Controversies and the Metaphysics of Mind

Controversies and the Metaphysics of Mind Controversies (CVS) Controversies includes studies in the theory of controv...

Author: Dr. Yaron M. Senderowicz

43 downloads 1084 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Controversies and the Metaphysics of Mind

Controversies (CVS) Controversies includes studies in the theory of controversy or any of its salient aspects, studies of the history of controversy forms and their evolution, casestudies of particular historical or current controversies in any field or period, edited collections of documents of a given controversy or a family of related controversies, and other controversy-focused books. The series also acts as a forum for ‘agenda-setting’ debates, where prominent discussants of current controversial issues take part. Since controversy involves necessarily dialogue, manuscripts focusing exclusively on one position will not be considered.

Editor Marcelo Dascal

Tel Aviv University

Advisory Board Harry Collins

University of Cardiff

Frans H. van Eemeren

University of Amsterdam

Gerd Fritz

University of Giessen

Fernando Gil †

Ecole des Hautes Etudes en Sciences Sociales, Paris

Thomas Gloning

University of Giessen

Alan G. Gross

University of Minnesota

Geoffrey Lloyd

Cambridge University

Volume 8 Controversies and the Metaphysics of Mind by Yaron M. Senderowicz

Kuno Lorenz

University of Saarbrücken

Everett Mendelssohn Harvard University

Quintín Racionero UNED, Madrid

Yaron Senderowicz Tel Aviv University

Stephen Toulmin†

University of Southern California

Ruth Wodak

University of Lancaster

Controversies and the Metaphysics of Mind Yaron M. Senderowicz Tel Aviv University

John Benjamins Publishing Company Amsterdam / Philadelphia

8

TM

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

Library of Congress Cataloging-in-Publication Data Senderowicz, Yaron M. Controversies and the metaphysics of mind / Yaron M. Senderowicz. p. cm. (Controversies, issn 1574-1583 ; v. 8) Includes bibliographical references and index. 1. Philosophy of mind. 2. Metaphysics. I. Title. BD418.3.S46 2010 128’.2--dc22 isbn 978 90 272 1888 9 (Hb ; alk. paper) isbn 978 90 272 8791 5 (Eb)

2010019176

© 2010 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

To the memory of Henry Chaim Ehrenberg Opatov, Poland 1917 – Knittlingen, Germany 2009

Table of contents Acknowledgments Introduction 1. Introduction 1 2. Rediscovering old truths in a new idiom 3 3. Metaphysics and metaphysical controversies 5 4. The Kantian approach to the problem 6 5. Metaphysical controversies and relevant alternatives 8 6. The structure of the book 11

xi 1

part one. Outline of a theory of metaphysical controversies chapter 1 The Kantian framework 1. Introduction 17 2. Metaphysics and Kant’s image of science 19 3. The required revisions 20 4. The modified framework 22 5. The first antinomy: The explicit argument 27 6. The first antinomy: The implicit argument 29 chapter 2 The idea of controversy and metaphysics 1. The limits of Kant’s theory 35 2. The nature of controversies as polemical dialogues 37 3. Presumptions and the burden of proof 40 4. Radically opposed views and chains of controversies 45 5. The addressee 50 6. Persuasion and the rational support of reasons 51 7. Epistemic function 52

17

35

 Controversies and the Metaphysics of Mind

chapter 3 Intuitions, thought experiments, and controversies 1. Introduction 59 2. Intuitions and beliefs 60 3. The fallibility of intuitions 63 4. The emendation of intuitions 64 5. Intuitions and the role of imagined ideal cases 66 6. Objectual imagining, propositional imagining, and imagining ideal cases 69 7. Imagined ideal cases and perception 70 8. Logical matters and “matters of fact” 72 9. Types of responses to conflicts of intuitions 74 10. Controversial relevant alternatives 75 11. Conclusion 77

59

part two. The knowledge argument chapter 4 The polemical character of the knowledge arguments 1. Introduction 81 2. Kripke’s modal argument for dualism as a polemical argument 82 3. Nagel’s argument 88 4. Jackson’s knowledge argument as a polemical argument 91

81

chapter 5 The antinomies of consciousness and their resolutions: Phenomenal concepts, representationalism, and two-dimensional semantics 97 1. Churchland’s response to the knowledge argument 97 2. Jackson’s response to Churchland 100 3. The causal efficacy of conscious qualitative states 103 4. The paradox of phenomenal consciousness 107 5. The phenomenal concepts strategy 109 6. Tye’s PANIC account of phenomenal consciousness 114 7. The knowledge argument and two-dimensional semantics 116 8. Mind and illusions 122 chapter 6 A priori knowledge and the explanatory gap 1. Introduction 125 2. Levine’s argument 126

125

Table of contents 

3. Jackson: Serious metaphysics and conceptual analysis 131 3.1 Entry by entailment 131 3.2 Conceptual analysis 133 4. Chalmers: Supervenience and scientific explanations 138 5. Block & Stalnaker’s attack 140 5.1 The arguments against the epistemic version 141 5.2 The apriority of reductive explanations and uniqueness 143 5.3 The explanatory gap and two-dimensional semantics 146 6. Chalmers & Jackson’s response 149 7. Chalmers’s two-dimensionalism refined 153 8. Conclusion 158 part three. Personal identity and revisionary metaphysics chapter 7 Personal identity, self-consciousness, and bodily identity 1. Introduction 161 2. Strawson’s Individuals 163 3. Williams: Personal identity and individuation 169 4. Shoemaker: Self-Knowledge and Self-Identity 173 chapter 8 From transcendental arguments to revisionary metaphysics 1. Wiggins: Relative identity and personhood 181 2. Shoemaker: Quasi-memory and revisionary metaphysics 187 2.1 Wiggins on identity 188 2.2 Persons and their pasts 190 chapter 9 Neo-Lockeanism, reductionism, and animalism: The emergence of a new debate 1. Introduction 197 2. Self-identity and self-concern 199 3. Parfit and “what matters” 202 4. Perry; identity and “what matters” 205 5. Wiggins’s animalism 211 6. Conclusion 218 Conclusion References Index

161

181

197

221 225 233

Acknowledgments I began examining the role of metaphysical controversies in the history of philosophy and metaphysics when I was part of the research project “Controversies in the République des Lettres” supported by the German-Israeli foundation. I wish to thank the German-Israeli foundation for their kind support. In particular, I would like to thank my collaborators in this research project Marcelo Dascal, Gerd Fritz, and Thomas Gloning. Marcelo Dascal’s exemplary knowledge of the history of early modern philosophy and of current philosophy of science, language, and mind, and his uncompromising criticism of my early views led me to improve this book significantly. The inspiring and intensive discussions with Gerd Fritz and Thomas Gloning helped me to formulate the thoughts expressed in this book. These ongoing conversations indicate how fruitful and stimulating an interdisciplinary intellectual project can be. Noa Naaman-Zauderer read parts of this book and her thoughtful comments helped me to shape its final version. I wish to thank her for many years of intellectual friendship. I wish to thank Shai Frogel, Ofer Gal, Cristina Marras, and Tzachi Zamir for their stimulating contributions to the meetings at Tel Aviv University and at Giessen University. I wish also to thank Liora Bilski, Eli Friedlander, Gideon Freudenthal, Ido Geiger, Zoe Gutzeit, Hagi Kenaan, Michal Kirschner, Vered Lev-Kenaan, Ofra Rechter, and Assaf Weksler for helpful discussions. Finally, I wish to thank Ilona, Assi, and Noa. This book would not have been written without their love and support.

