 304, 305, 311; logical and mathematical cognitions, 114; logical anthropologism, 85; logical character, 167; logical cognition, 123, 126, 129; logical consequence, 37, 42, 49, 65, 208, 311; logical entities, 14, 15, 133; logical form, 40, 90, 167, 168, 172, 243, 252, 267; logical glue, 218, 219, 222; logical grammar, 124; logical illusion, 290, 291,

294, 296–300; logical laws, 10, 83–90, 101, 108, 120, 121, 123, 124, 133, 161, 251; logical necessity, 70, 93–98, 123, 204, 210; logical or mathematical necessity, 210; logical principles, 57, 58, 88, 144, 249, 250; logical proposition, 115, 116, 126; logical psychologism, 88, 89, 91, 115–117, 121, 124, 126, 128, 130; logical relation, 37, 43, 58, 101, 166, 172, 208, 219; logical theory, 10, 15, 35, 52, 53, 75, 115, 117, 127, 217, 220, 266; logical truth, 8, 11, 16, 65, 125–127, 132, 142, 169, 206–210, 212, 213; logical truths, 11, 125–127, 142, 169, 209, 210; logical-consequence, 37, 42, 49, 65, 208, 311; logicomathematical, 256, 271 logician, 5, 24, 32, 33, 36, 39, 82, 86, 88, 101, 107, 119, 173, 208, 250–252, 290, 291, 304, 308 logicism, logicist, 90, 123, 289, 297, 305; logicist AI (LAI), 297–299, 308; see artificial intelligence Lord Verulam, 45 Lotze, Rudolf Hermann, 40, 48, 82, 83, 110, 111, 122, 151, 152 Luckmann, T., 142, 148 M MacColl, Hugh, 107, 109, 111, 152 machine, 120, 168, 297, 308, 312 Mackie, J.L., 234 Malebranche, Nicolas, 45 Mannheim, Karl, 134, 136, 138– 148, 151, 152, 154 many–valued logic, 13, 203

INDEX

Marcus, Ruth Barcan, 149, 226 Marcuse, Herbert, 138, 15 Margolis, Joseph, 115, 130, 195, 211–213 Marty, Anton, 5 Marx, Karl, 136, 138, 139, 153; Marxism, 138 Maslow, Alexander, 250–252, 261 Master of Balliol, 23 materialism, 138, 253, 254 mathematics, 1, 1, 2, 6–9, 15, 24, 41, 60, 61, 63, 72, 74, 77, 81, 109, 114, 115, 121, 129, 130, 148, 159, 173–176, 178, 190, 208–210, 214, 232–235, 271, 272, 281; mathematical, 2, 7, 15, 17, 38, 41, 45, 63, 65, 84, 107, 109, 110, 114, 117, 137, 148, 175, 197, 208, 210, 211, 213, 216–218, 220, 233, 234, 251, 252, 256, 259, 270–272 mathematical logic, 233, 251; mathematical models, 38; mathematical multiplicity, 252; mathematical necessity, 109, 210, 211, 213; mathematical truth, 114, 210, 234, 270–272; mathematician, 32, 39, 61, 73, 157, 172, 175, 176, 308 maxim, 62, 100, 157, 171, 174, 191, 248, 249 McDowell, John, 145,191,194, 285 McGinn, Colin, 253, 261 meaning, 3, 7–9, 18, 93, 104, 106, 115–118, 121, 123–126, 129, 130, 133, 137, 140, 142, 162, 185, 197–199, 201, 202, 205–508, 212, 223, 225, 234, 238–240, 243, 252, 253, 261, 263–266, 269, 270, 273–277,

327

280–284, 286, 287; meanings, 7, 8, 44, 116–119, 121, 123, 125, 128, 166, 176, 182, 199, 208, 221, 225, 238, 264–266, 272, 274, 275, 277–279, 283; meaningless symbols, 252 medicine, 12, 297 medieval trivium, 162 Meier, George Friedrich, 31, 32, 49, 152 Meja,V., 147–155 memory, 19,32,41,313 Menelaos, P., 21 mental, 1, 7, 8, 15, 18, 37, 39, 40, 60, 69, 70, 75, 76, 84, 115, 116, 119, 120, 123, 125, 127, 128, 130, 132, 141, 168, 170, 177, 182, 187, 190, 195–201, 205, 207, 208, 232, 264–268, 270, 274–281, 283–285, 289, 297, 303, 304, 306, 308, 311–313; mental act, 40, 70, 125, 127, 130, 177, 196; mental contents, 115, 190; mental entity, 15, 116, 120; mental experience, 116, 120, 132; mental life, 119, 123, 128, 200, 207, 280; mental logic (ML), 303, 304, 308, 312; mental metalogic (MML), 289, 304–306, 308, 309, 312, 315; mental models, 303; mental processes, 18, 75, 196, 198; mentalism, 8, 204–206, 276; mentalistic, 206, 281, 282 Mephistopheles, 26 Merkmale, 56, 57, 86 Merleau–Ponty, Maurice, 10 metaphilosophy, 213, 258, 286 metaphor, 44, 159, 175, 177, 277

