Computers and Cognition: Why Minds are not Machines PDF
Preview Computers and Cognition: Why Minds are not Machines
COMPUTERS AND COGNITION: WHY MINDS ARE NOT MACHINES STUDIES IN COGNITIVE SYSTEMS VOLUME 25 EDITOR James H. Fetzer, University of Minnesota, Duluth ADVISORY EDITORIAL BOARD Fred Dretske, Stanford University Charles E. M. Dunlop, University of Michigan, Flint Ellery Eells, Univeristy of Wisconsin, Madison Alick Elithom, Royal Free Hospital, London Jerry Fodor, Rutgers University Alvin Goldman, University ofA rizona Jaakko Hintikka, Boston University Frank Keil, Cornell University William Rapaport, State University of New York at Buffalo Barry Richards, Imperial College, London Stephen Stich, Rutgers University Lucia Vaina, Boston University Terry Winograd, Stanford University COMPUTERS AND COGNITION: WHY MINDS ARE NOT MACHINES by JAMES H. FETZER University 0/ Minnesota, Duluth, U.S.A. SPRINGER-SCIENCE+BUSINESS MEDIA, B.V. A c.I.P. Catalogue record for this book is available from the Library of Congress. ISBN 978-1-4020-0243-4 ISBN 978-94-010-0973-7 (eBook) DOI 10.1007/978-94-010-0973-7 Transferred to Digital Print 2001 Printed on acid-free paper All Rights Reserved © 2001 Springer Science+Business Media Dordrecht Originally published by Kluwer Academic Publishers in 2001 No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, incJuding photocopying, recording or by any information storage and retrieval system, without written permission from the copyright owner. To Janice E. Morgan CONTENTS Series Preface ix Foreword xi Acknowledgments xxi PROLOGUE 1. Minds and Machines: Behaviorism, Dualism and Beyond 3 PART I: SEMIOTIC SYSTEMS 2. Primitive Concepts: Habits, Conventions, and Laws 25 3. Signs and Minds: An Introduction to the Theory of Semiotic Systems 43 4. Language and Mentality: Computational, Representational, and Dispositional Conceptions 73 PA~T II: COMPUTERS AND COGNITION 5. Mental Algorithms: Are Minds Computational Systems? 101 6. What Makes Connectionism Different? A Criticial Review of Philosophy and Connectionist Theory 131 7. People are Not Computers: (Most) Thought processes are Not Computational Procedures 153 PART III: COMPUTER EPISTEMOLOGY 8. Program Verification: The Very Idea 183 9. Philosophical Aspects of Program Verification 221 10. Philosophy and Computer Science: Reflections on the Program Verification Debate 247 viii CONTENTS EPILOGUE 11. Computer Reliability and Public Policy: Limits of Knowledge of Computer-Based Systems 271 Index of Names 309 Index of Subjects 313 SERIES PREFACE This series will include monographs and collections of studies devoted to the investigation and exploration of knowledge, information, and data-process sing sys tems of all kinds, no matter whether human, (other) animal, or machine. Its scope is intended to span the full range of interests from classical problems in the philoso phy of mind and philosophical psychology through issues in cognitive psychology and sociobiology (concerning the mental capabilitiess of other species) to ideas related to artificial intelligence and to computer science. While primary emphasis will be placed upon theoretical, conceptual, and epistemological aspects of these problems and domains, empirical, experimental, and methodological studies will also appear from time to time. The domination of cognitive science by the computational conception of mind has been so complete that some theoreticians have even gone so far as to define "cognitive science" as the study of computational models of mind. The arguments presented in this volume clearly demonstate-definitively, in my judgment-that that paradigm is indefensible, not just in its details, but in its general conception. Minds are not digital machines and people are not computers. The conception of minds as semiotic systems appears to explain the phenomena--of consciousness, cognition, and thought-far more adequately than its predecessor ever could. The scope and limits of digital machines are delineated here. Others must judge if the time has come for a new paradigm of the mind. J. H. F. FOREWORD Contemporary research in artificial intelligence and cognitive science has been domi nated by the conception that minds either are computers or at least operate on the basis of the same principles that govern digital machines. This view has been elaborated and defended by many familarfigures within these fields, includingAUen Newell and Herbert Simon, John Haugeland, Jerry Fodor and Zenon Pylyshyn, Margaret Boden, Daniel Dennett, David J. Chalmers, Philip Johnson-Laird, Steven Pinker and scores unnamed, who have committed themselves to its defense. If general agreement were the measure of truth, it would be difficult to deny that computationalism, thus understood, appears to be true. Indeed, even most of those who disagree with the computational conception tend to accept some of its elements. The alternative known as connectionism, for example, which has been endorsed by David Rumelhart, Paul Churchland, and Paul Smolensky, among others, retains the conception of cognition as computation across representa tions, even while it rejects the contention that computationalism has properly under stood the nature of representations themselves. The difference between them turns out to be that connectionists hold that representations are distributed as patterns of activa tion across sets of neural nodes, where cognition is now taken as computation across distributed representations. Even granting the difference between nodular and distributed forms of representa tions, computationalists and connectionists can agree on many aspects of cognition. The classic formulation of the computational conception, for example, has been ad vanced by John Haugeland, who has suggested that thinking is reasoning, that reason ing is reckoning, that reckoning is computation, that computation is cognition, and that the boundaries ofc omputability define the boundaries oft hought. As long as the differ ences between nodular and distributed representations are acknowledged, there are no reasons-no evident reasons, at least-why a connectionist should not adopt these computationalist contentions. xii FOREWORD In order to establish whether minds are or operate on the basis of the same prin ciples that govern computing machines, however, it is necessary to accomplish three tasks. First, discover the principles that govern computing machines. Second, discover the principles that govern human minds. And, third, compare them to as certain whether they are similar or the same. That much should be obvious. But while leading computationalists have shown considerable ingenuity in elaborating and defending the conception of minds as computers, they have not always been attentive to the study of thought processes themselves. Their underlying attitude has been that no theoretical alternative is possible. Part I: Semiotic Systems. The essays collected here are intended to demonstrate that this attitude is no longer justified. The conception of minds as semiotic (or "sign-using") systems should lay this myth to rest, once and for all. According to this approach, which builds on the theory of signs advanced by Charles S. Peirce (1839-1914), minds are the kinds of things that are able to use signs, where signs are things that stand for other things (in some respect or other). Since there are at least three different ways in which things can stand for other things, there tum out to be at least three different kinds of minds, which use different kinds of signs. "Semiotic systems" as systems that can use signs ("minds"), of course, differ from "semiotic systems" as systems of signs ofk inds that semiotic systems can use ("signs"). The meaning of a sign for a system is understood on the basis of its effects on (actual or potential) behavior, which requires a dispositional conception of meaning that is non-behavioristic, non-extensional, and non-reductionistic. The semiotic conception thus satisfies Fodor's condition that a theory of cognition ought to connect the intensional properties of mental states with their causal influence on behavior. It would be a blunder to dismiss the semiotic account merely on the basis of mistaken preconceptions about dispositions. Thus, the three essays that appear in Part I provide an introduction to the theory of semiotic systems that includes a discussion of the theory of signs and of the kinds of minds that are distinguishable on that basis. Since there are at least three kinds ofs igns-namely, iconic, indexical, and symbolic-which have the ability to use things as signs on the basis of their relations of resemblance, of cause-or-effect, or of habitual association to other things, there also appear to be at least three kinds of minds-namely, iconic, indexical, and symbolic-which have the ability to use
The list of books you might like