Introduction 1. Introduction Metaphysics is an ancient discipline. In the first half of the 20th century, many philosophers considered it an outmoded discipline. Overcoming metaphysics is a theme expressed in different ways by logical positivists as well as by ordinary language philosophers such as Wittgenstein, Ryle, Austin, and their followers. Yet, in the second half of the 20th century, metaphysics was revived in various branches of analytic philosophy and, in particular, in the philosophy of mind. As is well-known, the logical positivists’ anti-metaphysical stance resulted from their attempt to provide an account of meaning that, in their view, was faithful to the methods and practice of the empirical and exact sciences. Philosophy had to provide criteria of demarcation that could distinguish between “real” science and “pseudoscience,” that is, between science and metaphysics. Interestingly, the philosophers who revived metaphysics at least partly share the logical positivists’ view regarding what the tasks of philosophy were. Many of them, who happily accept the title of ‘metaphysician’ today, especially philosophers of mind, locate metaphysics in the epistemic space that includes the empirical and exact sciences, even if they seem to believe that metaphysics is a discipline that has its own subject matter, or at least its own mode of approaching it. Moreover, these rather surprising relations between science and metaphysics are in many cases the subject of disputes between metaphysicians – philosophers who combine in their philosophical reflections both metaphysical contentions and information derived from scientific theories – and those philosophers who deny that this approach to philosophy is sound. The reunification of science and metaphysics in one of the most influential branches of analytic philosophy deserves to be understood. What makes this reunification possible? Is metaphysics distinct from the empirical sciences, and in what ways? How is it possible to integrate metaphysical contentions with scientific knowledge? These questions are rarely asked by contemporary metaphysicians. The reunification is usually seen as a reasonable step in the development of analytic philosophy. Quine’s (1953) attack on the analytic-synthetic distinction and the verificationist criterion of meaning is often pointed out as a significant step towards combining metaphysical reflections and empirical inquiries. By




undermining the basis for the criteria of demarcation, Quine’s arguments seem to facilitate the reunion. Yet the same logical and epistemic ideas led some philosophers to question not only the adequacy of metaphysics, but of philosophy as a whole.1 Moreover, according to the philosophers who practice metaphysics, the rehabilitation of metaphysics was required partly due to what they regarded as the demands of an adequate semantics for natural languages, which involved reinstating the traditional modal distinctions between the necessary and the possible. Some of these philosophers, most notably Jackson (1998) and Chalmers (1996), reinstate the a priori/a posteriori distinction. Needless to say, the distinction between necessity and contingency, and that between a priori knowledge and a posteriori knowledge were also within the scope of concepts attacked by Quine. A rational discipline that blurs the boundaries between metaphysics and the empirical sciences is bound to be puzzling. By addressing issues such as the mindbody problem, personal identity, the nature of mental representations, and similar topics, its domain of study seems to overlap with that of the relevant empirical sciences. By involving conceptual considerations or modal considerations, metaphysics seemed to be closer to logic and mathematics. But “matters of fact” and conceptual subject matter seem to be mixed together in the contentions that articulate metaphysical intuitions and in the reasons that support their acceptance.2 As is clear to anyone practicing metaphysics, the reasons that support the acceptance of metaphysical contentions, say, the claim that phenomenal properties are not physical properties, or that persons are basic particulars, are not empirical reasons in the ordinary sense. Similarly, the arguments that introduce the supporting reasons, although they apparently involve modal, conceptual, and logical considerations, are not of the same type of arguments as those used in the exact sciences. In other words, the reunion of metaphysics and the sciences is indeed a puzzling phenomenon. My intention in the present book is to fulfill one part of the complex task of describing the nature of the relations between science and metaphysics in

1.

See in particular Rorty (1980).

2. For example, according to Stalnaker (2001: 635–636): “Metaphysical disputes are of course always disputes about the natures of things – disputes where questions about what the world is like interact with questions about what we are saying when we say what the world is like. The positivists famously argued that metaphysics as an enterprise is founded on equivocation between semantic and factual questions – between questions that call for a decision about how to talk and questions to be answered by science or mathematics. If we are clear and avoid equivocation, they argued, metaphysics will go away. But if there is no general and absolutely neutral way to distinguish the semantic from the substantive questions, it may not be so easy to dispense with metaphysics.”

Introduction

contemporary philosophy. I will attempt to clarify the nature of metaphysical controversies and their significance to the rational nature of the metaphysics of mind. 2. Rediscovering old truths in a new idiom In the introduction to Individuals (1959), a landmark in the revival of metaphysics in analytic philosophy, Peter Strawson presented the following clarification of the significance and goals of what he called ‘descriptive metaphysics:’ The idea of descriptive metaphysics is liable to be met with skepticism. How should it differ from what is called philosophical, or logical, or conceptual analysis? It does not differ in kind of intention, but only in scope and generality. Aiming to lay bare the most general features of our conceptual structure, it can take far less for granted than a more limited and partial conceptual inquiry. Hence, also, a certain difference in method. Up to a point, the reliance upon a close examination of the actual use of words is the best, and indeed the only sure, way in philosophy. But the discriminations we can make, and the connections we can establish, in this way, are not general enough and not far-reaching enough to meet the full metaphysical demand for understanding. For when we ask how we use this or that expression, our answers, however revealing at a certain level, are apt to assume, and not to expose, those general elements of structure which the metaphysician wants revealed. The structure he seeks does not readily display itself on the surface of language, but lies submerged. He must abandon his only sure guide when the guide cannot take him as far as he wishes to go. (1959: 9–10)

The task of descriptive metaphysics is to reveal the structures that ground the basic features of understanding and meaning. One can “take for granted” that one knows only part of what this conceptual structure involves. Reliance upon the actual use of words “is the best, and indeed the only sure, way in philosophy” (ibid.) only up to a certain point. For “the discriminations we can make, and the connection we can establish...are not far reaching enough to meet the full metaphysical demand for understanding.”(ibid.) Descriptive metaphysics does not differ “in kind or intention” from philosophical, logical, or conceptual analysis but only in “scope and generality.” Nevertheless, there is a certain “difference in method.” According to Strawson, concepts – at least some of them – are susceptible to change. But it is erroneous to suppose that metaphysics is “an instrument of conceptual change:” For there is a massive central core of human thinking which has no history – or none recorded in histories of thought; there are categories and concepts which, in their most fundamental character, change not at all. Obviously these are not the specialties of the most refined thinking. They are the commonplaces of






the least refined thinking; and are yet the indispensable core of the conceptual equipment of the most sophisticated human beings. It is with these, their interconnections, and the structure that they form, that a descriptive metaphysics will be primarily concerned. (1959: 10)