328

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

metaphysics, 1, 6, 7, 32, 45, 46, 76, 78, 79, 115, 123, 135, 141, 160, 161, 178, 198, 212, 227, 245, 246, 250, 253–258, 260, 261, 277; metaphysics of morals, 245, 258, 260; metaphysical, 2, 25, 45, 51, 67, 68, 74, 115, 149, 174, 193, 198, 199, 216, 231, 250, 271, 284; metaphysical truths, 231; metaphysician, 254 metatheory, 308; metatheoretical, 10 method, 2, 11, 17, 28, 29, 33, 38, 45, 86, 87, 128, 130, 144, 146, 181–183, 188, 191, 194, 208, 212, 229, 230, 247, 251, 252, 254, 255, 267, 271; methods, 6, 10, 18, 86, 89, 90, 93–96, 100, 135, 158, 162, 182, 189, 191, 192, 240, 257, 287; methods of proof, 100; method of projection, 252; methodeutic, 162, 163, 175; methodological, 15, 161, 175, 183, 211, 268, 286; methodology, 39, 89, 102, 151, 178, 181, 182, 229, 245, 247, 256, 268; methodology of science, 89, 256 Meusel, A., 138, 152 Mill, James, 146 Mill, John Stuart, 6, 24, 47, 51–55, 58–60, 62, 64, 66, 69–76, 78, 79, 120, 161, 164, 165, 204, 210; Millian, 52, 53, 75 mind, 6, 7, 10, 11, 15–19, 21–30, 33, 34, 36–41, 43, 44, 58, 61, 62, 68, 71, 73, 76, 109, 114–116, 120, 121, 123–126, 129, 130, 133, 143, 149, 158, 161, 165, 166, 169, 171, 172, 177–179, 190, 194, 196, 202, 205, 213, 215, 218, 221, 225–227, 230,

232, 237, 239, 240, 250–252, 254–256, 260, 263–266, 278– 280, 283–286, 294, 296, 297, 299, 301, 303, 307, 312; minds, 11, 15, 114, 127, 160, 171, 175, 204, 215, 226, 229, 232, 234, 250, 254, 278, 311, 312; mindindependent entities, 6, 21 Misak, C.J., 178 ML, 303, 304; see mental logic MML, 289, 304, 305, 308, 309, 312; see mental metalogic modal logic, 13, 90, 206, 293; modal predicates, 92; modalities, 82, 90, 94, 98–100 modern logic, 10, 81, 82, 96, 100, 107; see logic modus ponendo ponens, 12, 71, 299 Moffett, Mark, 227 Mohanty, J.N., 130, 196–201, 204, 205, 207, 210, 211, 213, 249, 261 monad, 51; see Leibniz Montague, Richard, 216, 217, 225 Montaigne, Michel, 25 Moore, G.E., 232, 233, 312 morality, 158, 232, 233, 237; moral norms, 96; moral philosophy, 7, 246–248, 250; moral purpose, 96 Morick, H., 285, 286 Morscher, Edgar, 46, 49 Mugnai, M., 76, 79 Mulligan, Kevin, 17,19 Murphey, Murray, 178 Musgrave, Alan, 3, 19, 145, 151, 153, 154, 252, 261 N Nails, Debra, 256, 261 Natorp, Paul, 118, 124, 130, 132, 152, 260

INDEX

nature, 10, 17, 22, 24, 28, 29, 31, 37, 56, 57, 64, 66, 72–74, 76, 85, 88, 95, 113, 115, 117, 134, 136, 158, 159, 164, 165, 170–172, 175, 176, 184, 190, 194, 199, 217, 219, 229, 231, 234, 238, 246–248, 250, 252, 254, 258–260, 270, 279, 281, 283; natural history, 84, 174; natural logic, 22, 26, 29, 34; natural reason, 24, 27, 31; natural science, 6, 8, 18, 65, 84, 138, 146, 163, 182–184, 187, 188, 192, 238, 250, 256, 257, 266, 272, 280; natural sciences, 18, 138, 146, 256, 272; natural selection, 184, 185, 190, 192; naturalise, 142; naturalism, 7, 113–115, 127, 141, 181–183, 186, 187, 191–193, 196, 197, 201, 204, 205, 261, 266; naturalistic, 114, 115, 117, 128, 181, 183, 186, 187, 191–193, 196, 197, 201, 204, 205, 261, 266; naturalistic epistemology, 117, 181, 187, 193, 261; naturalistic framework, 186; naturalistic philosophy, 114; naturalized epistemology, 181, 185, 188, 190–194, 260, 263, 269, 274, 275, 279, 280, 285; naturalizers, 204 Navickas, Joseph L., 253, 261 necessity, 34, 35, 70, 71, 76, 78, 88, 91–100, 109, 123, 163, 175, 197, 204–213, 234, 246, 255, 264, 265; necessity-claims, 91 Neo-Kantian, 118; Neo-Kantians, 132, 134, 141, 144 Neurath, Otto, 138, 139, 145, 153, 188, 192

329

neurology, 281; neuropsychology, 12 Newton, Isaac, 123 Nicholas St. John Green, 174 noema, noesis, 125,128, 201, 210 Nola, Robert, 211 nominalism, 56; nominalist, 53, 54, 57; nominalistic, 76, 164 norms, 4, 96, 128, 158, 160, 162, 170, 178, 181, 184–186, 189–192, 196; normative, 3, 12, 70, 71, 74, 75, 84, 85, 87–89, 101, 109, 119, 122, 132, 158–160, 162–164, 169–171, 173, 175, 181, 184, 186, 190, 191, 196, 248; normative claim of logic, 84; normative function, 85, 88; normative psychologism, 84, 85, 87, 89, 101; normative task, 87, 101; normativity, 184, 186, 190–192; norms of discovery, 190 Notturno, Mark A., 19, 211, 213, 214, 261 noumena, 55, 59, 76 number, 15, 18, 38, 41, 44, 61, 132, 133, 205, 220, 226, 303; numbers, 15, 36, 39, 46, 196, 210, 217, 223, 227, 233, 290, 308

o O’Hear, Anthony, 191, 194 object, 7, 19, 35–37, 42, 54–61, 64, 68, 69, 76, 77, 84, 87, 89–91, 107, 116–118, 125, 126, 166, 167, 170–172, 174, 175, 177, 185, 202, 226, 227, 230, 233, 243, 285; objects, 14, 15, 36, 42, 46, 54, 55, 57–60, 62–68, 72–77, 102–104, 107, 111, 114, 117,

