Explaining the Brain This page intentionally left blank Explaining the Brain Mechanisms and the Mosaic Unity of Neuroscience Carl F. Craver CLARENDON PRESS · OXFORD 1 Great Clarendon Street, Oxford ox2 6dp Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries Published in the United States by Oxford University Press Inc., New York © Craver 2007 The moral rights of the authors have been asserted Database right Oxford University Press (maker) First published 2007 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this book in any other binding or cover and you must impose the same condition on any acquirer British Library Cataloguing in Publication Data Data available Library of Congress Cataloging in Publication Data Craver, Carl F. Explaining the brain : mechanisms and the mosaic unity of neuroscience / Carl F. Craver. p. ; cm. Includes bibliographical references and index. ISBN–13: 978–0–19–929931–7 (alk. paper) ISBN–10: 0–19–929931–5 (alk. paper) 1. Neurosciences—Philosophy. 2. Brain—Philosophy. I. Title. [DNLM: 1. Neurosciences. 2.Philosophy.WL100 C898e 2007] QP356.C73 2007 612.8—dc22 2006103230 Typeset by Laserwords Private Limited, Chennai, India Printed in Great Britain on acid-free paper by Biddles Ltd., King’s Lynn, Norfolk ISBN 978–0–19–9299317 10987654321 For Pamela This page intentionally left blank Preface There are neurophilosophers, and there are philosophers of neuroscience. Neurophilosophers use findings from neuroscience to address traditional philosophical puzzles about the mind. Philosophers of neuroscience study neuroscience to address philosophical puzzles about the nature of science. Philosophers of neuroscience are interested in neuroscience because it has distinctive goals, methods, techniques, and theoretical commitments. In this book, I propose a unified framework for the philosophy of neu- roscience. Because neuroscience is like other special sciences in many respects, this framework contains lessons for the philosophy of science generally. I develop this framework by addressing the following question: what is required of an adequate explanation in neuroscience? Debates frequently arise among neuroscientists and philosophers about whether a proposed explanation for a given phenomenon is, in fact, the correct explanation. Does Long-Term Potentiation (LTP) explain episodic memory? Do size differences in hypothalamic nuclei explain differences in sexual preference? Does the deposition of beta-amyloid plaques in the hippocampus explain memory deficits in Alzheimer’s disease? Do 40 Hz oscillations in the cortex explain feature-binding in phenomenal consciousness? While the answers to these questions depend in part on specific details about these diverse phenomena, they also depend on widely accepted though largely implicit standards for determining when explanations succeed and when they fail. My goal is to make those standards explicit and, more importantly, to show that they derive from a systematic and widespread view about what explanations are, namely, that explanations in neuroscience describe mechanisms. My project is both descriptive and normative. My descriptive goal is to characterize the mechanistic explanations in contemporary neuroscience and the standards by which neuroscientists evaluate them. This cannot be accomplished without attention to the details of actual neuroscience. I illustrate my descriptive claims with case studies from the recent history of neuroscience. For neuroscientists, I present enough detail to make the viii preface philosophical views concrete. For philosophers, I limit myself to the details required to demonstrate that the view corresponds to real neuroscience. This descriptive goal helps to keep the philosophical discussion targeted on issues relevant to the neuroscientists building the explanations. The goal of searching for mechanistic explanations is now woven through the fabric of neuroscience: it is taught through examples in classrooms and textbooks; it is propagated in introductions, discussion sections, and book chapters; and it is enforced through peer review, promotion, funding, and professional honors. To understand contemporary neuroscience, one has to understand this form of explanation. A second reason to pursue this descriptive project is that questions often arise about the adequacy of widely accepted strategies of explanation in neuroscience (see, for example, Uttal 2001; Bennett and Hacker 2003). We can address the question of whether the norms of neuroscience are justified only when we have an idea of what the norms are and of how they can be defended. The descriptive project, in other words, is the first step in a normative project: to clarify the distinction between good explanations and bad. As the body of neuroscience research continues to expand, it is worth pausing periodically to reflect on the goals of explanation and on the standards by which explanations should be evaluated. Similarly, as neurophilosophers learn more about neuroscience and seek to apply neuroscientific explana- tions to philosophical problems, they also need to learn to reflect critically on the standards for evaluating the explanations that they adopt. Here the philosopher of neuroscience can help. They can use the long tradition of philosophical literature about the nature of scientific explanation (see, e.g., Salmon 1989) to reveal crucial features of explanation in neuroscience specifically, and they can use neuroscience to reveal previously unrecog- nized features of explanation across the sciences (or at least the special sciences) generally. The relation between the descriptive and normative projects is complex, however. One cannot simply read off the norms of explanation in neuro- science from a description of what neuroscientists actually do when they form and evaluate explanations. Neuroscientists sometimes make mistakes. They sometimes disagree about whether a proposed explanation is adequate and even about what it would take to show that it is adequate. Explanatory standards change over time, and it is possible that the standards endorsed preface ix now might some day be rejected as inadequate. What role, then, can descriptions of explanations play in the search for norms of explanation? First, even if scientists often disagree about particular explanations, there are nonetheless clear-cut and uncontroversial examples of successful and failed explanations. Almost everyone (among scientists and philosophers) can agree that action potentials are explained by ionic fluxes, that some forms of neurotransmitter release are explained by calcium concentrations in the axon terminal, and that protein sequences are explained, in part, by DNA sequences. And almost everyone (scientists and philosophers) can agree that memory is not explained by the vibration of vital fluids through the cerebral ventricles, that the shape of a person’s skull does not explain their artistic talents, and that memory loss does not explain the deposition of beta amyloid in the cortex. Philosophical analyses of explanation should deliver the correct verdicts on these clear and uncontroversial examples unless there is compelling reason to suspect that the judgments of science are wrong. It is open to deny my verdict on these standard examples and to abandon widely accepted scientific ideas about what does and does not count as an explanation, but only at the risk of stretching the term ‘‘explanation’’ so far that it no longer looks at all like the scientific phenomenon that we are trying to characterize in the first place. Of course, people disagree about problem cases, but disagreement need not prevent one from using the agreed-upon examples as touchstones in formalizing an adequate account of explanation. The controversial cases can then be decided according to the account that best accommodates the central and uncontroversial cases. I argue (in Chapter 2) that many of the standard accounts of explanation in the philosophy of science fail to accommodate even the central and widely held examples of successful and failed explanation in science. In contrast, my account accommodates them directly. Second, in neuroscience, and in other sciences as well, explanations are not developed merely for the explainer’s intellectual satisfaction?—the ineffable ‘‘a ha’’ feeling that comes with understanding something. Such emotions and feelings are terrible indicators of how well someone under- stands something (see Keil and Wilson 2000; Trout 2002). Explanations in neuroscience are frequently developed with an eye