The metaphysician does not attempt to improve or revolutionize human thought and human language but rather to reveal the “commonplaces of the least refined thinking.” Strawson believes that it is unlikely that the metaphysician will be able to discover “new truths.” Nevertheless, the recovery of the “core of human thinking” is not a task that accomplishes once and for all what the metaphysician aims to realize: It has constantly to be done over again. If there are no new truths to be discovered, there are old truths to be rediscovered. For though the central subject-matter of descriptive metaphysics does not change, the critical and analytical idiom of philosophy changes constantly. Permanent relationships are described in an impermanent idiom, which reflects both the age’s climate of thought and the individual philosopher’s personal style of thinking. No philosopher understands his predecessors until he has re-thought their thought in his own contemporary terms... (1959: 10–11)

This oscillation between rediscovering the “central core of human thinking which has no history” and expressing it in “the new critical and analytic idiom that changes constantly” is both significant and puzzling. What is the method by means of which the central metaphysical core of human thinking is revealed? And what could be the basis for the assumption of the a-historical character of the metaphysical core of human language and thought, given that uncovering it “has constantly to be done over again,” since “permanent relationships are described in an impermanent idiom”? It could be assumed that the method Strawson has in mind is the method he used in his own investigations – a conceptual analysis that goes beyond the actual use of words, assisted also by philosophical thought experiments. The object of metaphysical inquiries – the conceptual structure that metaphysics aspires to reveal – is apparently not a historical entity. But metaphysical investigations are guided by the inquiries of previous philosophers. The descriptive metaphysician has to “rethink their thoughts in his own contemporary terms.” It should be noted, however, that ‘rethinking’ for Strawson is not equivalent to ‘interpreting others.’ ‘Rethinking’ apparently means to address the same conceptual structure that one’s predecessors aimed to reveal. Nevertheless, ‘rethinking’ cannot be severed from ‘interpreting others’, if the core idea that underlies the metaphysical endeavor at least includes rediscovering old truths in a new idiom. “No philosopher understands his predecessors until he has re-thought their thought in his own contemporary terms.”(ibid.) Yet, no philosopher can understand his predecessors without interpreting them. The question that naturally arises in this context is whether “rethinking” the “massive central core of human thinking which has no

Introduction

history” guided by interpreting one’s predecessors and one’s own contemporaries does indeed involve no novel components of content, as suggested by Strawson. 3. Metaphysics and metaphysical controversies Strawson’s approach to the practice and goals of metaphysics is not the only approach to metaphysics, but it does seem to express some widely accepted ideas. There is, however, one feature about which Strawson and those who share his views say almost nothing. The idea that in practicing metaphysics one aims to uncover the conceptual structure that has to be rediscovered again and again seems to imply either that there is general agreement regarding the content of this conceptual structure, or that there is a method available to all thinking beings by which it is possible to achieve general agreement. Needless to say, nothing is more remote from the present state of metaphysics. Nor can we say that this is a proper depiction of metaphysics in the past. There is not even one metaphysical issue on which there is general agreement. The disputed issues include the questions regarding the nature of truth, the nature of time, the existence of abstract entities, the nature of causality, of events and propositions, the nature of mental states and of the qualitative features of conscious mental states, the grounds that constitute personal identity, the fundamental concepts of existence and being, of the possible and the actual, and other metaphysical issues. Although metaphysicians appear to directly approach their subject matter from the depth of their own reflective capacities, in many cases, they are unable to argue for their views without arguing against a rival position. In other words, the past and present discourse of metaphysics is fraught with controversies. Understanding the nature of metaphysics cannot be disconnected from understanding the nature of metaphysical controversies and their significance to the rational and epistemic character of metaphysics. The fact that controversies are prevalent in metaphysics raises the question of whether there can be any progress in it. Indeed, it is not easy to comprehend the nature and possibility of progress in metaphysics. There are, however, some examples of significant changes in metaphysics that could be clues for understanding its nature. For example, in 1982, Frank Jackson, an eminent metaphysician, published “Epiphenomenal qualia,” in which he first presented the knowledge argument – the argument that aspired to establish the epiphenomenal character of qualia. Jackson’s argument was discussed and debated extensively. However, almost twenty years later, Jackson abandoned his earlier position.3 His new theory combines features related to a position labeled ‘representationalism’, together 3. See in particular his paper “Mind and illusion” (2004).






with Lewis’s (1983b, 1988) and Namirow’s (1980) ability hypothesis. As Jackson’s arguments (which will be discussed in Part Two) clarify, his revision of his position did not result from a blind leap of faith. As the extensive literature on the knowledge argument indicates, it would be equally mistaken to suppose that Jackson and those that followed him simply missed some good reasons that were already apparent in 1982. To be sure, these reasons were not beyond the reach of human understanding, and yet, at the time Jackson’s paper was published, they were not explicitly articulated by anyone in a way that clarifies their relevance to the problem at hand. These reasons only became available by virtue of painstaking and creative philosophical labor. It is widely agreed, I would assume, that Tye’s (1995, 2000) arguments for representationalism, Lewis’s (1983b, 1988) and Namirow’s (1980) arguments for the ability hypothesis, and Jackson’s (2004) arguments in his recent reappraisal of his earlier view are all novel philosophical contributions, even if one might disagree with what these theories assert. A close scrutiny of the development of the literature reveals two facts that need to be stated if one wishes to provide an account of the rational change that took place between these periods. Firstly, although the progress of the relevant empirical sciences formed a significant layer of the background knowledge that allowed these changes, the change itself was not generated by means of the discovery of new empirical evidence in psychology, neuroscience, or any other branch of empirical science. The arguments and counterarguments for and against the intended results of the knowledge argument and the philosophical method by which it is carried out do not directly involve the data contributed by the relevant empirical sciences. Rather, they involve puzzles that seem to go beyond what allows empirical warrant. These conceptual puzzles concern the nature of our modal concepts, meaning, conceptual analysis, scientific explanations, and scientific reduction. Secondly, the arguments that aim to warrant each of the respective competing positions are not merely based on conceptual analysis. They inherently involve criticism and denunciation of competing positions. As I hope to convince the reader, these two facts are connected to the nature of progress and changes in this field. One of my main claims here will be that the dialectical character of the metaphysical contentions can play a significant role in the reassessment of the rational nature of metaphysics, and the way in which metaphysics – together with the empirical and exact sciences – forms one unified body of knowledge. 4. The Kantian approach to the problem The claim that the pervasiveness of metaphysical controversies may serve as a clue to the discovery of the rational character of metaphysics is not novel in the