330

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

125, 166, 167, 182, 187, 201, 215, 216, 221, 226, 230, 232, 233, 235, 246, 247, 249, 252, 255, 258, 267, 276, 280, 304, 305; objects of consciousness, 232; object-directedness, 125, 126; object-level, 226; object– oriented, 126; objective, 2, 6–9, 14–16, 21, 39, 41, 42, 61, 63, 64, 72–74, 81, 83, 84, 92, 93, 114, 116, 118, 119, 127, 134, 135, 146, 148, 152, 159, 160, 167, 169, 176, 196–198, 208, 211, 226, 274–276, 282, 283, 290; objective judgment, 74; objectivist, 82–85, 87, 88, 91, 209, 243; objectivity, 11–13, 16, 83, 84, 114, 117, 124, 127, 128, 137, 138, 146, 154, 213, 245, 247, 254, 255, 256, 261, 270; objectivity for pure logic, 84; objectification, 118, 267 observation, 11, 13, 70, 141, 174, 175, 185, 191, 251, 273, 276, 279, 281–283; observation sentence, 185, 273, 279 Oliveira, Manfredo Araújo, 255, 261 ontology, 7, 46, 57, 64, 75, 144, 193, 194, 220, 227, 253, 255, 285; ontological, 2, 19, 22, 46, 55–58, 117, 121, 140, 141, 194, 195, 213, 217, 218, 220, 221, 227, 255, 261, 286; ontological status, 22, 46 ordered pair, 217, 219; see class, classes; set, sets ordinary language, 225, 265, 266, 270, 271 OSCAR, 308; see Pollock

P pain, 186, 236, 248 Pap, Arthur, 3, 19 Pappus, 29, 45 paradox, 124, 125, 130, 259, 286; paradoxes of existential import, 107 Parsons, Charles, 213 Parsons, Terence, 227 Paton, H.J., 249, 261 Peckhaus, Volker, 108, 111 Peirce, Benjamin, 157 Peirce, Charles Sanders, 63, 157–179 Penrose, Roger, 308 perception, 9, 15, 59–63, 77, 115, 127, 205, 230, 232, 254, 259, 265, 269, 283; perceptual apparatus, 185; perceptual manifold, 72, 74 Perloff, Michael, 109 person, 4, 11, 34, 37, 114, 119, 120, 161, 171, 174, 188, 223, 255, 257, 272, 284, 290 perspective, 2, 52, 59, 61, 66, 90, 137, 139, 166, 169, 175, 192, 241–243, 254, 261, 268, 274, 298, 304; perspectives, 19, 142, 179, 194, 211, 213, 214, 227, 242, 261, 285 phenomena, 68, 76, 77; phenomenal, 35, 57, 59, 68, 76, 230–232 phenomenology, 7, 10, 17, 128–130, 143–145, 154, 173, 175, 211–213, 245, 249; phe– nomenological, 11, 12, 48, 73, 118, 128, 129, 142–144, 148, 173, 196, 201, 212, 249, 261;

INDEX

phenomenological intuition, 128; phenomenologists, 141, 196, 204 Philalethes, 24 Philipse, Herman, 17, 19, 258, 261 philosophy of language, 1, 9, 181, 198,212 philosophy of logic, 6, 8, 9, 13, 16, 19, 51–53, 66, 69, 82, 83, 196, 252, 258 philosophy of mind, 17, 19, 196, 221, 266 philosophy; philosophical analysis, 279; philosophical norms, 190; philosophical semantics, 7, 9, 253; philosophical truth, 229, 236; philosophical truths, 229 physics, 8, 10, 12, 23, 88, 89, 163, 183, 189, 210, 231, 235, 246, 257, 258, 265, 266, 281, 297; physics of thinking, 88, 89; physical entities, 115; physical objects, 182, 187, 255, 267; physical sciences, 10, 38, 284; physical theory, 273, 280; physicalism, 18, 182, 183, 191, 273, 283 physiology, 26, 164, 183, 189, 273, 277, 283; physiological, 26, 114, 160, 165, 276, 279, 283; physiological psychology, 114 picture theory of meaning, 252 Pinker, Steven, 280, 285 Platner, Ernst, 49,153 Plato, 21; platonic entities, 217, 220; platonism, 6, 195, 200, 206 pleasure, 45, 236, 248, 271 Plessner, H., 153 Polanyi, Michael, 167 polemic, polemics, 4, 6, 25, 38, 41, 136, 249 Pollock, John, 308

331

Pope, Alexader, 24 Popkin, Richard H., 259, 261 Popper, Karl, 145–148, 153, 193; Popperian, 19 posit, 263, 268, 269, 275; positing, posits, 8, 189, 257, 264, 265, 268–270, 275, 286 positivism, 207, 242 possibility, 54, 74, 76, 78, 91–93, 97–100, 109, 114, 201, 203, 258, 275, 293, 312; possibility from ignorance, 98, 100 post-modernism, 235 Posy, Carl, 79 Potter, Vincent J., 178, 179 pragmatic, 85, 99, 108, 121, 157, 174, 179, 243, 255; pragmatic critique, 121; pragmaticism, 174, 177; pragmatism, 157, 171, 174, 177, 178, 206; pragmatist, 2, 63, 206; practical activity, 7; practical anthropology, 258 pre-established harmony, 185, 242 pre-linguistic properties, 77, 78 precepts, 24, 30, 70, 71,74 predicate, 39, 40, 55–5967, 76, 93, 94, 100, 238, 239, 305, 308, 312; predicate concept, 57, 59; prediction, 74, 184, 191 prejudice, 26, 31, 32, 34, 270 prereflective act, 126 prescriptive, 11, 12, 18, 19, 87, 185, 234, 248, 261; prescriptive function, 19, 87, 261 Price, H.H., 259, 261 Prichard, 233 prima philosophia, 161 Principal of St. Edmund’s Hall, 23 principle of contradiction, 77, 88, 121; principle of identity, 209;