Introduction

modern history of metaphysics. One of Kant’s main goals in the Critique of Pure Reason and in related writings was to reveal the connection between metaphysical controversies and the rational and epistemic characteristics of this discipline. In Kant’s view, the prevalence of metaphysical controversies raises the question of whether it is possible for metaphysics to be a science. As is well-known, Kant believed that metaphysics, if it is possible, is a synthetic a priori science. As clarified by the arguments included in the “Dialectic of Pure Reason,” when one attempts to make objective judgments merely on the basis of the conceptual resources of pure reason, one is bound to be entangled in unresolved conflicts. Pure synthetic a priori judgments are possible only within the limits of possible experience. But Kant’s critique of metaphysics does not end with this contention. Its most interesting feature probably consists in the explanation of the role of pure reason in experience. In Kant’s view, a transcendental use of the ideas of pure reason and the concepts of the understanding, which transgresses the limits of possible experience, is unavoidable. This type of use leads to transcendental illusions that are rooted in our rational capacities. Although these illusions are unavoidable, they can be cured by means of a critique that assigns the proper role to each representation. As Kant’s arguments clarify, his transcendental idealism – the ontological position developed throughout his critical writings – is necessarily required for any possible resolution of the conflicts within reason. Though they generate transcendental illusions, the ideas of pure reason – the concepts of objects of reason (objects that we cannot know) – are not “mere figments of the brain.” (CPR B 371) Pure reason has a necessary but regulative role in the gappy continuum of objective knowledge due to the demand for a total and all-inclusive explanation of observed phenomena. Without using the concepts of reason, the experience of objects qua rule-governed enterprise would not be possible. And yet, the role of the concepts of reason differs from the role of pure and empirical intuitions as well as from the role of the pure and empirical concepts of the understanding. In examining the connection between metaphysical controversies and the rational character of metaphysics, my approach to the role of controversies in metaphysics is inspired by Kant’s theory. Following Kant, I believe that the fact that metaphysical controversies are no less prevalent today than in Kant’s time is important in order to reveal the rational nature of metaphysics. However, I distinguish between Kant’s explicit position regarding the role and nature of metaphysical controversies and his implicit position, which is revealed by the arguments that can be reconstructed from his writings. One type of argument that will be examined in the present book is connected to Kant’s antinomies and the solution he offers for them. An antinomy consists of two conflicting contentions, each of






which can be traced to a plausible metaphysical intuition.4 The arguments that support each of the conflicting contentions inherently involve the refutation of the opposed contention. Kant’s account of the antinomies uncovers the conflicts and their rational sources. But it also includes an argument for transcendental idealism. Kant justifies the acceptance of his novel ontological position also by claiming that a resolution of the antinomies is possible on the basis of his transcendental idealism. Interestingly, Kant does not present this argument as a metaphysical argument. Rather, he places it outside his discussion of the nature and possibility of metaphysics. Yet my claim is that the overall argument that begins by revealing the antinomies and then proceeds by arguing for a novel metaphysical position exemplifies a significant type of argument that is widely used in current metaphysics. This type of argument will be explored throughout this book.5 5. Metaphysical controversies and relevant alternatives One of the main concepts that I coin in this book in order to clarify the nature of progress in metaphysics is the concept of a controversial relevant alternative. I suggest that establishing a position as a relevant alternative is the type of epistemic achievement that best fits the actual way in which ‘progress’ is grasped by metaphysicians. Recognizing a position as a relevant alternative does not entail maintaining that the position is true. Rather, it usually means that the position is recognized as a plausible competitor in the field, that is, as a position supported by persuasive arguments. In addition, it means that any attempt to establish a position 4. In the present context, by metaphysical intuition I do not mean intuition in Kant’s technical sense, but rather ‘intuition’ in the sense that is prevalent in the current metaphysical discourse. See, for example, Jackson (1998). I discuss this concept in Chapter 3. 5. Why didn’t Kant explicitly discuss the features of the overall argument that he himself used in his critical writings? I do not wish to engage in speculative exegesis regarding his underlying intentions. Let us merely note that viewing the implicit argument as a metaphysical argument would have placed Kant’s theory within the dialectical process and not outside it. Kant believed that the antinomies of pure reason originate from a-temporal, abstract structures of pure reason. He believed that reason itself could not be dialectical in its deepest level. In the Critique of Pure Reason, he wanted to end the conflicts of reasons, that is, to bring reason to the state of “eternal peace”. Placing himself within the dialectical process apparently conflicts with this goal. Needless to say, metaphysical and ontological controversies did not end after the publication of Kant’s critical writings. Already in his lifetime, his own metaphysics and ontology were furiously debated. Yet, philosophers who refused to accept Kant’s novel ontology were nevertheless motivated by his arguments to develop their own positions. Kant’s arguments persuaded them to consider his theory as a relevant alternative to their own theories even though it was unacceptable to them.

Introduction

that competes with the relevant alternative has to address the reasons involved in the arguments that support the relevant alternative. A position is established as a relevant alternative not merely by virtue of its characteristic contentions, but rather by virtue of the arguments that support it. These arguments also usually aim to clarify why, given the shared background of the competing positions, the position is more plausible than its competitors. In other words, the arguments that establish positions as relevant alternatives are inherently dialectical arguments. I will discuss the concept of a relevant alternative in Chapter 3.6 I suggest that the following steps form a structure which clarifies the rational role of controversies in the epistemic changes in metaphysics: a. A metaphysical controversy emerges. The controversy has the following features: (1) proponents of each of the conflicting positions persuasively argue for contentions that articulate metaphysical intuitions emphasized by the position that they support; (2) the arguments that support each of the conflicting contentions inherently involve reasons that apparently refute the contentions supported by proponents of the competing position; and (3) given the state of knowledge within which the controversy emerges, it is not possible to rationally resolve the conflict by favoring one of the positions involved. b. Philosophers (the contenders themselves or other philosophers) present arguments that attempt to resolve the controversy by including various types of conceptual innovations. These arguments and the reasons used are motivated by the paradox or the aporia revealed by the controversy. The attempts to resolve the conflict either reinforce one of the old positions (by presenting a new version of it), or introduce new positions. The successful resolutions of the conflict, which are widely recognized as persuasive, establish (or reestablish) positions as relevant alternatives to existing positions. c. A new controversy emerges which involves the relevant alternatives that are introduced in stage (b) above. Kant’s implicit argument in “The Antinomy of Pure Reason” is a metaphysical argument that has the features of stages (a)–(b). However, his argument generated several metaphysical controversies in which he himself was involved either directly or indirectly, in conformity with stage (c).7

6. Clearly, a theory could also be considered important enough to require a response because one believes that it is bound to have “a bad influence,” or that it contains mistakes that are bound to tempt one to stride in hopeless routes of thought. But this is not the only possible response to a relevant alternative. The more interesting response inherently involves genuine metaphysical innovations. 7.

On this subject, see Allison (1973); Beiser (1987); Saner (1973); and Senderowicz (1998).