332

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

principled disjunction, 195, 196, 198, 201; principles of syllogism, 120 probability, 41, 208, 212, 251, 252, 257; probability theory, 41, 251 property, 40, 55, 125, 218–220, 223, 224, 226, 227, 265, 274, 282; properties, 27, 40, 46, 55–57, 64, 73, 76, 93, 102, 175, 215–220, 222–227, 252, 255, 256, 259, 265, 275, 304; property-level, 222 proposition, 15, 39–41, 44, 46, 58, 65, 71, 76, 92, 109, 116, 118, 121, 125–127, 133, 176, 215–226, 252, 292, 301, 312; propositions, 7, 16, 21, 22, 32, 37–42, 44, 46, 65, 75–77, 87, 91, 93, 115, 118–120, 123, 125, 126, 167, 168, 215–217, 219–227, 234, 243, 253, 289, 292; prepositional constituency, 217; propositional representation, 221, 224; propositionality, 218, 221, 222 propositional and predicate calculi, 308 psychologism, 1–11, 13–19, 21, 22, 37, 38, 42–44, 46, 51, 53, 66, 69, 81–91, 101, 107, 108, 110, 111, 113–136, 140–143, 145, 147–149, 151, 157–159, 161, 163–167, 169–171, 173, 177, 195–197, 199–211, 213–215, 224, 227, 229, 230, 232, 234, 235, 237, 239, 242–253, 255–258, 261, 289–292, 296, 297, 303, 305, 308, 309; psychologism-cum-relativism,

121; Psychologismus, 130, 149, 160, 161; psychologistic, 1, 3-5, 7, 16, 18, 22, 51–53, 66, 69, 73, 82–84, 86–90, 107, 108, 116–123, 131, 132, 143, 145, 158, 159, 164, 196, 198, 204–208, 211, 215, 229, 230, 232, 234, 247, 249–252, 254, 255, 258, 292; psychologistic logic, 81, 84, 86, 89, 119, 168; psychologizers, 204 psychology, 1, 2, 4–18, 21–23, 29, 31, 33–36, 38, 39, 41–44, 69–72, 82, 83, 85–89, 101, 114–116, 119–124, 129, 132–134, 139, 141–143, 147, 157, 158, 160, 161, 163, 164, 169–172, 174–176, 182, 183, 187, 189, 196, 204, 229, 230, 234, 245, 246, 248, 250–253, 256–258, 260, 263, 266, 268, 273, 276, 279, 282, 289, 290, 296, 297, 303, 308, 309, 311–313; psychological, 1–3, 6–11, 14–16, 18, 21, 28, 29, 32, 34–38, 41–43, 51, 70, 71, 82–91, 101, 108, 113–115, 118–120, 123–128, 133, 137, 142–144, 151, 157, 158, 160, 161, 163, 165, 167–172, 174, 176, 185, 195, 197, 201, 203, 205, 245–254, 256–258, 266, 267, 282, 308; psychical, psychical event, 86, 169, 171; psychophysical, 100; psychological apperception, 128; psychological entities, 15, 133; psychological factors, 142, 163, 201, 205, 250; psychological investigation, 34;

INDEX

psychological processes, 142, 163, 201, 205, 250; psychological properties, 256; psychological science, 1, 6, 8, 10; psychological states, 256, 267; psychological subjects, 14, 15, 257 pure forms of intuition, 230, 247 pure logic, 32–36, 69, 84, 86, 90, 108, 115, 119, 123, 124, 130 pure reason, 32, 34, 59, 60, 63, 64, 66, 68, 72, 74, 78, 159, 175, 246, 247, 255, 260 Putnam, Hilary, 75, 79, 200, 210, 211, 213, 281, 285 Q

quantifiers, 68, 223, 239; quantificational logic, 229 quantum physics, 10 Quine, W.V.O., 6–8, 17, 19, 145, 153, 181–194, 203–214, 216, 225–227, 235, 237, 239, 243, 256–258, 261–266, 268–270, 272–274, 276–283, 285, 286; Quinean, 181, 192, 204, 235, 264, 277–279, 283, 284; Quinean naturalism, 279 R

radical translation, 265, 267; Quine Radnitzky, Gerhard, 4, 19 Ramus, Peter, 45 Rappaport, Steven, 281, 286 Raspa, Venanzio, 109, 111 Rath, Mathias, 17, 19, 108, 153, 249, 261 rational, 23, 57, 84, 116, 126, 160, 162, 169, 174–177,

see

111, 127, 181,

333

190, 196, 209, 245–247, 249, 254, 255, 258, 267, 282; rationality, 84, 162, 175, 193, 211, 240, 243, 267, 272, 273, 286, 313; rationalism, 146, 198; ratiocination, 31, 168 reality, 34, 46, 54, 55, 58, 69, 76, 87, 115, 137, 138, 141, 148, 175, 190, 195, 198, 199, 206, 215, 230–233, 254, 256, 261, 263–265, 271, 275, 280, 286, 308; realism, 56, 66, 75, 148, 182, 193, 198, 199, 210, 211, 213, 229, 230, 232–234, 236, 243, 259, 262 reason, 11, 12, 14, 17, 22–28, 30–36, 45, 57, 59, 60, 62–66, 69, 72–76, 78, 79, 84, 87, 90, 92, 96, 99, 103, 109, 131, 158, 159, 163, 168, 175–177, 179, 199, 201, 203, 205, 209, 211, 215, 216, 219, 220, 224, 237, 239–242, 246–248, 254–256, 259–261, 274, 281, 283, 290, 291, 294, 297, 301, 303, 304; reasoning, 4, 6, 11–14, 16, 17, 22–24, 26, 31, 34, 35, 53–55, 63, 65, 68–71, 75, 76,78, 110, 148, 161–164, 168, 175, 176, 204, 207, 234, 245, 247, 248, 251, 252, 255, 257, 258, 267, 289–291, 296–298, 303, 304, 308, 309, 311–313 reciprocal containment, 8, 188, 257 reconceptualization, 170, 184 reductio ad absurdum, 296, 303 reductionism, 43. 160, 165, 208 reference, 26, 37, 41, 45, 47, 52, 55, 58, 60–69, 75, 93, 96, 97, 104, 109, 130, 133, 148, 158, 168, 169, 171, 172, 176, 187, 194,

334

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

200, 201, 204, 208, 209, 227, 234, 241, 249, 258, 271, 272; referential, 51, 52, 55, 58–60, 62, 64, 65, 68, 25, 201, 221, 239; referential and cognitive semantic theories, 51, 52; referentialism, 52, 65 Rehmke, J. 153 Reid, Thomas, 26, 27, 45, 49, 53, 153, 262 Reimarus, H.S., 46, 49, 153 relation, 14, 27, 40, 46, 54, 55, 58, 67, 82, 88, 89, 98, 104, 107, 117, 125, 127, 130, 177, 183, 186, 188, 192, 208, 215, 216, 218–223, 226, 227, 243, 277, 279, 283, 293; relations, 7, 15, 21, 27, 37, 39, 40, 43, 46, 67, 73, 76, 79, 85, 87, 93, 94, 101–104, 115, 147, 154, 159, 208, 210, 215–219, 222, 223, 227, 230, 243, 253, 263, 282; relations of ideas, 27; relata, 46, 223, 227 relativism, 15, 19, 116, 117, 199– 122, 125, 133, 137, 139, 144, 147, 193, 201, 204, 211, 213, 229, 235, 236, 238–243; relativist conceptions of truth, 229; relative necessity, 94–96; relativization, 235 remembering, 187, 278 representation, 35, 41, 53, 58, 59, 77, 78, 220, 221, 224–226, 285, 299–301, 305, 312; representations, 35, 36, 38, 40, 44, 67, 71, 72, 224, 225, 302; representational, 217, 218, 220; representationalism, 217, 218, 220; representative, 161, 245, 249, 280

Rescher, Nicholas, 109, 243, 255– 258, 261 Rickert, H., 135, 151, 153 Ricketts, Thomas, 272–274, 282, 286 Riemannian geometry, 13 Rinella,K.,308,312 Rips, L., 261, 296, 303, 304, 312, 313 Rorty, Richard, 186, 194 Roth, Paul A., 263, 278–280, 283, 286 rule, rules, 3, 22–25, 28, 31, 32, 34, 41, 70, 71, 75, 83, 86, 87, 120, 124, 133, 164, 169, 175, 177, 179, 182, 186, 190, 196, 198, 200, 201, 203, 235, 241, 252, 258, 268–274, 278, 279, 281, 282, 285, 294–296, 288, 301, 303–305, 311, 312; rules of thought, 196; rule-governed, 252 Rusnock, Paul, 45, 46, 48 Russell, Bertrand, 8, 145, 182, 187, 188, 208, 210, 213, 222, 226, 227, 232, 233, 311, 313 Ryle, Gilbert, 232 S sameness, 185, 270, 276, 282, 283; see identity Satz an sich, 39; see sentences in themselves Saussure, Ferdinand de, 176, 179 Savan, David, 158, 173, 179 Savary, F., 290, 311, 312 Scheler, Max, 134–136, 141–144, 146, 153 Schelting, A. von, 139, 153 schematic, 61, 62, 72; see Kant

INDEX

Schiller, Johann Christoph Friedrich von, 175 Schilpp, Paul Arthur, 18, 184, 193, 212, 213 Schlick, Moritz, 5, 133,153 Schofield, Malcolm, 254, 261 scholastic distinction, period of philosophy, 23–25, 27–29, 32, 122 Schopenhauer, Arthur, 115 Schröder, Ernst, 101–103, 111, 154 Schuppe, W., 122, 130 Schütz, A., 143, 153 science, 1, 6–8, 10–12, 14, 17–19, 23, 25, 26, 32–35, 38, 39, 41–45, 47, 65, 66, 69, 70, 77, 81, 84, 86, 87, 89, 90, 93, 100, 101, 108, 110, 118–121, 131, 135, 136, 141, 142, 147–149, 153, 154, 157, 159, 163–165, 169, 170, 171, 174–176, 178, 181–194, 205, 212, 230, 232, 234, 235, 238–240, 245, 246, 250–252, 254, 256, 257, 261–266, 269, 271, 272, 274, 277–280, 284, 286, 308, 309, 312, 313; science of logic, 17, 164, 170, 171, 251, 252, 262; science of thinking, 44; sciencefriendly, 240; sciences, 2, 2, 10, 12, 14, 18, 23, 28, 34, 38, 41, 47, 78, 86, 93, 100–102, 118, 130, 136–138, 144, 146, 158, 162, 164, 169, 174, 175, 178, 203, 212, 246, 247, 251, 254, 256, 260, 263, 272, 284, 287, 297, 313; scientific approach, 38; scientific discovery, 28, 193; scientific epistemology, 181, 183; scientific explanation,

335

110, 149, 263, 264, 270, 275; scientific knowledge, 101, 146, 147, 183, 184; scientific language, 93; scientific method, 38, 181, 183, 188, 254; scientific picture, 278; scientific psychology, 6, 12, 13, 17, 43, 87; scientific rationality, 272; scientific study, 16, 283; scientific theory, 9, 199; scientific worldview, 284 Seebohm, Thomas, 114, 130, 204, 211, 213 self, 35, 36, 62, 175, 177, 259; selfcorrective, 169; self-evident, 121, 246, 275 Sellars, Wilfred, 145, 159, 266, 267, 281, 286, 287; Sellarsian, 261, 266–268, 276, 278 semantic, semantics, 1–4, 6–9, 19, 48, 49, 51–53, 55–60, 64–69, 75, 197, 198, 202, 205–207, 212, 214, 221, 225, 226, 253, 258, 276, 286, 304, 305, 308, 311, 312; semantic humanism, 69; semantic relation, 55; semantic theory, 52, 55, 58, 67 semeiotic, 162–164, 170, 173, 175, 176, 178, 179; semiotic, 159, 176, 177, 179 sense-data, 6, 187, 232, 257; sensory data, 182, 183, 188, 192; sensible intuition, 67, 68, 78; sensible manifold, 247; sensible world, 67, 68; sensitive intuition, 68; sensory evidence, 182; sensory perception, 59; sensualism, 43