Probably not all metaphysical controversies have the above structure. Nevertheless, as I hope to show in Parts Two and Three of this book, the above structure can be discerned in some of the most interesting and stimulating controversies in contemporary metaphysics of mind. It might be argued that this structure is in fact a Hegelian dialectic in disguise. But this supposition is mistaken. Specifically, controversies that reach antinomic states do not impose their resolution, and in particular, they allow for more than one resolution. Controversies that reach an antinomic state may motivate attempts to resolve them. These attempts may share the recognition of the controversy’s underlying antinomic state and the fact that each position involved is a relevant alternative. But even this is not a necessary feature of metaphysical controversies. Normally, there are varied and conflicting ways of responding to a metaphysical controversy that, among other things, depend on the way the controversy is interpreted, and (if it is interpreted as an antinomy) on how the antinomy is interpreted. One of the interesting results of the theory developed in this book is that progress in metaphysics cannot be represented as a change from stages in which some conflicting positions are discernable to new stages in which novel positions supersede the former positions. The arguments that establish the acceptance of controversial relevant alternatives seem to be much more significant to the development of metaphysics. As noted above, a position is recognized as a relevant alternative not merely by virtue of the contentions that characterize it. Analytic philosophers in the 1940s and 1950s regarded Cartesianism, for example, as relevant to philosophical theories of the mind merely as the straw man that served to point out what was wrong with earlier theories. They did not view it as posing a challenge that raised intriguing puzzles conceived as relevant to the tasks the theories themselves were meant to accomplish. But Cartesianism reemerged as a powerful and intriguing relevant alternative to physicalist theories of the mind in the 1970s and 1980s. To be sure, this did not happen by virtue of Descartes’s original arguments. It became a relevant alternative to the contemporaneous positions by virtue of original arguments presented by some of the finest philosophers in the last four decades, some of which will be explored in Part Two of this book. The historical fate of Cartesianism or of Aristotelianism is a clear indication of the fact that positions need not be eliminated from metaphysical discourse when a competing alternative is recognized by most experts as a position that has better rational support. In contrast to the nature of scientific theories in the empirical and exact sciences, metaphysical positions may reappear time and again, not merely due to blind leaps of faith but because of reasons introduced by arguments that establish these particular metaphysical positions as relevant alternatives to the existing positions in particular discursive contexts. This inordinate

Introduction

history of appearance and reappearance of metaphysical positions has a rational character that at least partly consists in the polemical arguments that reintroduce them qua relevant alternatives. In other words, the intellectual history of the metaphysics of mind should not merely be depicted as a history of positions replaced by opposite positions, but rather, at least partly, as a history of their status as relevant alternatives. This history is constituted by dialectical arguments that establish certain positions as relevant alternatives at a given historical period and that undermine their power as such at latter times. 6. The structure of the book The book is divided into three parts. In the three chapters of Part One, I will present a theoretical framework that clarifies the rational and epistemic role of controversies in metaphysics. The framework consists of a modified Kantian account of metaphysics combined with issues pertaining to the pragmatics of controversies. In Chapter 1, I will develop a modified Kantian account of ampliative arguments in metaphysics. I will begin by pointing out the intricate features of Kant’s image of science and the place of metaphysics within it. I will then analyze Kant’s explicit argument in the “The Antinomy of Pure Reason, First Conflict of the Transcendental Ideas” and what I consider to be his implicit argument. Kant’s explicit and implicit views regarding the rational and epistemic role of metaphysical controversies in metaphysics leave out the features of real polemics. In Chapter 2, I will present some themes of the pragmatics of controversies that are relevant to controversies in the metaphysics of mind. I will analyze the features of the central dialectical argument – which derives from Kant’s notion of antinomy – which will be examined throughout the book. In Chapter 3, I will discuss the role of metaphysical intuitions, and of conflicting metaphysical intuitions in metaphysical controversies and in metaphysics. Given the fundamental role of the appeal to intuitions, the idea of conflicting intuitions seems to jeopardize metaphysics by opening it up to the hazards of ontological and metaphysical relativism. In this chapter, I will suggest a way out by distinguishing between shared rational intuitions and cognitive dispositions that vary with the cognitive differences between individuals. I will clarify the sense in which conflicting responses to possibilities are motivated by the incomplete and indeterminate character of our concepts as expressed in language. I will address the role of the unique type of thought experiments involved in metaphysical inquiries, and will discuss one of the distinctive features of metaphysical intuitions, that is, the merging of “matters of fact” and logical matter. In the final






section of this chapter, I will discuss the concept ‘relevant alternative’ which underlies my account of progress in the metaphysics of mind. In Parts Two and Three, I will apply the theory developed in Part One to two examples of chains of controversies in current philosophy of mind. In Part Two, I will interpret a chain of controversies regarding the nature of consciousness that are motivated by Nagel’s and Jackson’s knowledge arguments and by Kripke’s modal argument. The epistemic changes related to this chain of controversies exemplify the strategy of descriptive metaphysics.8 In Chapter 4, I will interpret Kripke’s modal argument and Jackson’s knowledge argument as dialectical arguments of the type discussed in Chapter 2. In Chapter 5, I will address some of the polemical exchanges that refer to Jackson’s argument and to Kripke’s argument. These exchanges motivated interpreting these controversies as expressing antinomies (in the sense discussed in the first three chapters). The resolutions of these antinomies generated other controversies that widened the scope of topics addressed and added issues pertaining to logic and modality, scientific explanation, and related issues. These implications of the knowledge arguments and Kripke’s modal argument will be discussed in Chapter 6 where I will address the debates on the explanatory gap, on two-dimensional semantics, and on the concept of scientific explanation that evolved from the debates on the nature of consciousness. In Part Three, I will examine another chain of controversies related to the question of the connection between personal identity, self-consciousness, and bodily-continuity. I suggest that the epistemic changes motivated by these polemical exchanges constitute a revisionist approach to metaphysics, which differs from the descriptive approach. The debates to be addressed have motivated novel types of accounts of personal identity that transformed the questions addressed by metaphysicians. Whereas the early debates were between adherents of accounts of personal identity based on criteria of bodily-continuity (Williams), those that based such accounts on criteria of psychological continuity (Lockeans), and those that viewed persons as basic particulars (Strawson), the current debate is between the neo-Lockeans (whose views differ significantly from their predecessors’ views), reductionists (Parfit and his followers), and animalists (Wiggins, McDowell, Olson, and others). In Chapter 7, I will examine the positions of three philosophers who first started the debates on personal identity in analytic philosophy: Strawson, Williams, and Shoemaker. Particular attention will be given to Shoemaker’s 8. In the sense to be used here, descriptive metaphysics consists in the attempt to resolve metaphysical conflicts by constructing concepts and theories assumed all along to be implicit in our ordinary conception; that is, making these concepts explicit is at least partly motivated by the relevant metaphysical controversies.

Introduction

Brown-Brownson (body exchange) thought experiment, and the way in which it influenced the debates on personal identity and bodily continuity. In Chapter 8, I will examine the exchange between Wiggins and Shoemaker on the identity of persons. I will show how Shoemaker’s earlier position could be interpreted as involving the notion of ‘relative identity’ that is attacked by Wiggins, and how his revised account aims to avoid the commitment to relative identity. This criticism was one of the grounds that motivated Shoemaker’s shift from quasi-transcendentalism to his neo-Lockean physicalist-functionalist account of the identity of persons. Shoemaker’s novel account introduced the concept of ‘quasimemory’ into the debate. This is a concept of a memory-like state the existence of which does not imply personal identity, and that therefore seems to make a noncircular definition of personal identity in terms of such states possible. This concept, the content of which apparently conflicts with our ordinary grasp of personal identity, had a significant role in setting a new agenda for the debates on personal identity. In Chapter 9, I will clarify how the new stage in this chain of debates evolved from the early stages. I will first examine Williams’s puzzle of personal identity as presented in his paper “The Self and the Future.” (1970). This puzzle has the features of the modified Kantian antinomy. I will proceed by interpreting Parfit’s reductionist approach to persons and personal identity in “Personal Identity” (1971) as an attempt to resolve this puzzle. I will then show how Perry’s (1972, 1975, 1976) and Lewis’s (1976) neo-Lockean modifications of Shoemaker’s definition of identity and of the continued existence of persons can be viewed as a polemical response, both to Shoemaker and to Parfit’s arguments. I will conclude this chapter by examining Wiggins’s arguments in Chapter 6 of the first edition of Sameness and Substance, which aim to clarify why his animalistic position should be preferred over the neo-Lockean position and the bodily-continuity position. Taken together, I hope that the three parts of the book clarify the role of controversies in the metaphysics of mind. Before I embark on the tasks set out in this introduction, may I add one last preliminary word. In this book, I have chosen to focus on controversies and the metaphysics of mind, and on a particular type of argument that is significant in facilitating an understanding of the significance of controversies in the metaphysics of mind. I wish to stress, however, that this by no means entails that controversies are significant only to the metaphysics of mind. I hope that what I will present in the following chapters will at least indicate how the type of analysis conducted here can be extended to other branches of metaphysics.



part one

Outline of a theory of metaphysical controversies

chapter 1

The Kantian framework 1. Introduction Throughout his career, Kant’s account of the nature and possibility of scientific and ordinary knowledge was linked to his account of the possibility of metaphysical knowledge. In the Prolegomena, Kant declared that Hume’s doubts regarding the knowledge of causal relations arose him from his (pre-critical) dogmatic slumber. Yet, in a 1798 letter to Christian Grave, six years before his death, Kant writes that it was the Antinomy of Pure Reason that “first arose me from my dogmatic slumber and drove me to the critique of reason itself in order to resolve the ostensible contradiction of reason with itself.” (Br 12: 258; 552) Kant’s antimonies articulate abstract conflicts of pure reason, and yet they can easily be traced to real polemics. As he expresses in the question that he raises regarding the possibility of metaphysics as a science in the passage below, Kant attributes considerable significance to the fact that metaphysical contentions do not enjoy universal and lasting acclaim: If metaphysics is a science, why is it that it cannot, like other sciences, attain universal and lasting acclaim? If it is not, how does it happen that, under the pretense of a science it incessantly shows off, and strings along the human understanding with hopes that never dim but are never fulfilled? Whether, therefore, we demonstrate our knowledge or our ignorance, for once we must arrive at something certain concerning the nature of this self-proclaimed science; for things cannot possibly remain on their present footing. It seems almost laughable that, while every other science makes continuous progress, metaphysics, which desires to be wisdom itself, and which everyone consults as an oracle, perpetually turns round on the same spot without coming a step further. Furthermore, it has lost a great many of its adherents, and one does not find that those who feel strong enough to shine in other sciences wish to risk their reputations in this one, where anyone, usually ignorant in all other things, lays claim to a decisive opinion, since in this region there are in fact still no reliable weights and measures with which to distinguish profundity from shallow babble. (4: 256)

The failure to obtain universal and lasting acclaim is contrasted with “hopes that never dim but are never fulfilled.” In Kant’s view, the tantalizing oscillation between hopes that motivate one to pursue metaphysical knowledge, and the




recurring failures to achieve such knowledge does not amount to a history that manifests “real” progress. Metaphysics has obtained nothing that is “universal and lasting.” Yet, in his view, metaphysics is open to genuinely new ideas, the path to which involves uncovering the nature of metaphysical controversies.1 Clearly, in Kant’s view, the rational task faced by philosophers is to overcome the recurring controversies, which was a task he believed his critical theory to have fulfilled. Yet, although his theory did not enjoy the universal and lasting recognition he aspired for, I will attempt to show that his implicit views can serve as a valuable framework for an account of the rational role of controversies in metaphysics. The term ‘controversy’ will hereafter be used as involving defeasible reasons for holding opposed views, the adequacy of which depends on the historical context. Although Kant does not explicitly acknowledge the defeasibility of metaphysical theories, this idea is implicit in his writings. The apparent “conflicts of pure reason” involve contentions supported by equally plausible arguments. The balance of reasons seems to lead to skeptical despair, and yet, by introducing his novel ontology he explains away the plausibility of these reasons, and resolves the conflict. The revolutionary shift does not render these contentions meaningless. Yet Kant’s alternative rules out the reasons that support them qua good reasons. Hence, the reasons used by the arguments are defeasible in a unique sense; they lose their hold upon us by means of a theory that undermines them qua reasons.2 The idea that metaphysical theories are defeasible seems to entail that they express possibilities similar in kind to empirical possibilities, the only difference being that they are not empirically testable or refutable.3 Yet the philosophical theory that analyzes the logical and epistemic features of metaphysics – Kant’s critical theory – was not conceived by him as being defeasible in the same sense.4 1. Kant’s controversy with Eberhard addressed Kant’s claim to novelty. See Allison (1973), Eberhard (1789a, 1789b, 1789c, 1790), Maaß (1789, 1791). On this subject, see also Senderowicz (1998). 2. In fact, Kant seems to distinguish between three distinct epistemic levels or rational perspectives: the rational perspective, which Kant dubbed ‘the dogmatic view’, that is assigned to holders of opposed views involved in a controversy; ‘the skeptical view’ from which one recognizes the insolvability of the controversy; and the critical view that aims to clarify how the rational debate could be settled, why it is unavoidable, and in what sense is it rationally valuable. The reasons involved in each of the first two rational perspectives, the dogmatic conviction that the controversy must be resolvable in one’s own favor, and the skeptical conviction that it is irresolvable are defeasible reasons, although they are not recognized as such by those entrapped in these rational perspectives. 3. See Popper (1963a, 1963b) 4. As we shall see in Chapter 3, it is indeed difficult to grasp how such a theory is defeasible. Nevertheless, the interplay between ‘defeasible’ and ‘not defeasible’ is an inherent feature of metaphysical inquiries and theories.