336

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

sentence token, 224, 225; sentences in themselves, 21, 46; see Satz an sich sequential inferences, 304 set, sets 10, 12, 25, 35, 36, 38, 40, 56, 57, 61, 67, 89, 95, 97, 98, 105, 123, 136, 162, 165, 175, 186, 208, 216, 219, 226, 248, 269, 274, 294–296, 298, 301, 311 shibboleth, 1, 5, 312 Shimony, Abner, 256, 261 Short, T.L., 172, 179 Sider, Ted, 226 sign, signs, 61, 107, 157–160, 162–168, 170–173, 175–177, 187, 250, 252, 270 Sigwart, Christoph, 82, 85, 86, 89–96, 98–100, 107–109, 111, 119, 120, 123, 124, 130–133, 154; Sigwartian, 90, 94 Simon, Herbert, 308, 309; Simon’s dream, 308 simplicity, 19, 199, 240, 281 Singer, K., 140, 154 singular concept, 57, 77 Skagestad, Peter, 177, 179 skeptic, 181, 188, 189; skeptical, 202, 223, 254, 265, 280, 282; skepticism, 15, 76, 114, 188, 189, 191, 195, 259, 277, 282; sceptic, 262; scepticism, 31 Skorupski, John, 211, 214 Sluga, Hans, 114 Smith, Barry, 124–126, 130, 213 Smith, John E., 175, 176, 179 Soames, Scott, 227 Sober, Elliott, 145, 193 sociology, 1, 19, 22, 43, 131–149, 151, 152, 211, 213, 249, 261;

sociology of knowledge, 131, 134–140, 142–149, 152, 211, 213; sociological, 2, 5, 136, 140, 141–143, 145, 209; sociologism, 131, 134, 135, 140, 141, 143–147; socio-philosophical, 142 solipsism, 15, 232 Sombart, W., 138, 154 soundness, 83–85, 87, 88, 90, 106, 109, 304 space, 21, 31, 39, 41, 63, 68, 73–75, 134, 157, 159, 170, 173, 231, 247, 277, 300, 312; space and time, 39, 73, 74, 247; spatio-temporal, 64, 67; space of signs, 157, 159, 170, 173 specific relativism, 120, 121 speculative metaphysics, 123 Speier, H., 139, 154 Spengler, O., 134, 154 Spinoza, Benedictus (Baruch) de, 254, 255, 262 Spranger, E., 140, 154 Standard Social Science Model (SSSM), 280 Stanovich, K.E., 311,313 state of affairs, 252, 292 stechiology, 162, 163 Stelzner, Werner, 108, 109, 111 Stern, G., 139, 154 Stevenson, 234, 237 stimulus meaning, 7, 185, 281, 283 Stoa, 254 strong psychologism, 18, 108; see weak psychologism subextensiveness, 219, 222 subject, 1, 8, 11, 12, 15, 22, 23, 25–27, 30, 32, 33, 35, 37, 39, 40, 42, 44, 45, 55–57, 67, 69,

INDEX

70, 76, 77, 92–94, 99, 100, 103, 117, 125, 168, 172, 175, 177, 182, 188–190, 192, 222, 229, 245, 246, 251, 255, 257–271, 279, 281, 282, 311; subjects, 8, 14, 15, 27, 31, 33, 42, 60, 87, 88, 93, 94, 99, 144, 195, 222, 257, 279, 291, 292, 296, 299, 301, 311; subjective, 2, 8, 10, 13, 15, 16, 34, 35, 42, 55, 60, 61, 63, 69, 93, 114, 117, 118, 127, 135, 152, 184, 195, 196, 198, 205, 229, 240, 245–247, 255, 257; subjective psychology, 257; subjectivity, 9, 11, 12, 14, 124, 127, 142, 247, 259, 261, 281; subjectification, 267 sufficient reason, 57 Swoyer, Chris, 249, 262 syllogism, 23, 25, 33, 45, 118, 120, 167, 280, 291 symbol, 96, 106, 168, 252; symbols, 167, 168, 172, 174, 176, 250–252; symbolic, 10, 81, 101, 103, 111, 152, 257, 293, 304, 306, 308; symbolic logic, 10, 111, 152, 257, 293, 304, 306, 308 synonymy, 212, 273 syntactic, 91, 201, 205, 206, 274, 304, 305, 308, 311; syntactic rules, 304; syntactical, 91, 201, 274, 311 T

Tarski, Alfred, 40, 49, 202, 211, 233, 234; Tarski-like, 202 tautology, 43, 89, 209 taxonomy, 1, 12, 114, 117 Taylor, Charles, 175

337

technology, 119, 184, 190, 313; technological, 190, 236 teleological necessity, 95–98; teleological possibility, 98 temporality, 73 terms, 7, 8, 15, 21, 22, 25, 35, 36, 40, 44, 52, 66, 75–77, 83, 84, 90, 96, 104, 118, 140, 158, 162, 174, 187, 188, 197, 198, 200, 201, 205–211, 226, 234, 248, 256, 265, 267, 268, 270, 272–274, 276, 277, 279–282, 300 tertium non datur, 197, 199, 202, 209, 211 theory, theoretical; theoretical restrictions, 282; theoretical science, 119, 184; theory of knowledge, 1, 91, 118, 142, 144, 256; theory of mathematics, 61; theory of meaning, 115, 116, 118, 124–126, 130, 199, 252; theory of syllogisms, 78; theory revision, 240 thought, 1–4, 6–11, 13–16, 18, 21, 22, 24–30, 35–46, 48, 53, 55, 58, 59, 61, 64, 66, 68–71, 74–77, 83, 85, 88, 95, 101, 110, 113, 115, 119, 120, 125, 126, 128–130, 132, 135, 136, 138, 141–143, 146, 148, 151, 160–163, 165–169, 172–178, 187, 193, 195–201, 204–211, 213, 215, 220, 224, 226, 229, 231–233, 235, 237, 238, 241–243, 249–254, 259, 267; thoughts, 7, 14, 19, 21, 23, 29, 37, 42, 46, 76, 116, 125, 161, 167, 177, 196, 215; thought experiment, 241; thinking, 3, 6,