Chapter 1. The Kantian framework 

Kant’s explicit view appears to be that the series of attempts ceaselessly frustrated by contrary attempts is a sign not of progress, but rather of epistemic stagnation. Nevertheless, I hope to show that one can extract an implicit argument from his account that allows one to locate the notion of progress in the historical process itself. In order to reveal Kant’s implicit argument my first task will be to differentiate between those features of Kant’s position that are responsible for the supposedly ahistorical nature of metaphysics, and those features relevant to the possibility of genuine epistemic change. After stating the main themes of my reconstruction of Kant’s theory, I will single out the implicit argument by interpreting one antinomy. 2. Metaphysics and Kant’s image of science According to Kant’s theory, our a priori capacities allow one to reveal the comprehensive system of concepts and questions with which metaphysics deals at all times. These include the ontological concepts that are part of all claims to knowledge. The differences between the validity of the metaphysical claims and mathematical or empirical claims require an account of the differences between the types of use of the ontological (pure) concepts that equally belong to the various rational fields. One type of use, namely, empirical use, constitutes the conditions for the possibility of empirical knowledge. Another type of use, the transcendental use, is empty use.5 The criteria of demarcation between science and metaphysics are linked to the distinction between these types of use. The critical theory within which one purportedly solves the relevant metaphysical disputes “once and for all” involves an account of the nature of mathematical claims and empirical (scientific) claims not merely in order to uncover the features of the “good” sciences and to spell out what is “bad” in metaphysics, but also in order to set the limits with which the metaphilosophical theory is not allowed to conflict. Even if one assumes (as does Kant) that knowledge gained in the empirical sciences and in mathematics is incomplete, this incompleteness does not allow one to rationally hold views that conflict with what is known in the empirical sciences and mathematics. As noted above, the criteria of demarcation between science and metaphysics are linked to the distinction between the types of use allowed by the ontological concepts. Yet, although these criteria single out the epistemic gap between the 5. See in particular the chapter “on the ground of the distinction of all objects in general into phenomena and noumena” in Kant’s Critique of Pure Reason. On this subject, see also Senderowicz (2005)




validity of the ordinary scientific claims and the metaphysical claims, they do not lead to the elimination of metaphysics. In this respect, Kant’s image of science is notoriously complex. On the one hand, knowledge is feasible only in mathematics and the empirical sciences. On the other hand, empirical and mathematical knowledge is incomplete. The gap between what can be known and what one aspires to know is filled by illusions of knowledge and generates metaphysical controversies, but it also indicates the necessity of the regulative use of reason involved in ordinary and scientific empirical knowledge. In other words, the demarcation between science and metaphysics separates the empirical and exact sciences and metaphysics, but does not undermine the unity of the rational enterprise that involves them both. To be sure, this is not a harmonious unity. As I hope to demonstrate in the following chapters, conceiving of science and metaphysics as separated by a gap that does not undermine the unity of the epistemic endeavor that includes them both is a powerful and fruitful idea. 3. The required revisions As noted above, Kant believed that a priori grounds distinguish science from metaphysics. My suggestion is to suspend judgment on this matter. As far as the goals of the present inquiry are concerned, we cannot make a decision on this matter in advance without assuming that the controversy regarding the possibility of metaphysics has been resolved. The normative claim to a priori demarcation may therefore be replaced with the descriptive claim that a gap between science and metaphysics could be detected by spelling out the way in which new contentions are established in metaphysics and in the sciences.6 The gap may shift with the historical evolution of science. It has to be rediscovered and restated in each historical context. We must assume, however, that throughout history, the various disciplines conceived as metaphysical have been similar enough to justify their being depicted as such. Some of the features of Kant’s official position therefore need to be modified. These features are: 1. The scope of the subject matter of metaphysics; 2. The claim that there are criteria for demarcating or distinguishing between metaphysical claims to knowledge and empirical or scientific claims to knowledge; and 3. The implicit claim that metaphysics does not allow real progress. 6. I will address the alternative descriptive demarcation between science and metaphysics in what follows.

Chapter 1. The Kantian framework 

Regarding the first feature, Kant’s theory of metaphysics comprises three themes: rational psychology, rational cosmology, and rational theology. Metaphysics can be separated from logic and ontology.7 Yet, separating between the questions regarding the nature of truth, meaning, the infinite, abstract objects, the reality of relations, and similar questions, and the issues addressed by metaphysics seems to be particularly wanting when one considers the metaphysical controversies in which Kant was personally involved.8 Moreover, the fact that the questions regarding the nature of meaning and truth are involved in the disputes that Kant addresses in the dialectic is manifested in the solutions that he offers for these disputes.9 These solutions require substantial changes, debatable in their own right, in the concepts of objectivity, empirical knowledge, the nature of space and time, and all the other issues related to the doctrine of transcendental idealism.10 Finally, the 7. In his prize essay “What Real Progress Has Metaphysics Made in Germany since the Times of Leibniz and Wolf?,” Kant defines metaphysics as “the science of advancing by reason from knowledge of the sensible to knowledge of the supersensible.” (Ak. 20: 260) In contrast, “ontology (as part of metaphysics) is the science that comprises a system of all concepts and principles of understanding, but only insofar as these extend to objects given by the senses and can, therefore, be justified by experience.” (ibid.) According to this classification, ontology is part of metaphysics. Yet this characterization presumably results from the polemical intentions of this essay. For in the Metaphysics Mrongovius (from 1782–1783) ontology is characterized as the science of the properties of all things in general (Ak. 29: 784) and is, therefore, more general in scope than metaphysics. 8. The controversy between Kant and his collaborators and Eberhard and his collaborators specifically addressed the features of Kant’s theory of judgments, and in particular, Kant’s distinction between synthetic and analytic judgments. Kant believed that his theory was not controversial, but that it was, rather, part of a well-established logical theory. Yet as Kant’s controversy with Eberhard demonstrated, Kant’s theory of judgment was controversial. On this subject, see Allison (1973), Senderowicz (1998). 9. Kant’s solution to some of the metaphysical disputes addressed in the first Critique involves the account of meaning presupposed by him. The fallacy involved in the paralogism is characterized in the second edition of the first Critique as a fallacy per sophisma figurae dictionis. It presupposes that the subject term could be considered as having two meanings, as a transcendental I and as an object that could be given in intuition. See CPR B 410–412. Kant detects a similar dialectical fallacy per sophisma figurae dictionis in the case of the cosmological inferences that are involved in the antinomies of pure reason. See CPR A 499–500/B 527–528. 10. The ideas that ‘object’ has two meanings, and that a priori concepts have two types of meanings related to two types of use – transcendental meaning and empirical meaning related to transcendental use and empirical use of a pure concept – are essential to Kant’s theory as well as to the solution he offers for the metaphysical disputes that he addresses in the first critique. The two meanings in which an object should be considered, a distinction that is sometimes expressed as a distinction between objects in the logical sense and objects in the real “weighty” sense, is essential for the division between things in themselves and appearances. As I showed elsewhere, the very idea of transcendental idealism involves a distinction