338

PHILOSOPHY, PSYCHOLOGY, AND PSYCHOLOGISM

11, 17, 21, 22, 24, 25, 34, 36, 38, 42, 44–47, 65, 70, 76, 77, 81, 83–89, 95, 99, 101, 104, 113, 116–118, 120, 122, 123, 125, 126, 128, 132, 135, 138–141, 146, 166, 167, 176, 177, 191, 196, 197, 199–202, 208, 209, 216, 227, 230–232, 236, 240, 243, 246, 248, 250–252, 301; thinkers, 4, 15, 117, 160, 221, 247, 248, 251, 291 Thomasius, Christian, 31,49 Thompson, Manley, 164, 179 Tillich, Paul, 139, 154 time, 30–32, 39–44, 52, 58, 60, 63, 68, 73, 74, 82, 88, 99, 100, 103, 113, 119, 124, 128, 131, 137, 142, 144, 146, 159, 160, 166, 170, 172, 176, 190, 210, 216, 231, 235–237, 241, 247, 248, 268, 290, 301, 311; timelessness, 119; see space and time; spatio-temporal token, 224, 253, 283; see type totality, 68, 104–107, 284 traditional logic, 12, 23, 24, 28, 29, 33, 38, 81–83, 85, 86, 89, 107, 108; traditional logicians, 81, 82, 85 transcendent, transcendence, 1, 2, 241, 242; transcendent notion of truth, 241, 242; transcendental, 2, 34, 36, 43, 63–66, 72, 74, 77, 85, 88, 108, 114, 115, 124, 127–129, 189, 196, 205, 211–213, 247, 249, 255, 259; transcendental aesthetic, 247; transcendental deduction, 63, 72; transcendental idealism, 64;

transcendental illusion, 77; transcendental logic, 34, 85, 114, 115, 124, 128, 129; transcendental phenomenology, 128, 212, 249; transcendental psychologism, 114, 127, 128; transcendental realism, 66; transcendental reasoning, 255; transcendentalism, 85, 108 truth, 2, 3, 8, 12, 13, 16–19, 23, 25, 28–30, 41, 42, 45, 46, 52–55, 57–60, 62–67, 69, 71, 74–79, 83, 85, 89–92, 99, 115–119, 121, 123, 126, 127, 129, 132–135, 137, 138, 141, 143, 144, 158, 163, 168, 171, 175, 177–179, 181–183, 194–200, 202–204, 206–217, 220, 221, 224–226, 229, 231, 233–236, 238–244, 252, 253, 255, 256, 261, 266, 270, 272, 280–282, 290, 292, 296, 297, 303, 312; truth referentially, 64; truth table definition, 270; truth-claims, 91, 92 type, 11, 14, 135, 216, 217, 250, 251, 253, 256, 257, 269, 275; see token U

unconditional normative psychologism, 73–84 underdetermination, 265, 269, 280, 281 understanding, 7–9, 23–26, 28, 29, 32–36, 43, 44, 48, 72–74, 77–79, 86, 108, 115, 118, 128, 131, 151, 158, 160, 163, 165, 169, 170, 173, 183, 192, 229, 230, 232, 246, 248, 250, 257,

INDEX

259, 267, 277–280, 284, 285, 296 unity of consciousness, 42 universals, 7, 30, 270; universality, 13, 207, 247; universalization, 62, 72, 249; universal law, 248, 249; universal soundness, 88 unregenerate realism, 182 urteil, 83, 111, 154; urteilen, 44 Utopia, 137, 152 V valid, validity, 3, 27, 35, 53, 55, 62–72, 75, 83, 87, 117–119, 121, 133, 139–144, 160, 162, 164, 165, 203, 208, 210, 246, 248, 250, 251, 257, 292, 294, 300, 301, 303, 304; valid argument, 55, 300, 302, 303; valid thinking, 70 value theory, 7 Vasil’ev, Nicolai, 82, 107–109, 111 veil of perception, 182; see Quine Verhältnis, 148, 161, 260 Vergauwen, Roger, 109, 111 voluntative logic, 95 Vorstellung, 83, 113, 258 W Ward, Seth, 30, 49 Warden of Wadham, 24 Watkins, J.W.N., 147, 154 weak psychologism, 18; see strong psychologism Weber, Alfred, 47, 138, 148, 154 Welby, Victoria, 171, 172 West, R.F., 311, 313 Wiener, P., 171 Wiggins, David, 243

339

Willard, Dallas, 124–126, 130, 145, 154, 205, 211, 214, 227, 259 Williams, Bernard, 235, 238, 239, 243, 244 Wilson, Fred, 259, 262 Windelband, Wilhelm, 132, 154, 155 Wittfogel, K.A., 139, 155 Wittgenstein, Ludwig, 2, 8, 19, 145, 148, 165, 177, 179, 183, 209, 232, 249, 250, 252, 253, 261, 262, 277, 278, 281–285; Wittgensteinian, 125, 206, 211, 278, 284, 285 Wolff, Christian, 26, 29–31, 45, 48, 49, 151 world-view, 238, 243 Wright, G.N., 49,153 Wundt, Wilhelm, 82, 85, 86, 89, 100–108, 111, 123, 130–132, 141, 143, 145, 146, 155 Y Yang, Yingrui, 308, 311–313 Yolton, John W., 259, 262 Z Zeman, J.Jay, 172, 179 Zeno of Citium, 254, 255, 260–262 Zheng, Lan, 164, 179

This page intentionally left blank

PHILOSOPHICAL STUDIES SERIES 52. 53. 54.

55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.

70. 71. 72.

73. 74. 75. 76.

Jesús Ezquerro and Jesús M. Larrazabal (eds.): Cognition, Semantics and Philosophy. Proceedings of the First International Colloquium on Cognitive Science. 1992 ISBN 0-7923-1538-3 O.H. Green: The Emotions. A Philosophical Theory. 1992 ISBN 0-7923-1549-9 Jeffrie G. Murphy: Retribution Reconsidered. More Essays in the Philosophy of Law. 1992 ISBN 0-7923-1815-3 Phillip Montague: In the Interests of Others. An Essay in Moral Philosophy. 1992 ISBN 0-7923-1856-0 Jacques-Paul Dubucs (ed.): Philosophy of Probability. 1993 ISBN 0-7923-2385-8 Gary S. Rosenkrantz: Haecceity. An Ontological Essay. 1993 ISBN 0-7923-2438-2 Charles Landesman: The Eye and the Mind. Reflections on Perception and the Problem of Knowledge. 1994 ISBN 0-7923-2586-9 ISBN 0-7923-2595-8 Paul Weingartner (ed.): Scientific and Religious Belief. 1994 Michaelis Michael and John O’Leary-Hawthorne (eds.): Philosophy in Mind. The Place of Philosophy in the Study of Mind. 1994 ISBN 0-7923-3143-5 William H. Shaw: Moore on Right and Wrong. The Normative Ethics of G.E. Moore. 1995 ISBN 0-7923-3223-7 T.A. Blackson: Inquiry, Forms, and Substances. A Study in Plato’s Metaphysics and Epistemology. 1995 ISBN 0-7923-3275-X Debra Nails: Agora, Academy, and the Conduct of Philosophy. 1995 ISBN 0-7923-3543-0 Warren Shibles: Emotion in Aesthetics. 1995 ISBN 0-7923-3618-6 John Biro and Petr Kotatko (eds.): Frege: Sense and Reference One Hundred Years Later. 1995 ISBN 0-7923-3795-6 Mary Gore Forrester: Persons, Animals, and Fetuses. An Essay in Practical Ethics. 1996 ISBN 0-7923-3918-5 K. Lehrer, B.J. Lum, B.A. Slichta and N.D. Smith (eds.): Knowledge, Teaching and Wisdom. 1996 ISBN 0-7923-3980-0 ISBN 0-7923-4033-7 Herbert Granger: Aristotle’s Idea of the Soul. 1996 Andy Clark, Jesús Ezquerro and Jesús M. Larrazabal (eds.): Philosophy and Cognitive Science: Categories, Consciousness, and Reasoning. Proceedings of the Second International Colloquium on Cogitive Science. 1996 ISBN 0-7923-4068-X ISBN 0-7923-4401-4 J. Mendola: Human Thought. 1997 J. Wright: Realism and Explanatory Priority. 1997 ISBN 0-7923-4484-7 X. Arrazola, K. Korta and F.J. Pelletier (eds.): Discourse, Interaction and Communication. Proceedings of the Fourth International Colloquium on Cognitive Science. 1998 ISBN 0-7923-4952-0 E. Morscher, O. Neumaier and P. Simons (eds.): Applied Ethics in a Troubled World. 1998 ISBN 0-7923-4965-2 R.O. Savage: Real Alternatives, Leibniz’s Metaphysics of Choice. 1998 ISBN 0-7923-5057-X Q. Gibson: The Existence Principle. 1998 ISBN 0-7923-5188-6 F. Orilia and W.J. Rapaport (eds.): Thought, Language, and Ontology. 1998 ISBN 0-7923-5197-5

PHILOSOPHICAL STUDIES SERIES 77. 78. 79. 80. 81.

82. 83. 84. 85. 86. 87. 88. 89. 90. 91.

J. Bransen and S.E. Cuypers (eds.): Human Action, Deliberation and Causation. 1998 ISBN 0-7923-5204-1 R.D. Gallie: Thomas Reid: Ethics, Aesthetics and the Anatomy of the Self. 1998 ISBN 0-7923-5241-6 K. Korta, E. Sosa and X. Arrazola (eds.): Cognition, Agency and Rationality. Proceedings of the Fifth International Colloquium on Cognitive Science. 1999 ISBN 0-7923-5973-9 M. Paul: Success in Referential Communication. 1999 ISBN 0-7923-5974-7 E. Fischer: Linguistic Creativity. Exercises in ’Philosophical Therapy’. 2000 ISBN 0-7923-6124-5 R. Tuomela: Cooperation. A Philosophical Study. 2000 ISBN 0-7923-6201-2 P. Engel (ed.): Believing and Accepting. 2000 ISBN 0-7923-6238-1 W.L. Craig: Time and the Metaphysics of Relativity. 2000 ISBN 0-7923-6668-9 D.A. Habibi: John Stuart Mill and the Ethic of Human Growth. 2001 ISBN 0-7923-6854-1 M. Slors: The Diachronic Mind. An Essay on Personal Identity, Psychological Continuity and the Mind-Body Problem. 2001 ISBN 0-7923-6978-5 L.N. Oaklander (ed.): The Importance of Time. Proceedings of the Philosophy of Time Society, 1995–2000. 2001 ISBN 1-4020-0062-6 M. Watkins: Rediscovering Colors. A Study in Pollyanna Realism. 2002 ISBN 1-4020-0737-X W.F. Vallicella: A Paradigm Theory of Existence. Onto–Theology Vindicated. 2002 ISBN 1-4020-0887-2 M. Hulswit: From Cause to Causation. A Peircean Perspective. 2002 ISBN 1-4020-0976-3; Pb 1-4020-0977-1 D. Jacquette (ed.): Philosophy, Psychology, and Psychologism. Critical and Historical Readings on the Psychological Turn in Philosophy. 2003 ISBN 1-4020-1337-X

KLUWER ACADEMIC PUBLISHERS – DORDRECHT / BOSTON / LONDON