questions pertaining to the nature of meaning and truth are crucial to the most central distinction in Kant’s transcendental philosophy, namely, that between analytic and synthetic judgments. This distinction cannot be disconnected from the issues with which metaphysics is concerned. Kant’s views on this matter may very well be plausible, but they are not indisputable, and the controversies concerning these topics cannot be divorced from the dispute on other metaphysical themes.11 The second issue that requires revision concerns the criteria that allegedly distinguish between metaphysical knowledge and empirical (scientific) knowledge. According to Kant, synthetic judgments involve a relation between concepts and sensible intuitions. One can have knowledge of objects and objective states of affairs only by means of judgments that subsume sensible intuition under concepts either directly or derivatively, that is, by means of the schematic ostensive exhibition of a concept “in concreto,” or by being connected to experience. According to Kant, this condition cannot be satisfied by metaphysical judgments. This theme is connected to the previous theme, and it has to be suspended together with the first theme. When one tries to account for metaphysical controversies, it is preferable not to suppose that this epistemic criterion belongs to the indisputable background of general agreement presupposed by the parties involved in the controversy. Given the realistic stance of current metaphysicians, the verificationist and anti-realist views on meaning that are linked to this criterion are controversial, to say the least. The last issue that requires revision concerns the possibility of progress in metaphysics. The previous revisions facilitate the conceivability of epistemic change in metaphysics. As I already noted, whether or not it actually provided comprehensive answers to the problems of metaphysics, Kant presented a new metaphysical account of space and time, attributed new meaning to the concept of an object, to the subjective features of experience and to many other metaphysical topics.12 4. The modified framework So far I have addressed those features of Kant’s theory that need to be amended. Below are the features of Kant’s account that need to be sustained in the proposed modified framework. These include: related to the distinction between ‘mere possibility’ and ‘real possibility.’ This distinction is based on the distinction between the transcendental meaning and the empirical meaning of the pure concepts. See Senderowicz (2005), Chapter 2. 11. I argued at length for their plausibility in Senderowicz (2005). 12. The concept of progress that is suitable for metaphysical inquiries will be discussed in detail in Chapter 3.

Chapter 1. The Kantian framework 

4. The idea that the rational demand for complete explanation connects the empirical sciences and metaphysics. 5. The differences between the methodologies and grounds of evidence related to the exact and empirical sciences and to metaphysics. 6. The view that coherentism is the type of epistemic justification that fits metaphysical claims. The explanatory link that connects empirical and metaphysical issues involves two ideas, one that is central to Kant’s account of experience and another that represents the “pure employment of reason.” Regarding experience, Kant makes two distinct claims. Experiences of objects must include causal relations. Knowledge of causal relations is limited to the application of concepts to indefinitely extendable temporal series of sensuous objects, events, and processes. Each event causes other events, and in turn has a cause. The occurrence of the cause explains the occurrence of the effect, and is itself explained by the occurrence of its own cause. Yet, in Kant’s view, that knowledge of causal relations of this type is necessary in order to have experiences of objects and objective states of affair does not entail that a comprehensive explanation of the occurrence of an event is feasible.13 In Kant’s theory, the notion of a comprehensive explanation is based on the pure employment of the ideas of reason. In contrast to the understanding – the faculty of judgments – reason is the faculty of inferences. It does not unify sensible data, but rather the judgments of the understanding. Although these two conceptual capacities are relatively independent, they are not mutually exclusive or unconnected. The epistemic continuum consists of judgments that subsume objects of sense perceptions under pure and empirical concepts as well as reason qua faculty of inferences that involves the idea of complete (total) explanation. The concepts of reason cannot be exhibited in perception, in contrast to the concepts of the understanding. Yet the formation of empirical concepts, which are used by the understanding in order to make empirical objective judgments, is regulated by the rational capacities of reason. Representations that belong to the faculty of reason seem to stand for objects and objective properties, and nevertheless, knowledge of their existence is empirically undecidable.14 According to Kant’s theory, one could explain why an event E occurred by referring to its cause C. C cannot explain the occurrence of E independently of the fact that the occurrence of C itself requires explanation. But the concept of an 13. It should be noted that this rational gap does not necessarily entail the demarcating criteria between science and metaphysics. 14. To be sure, this gap is related the underdetermination argument, of which Kant was implicitly aware, as indicated by the relevant passages from the appendix to the dialectic of pure reason and the first and second introductions to the Critique of Judgment.




indefinitely extendable series of sensible events related as explanandum and explanans does not require an unconditioned explanans as one of its conditions. As Kant notes, a conditioned analytically entails some condition, not an unconditioned. (CPR A 308/B 364) The additional demand for an unconditioned explanans is based on (a) the principle of pure reason, according to which “...when the conditioned is given, then so is the whole series of conditions, subordinate to one another, which is itself an unconditioned, also given (i.e., contained in the object and its connection),” (CPR A 307–308/B 364) and (b) on the regulative use of the concepts of reason in constructing scientific theories. Features such as simplicity, overall unity, and parsimoniousness guide the formation of empirical concepts and the construction of scientific theories. Though these are clearly relevant to the content of the causal concepts involved in causal explanations, they are not concerned with the application of concepts to experiences, but rather with the way in which empirical concepts are arrived at. The account of the concepts used in ordinary and scientific judgments must involve the inferential (explanatory) links between these judgments that are inherently connected to the idea of an unconditioned, but should not undermine the relative autonomy of the understanding qua faculty through which concepts are applied to sensible data. The pure use of the concepts of reason seems to have existential implications. These are due to the link between the logical maxim of pure reason that requires one “to find the unconditioned for conditioned cognitions of the understanding, with which its unity will be completed” (CPR A 307/B364) and the principle of pure reason mentioned above. Kant believed that the link between the logical principle of pure reason and the logical maxim of pure reason is both natural and deceptive. The maxim of pure reason presupposes the principle of pure reason. But the link between the logical maxim and the logical principle of reason responsible for the alleged existential implication (the givenness of the unconditioned series) is in fact the main source for the various transcendental illusions.15 The regulative employment of the concepts of reason is unavoidable. But it must be clearly distinguished from the constitutive role of the empirical use of the concepts of the understanding.16 For Kant, this distinction delineates two distinct intellectual fields with distinct types of validity. Yet, given my previous suggestions for revision, we may sustain the epistemic distinction between the two respective layers of explanation by pointing out the distinct types of grounds of evidence connected to them without rendering the existential implications of the 15. On this subject, see Grier (2001) and Allison (2004: Chapter 11). 16. The account of how the content of empirical concepts used in scientific explanations is connected to the demand of reason for a complete unconditioned explanation is explicated in detail in the appendix to the Transcendental Dialectic.

Chapter 1. The Kantian framework 

claims involved in the higher level illusory. In other words, we may maintain the epistemic continuum of the two respective intellectual fields while acknowledging the epistemic gap that this continuum contains. The demand for knowledge of the unconditioned applies to the domain of empirical knowledge, that is, to what is given in empirical perception. But the theories and explanatory moves that are typically part of the metaphysical inquiries are separated from those of the empirical sciences.17 Being “merely” conceptual does not entail that one’s knowledge in these matters is a priori certain.18 It indicates the partial autonomy of the activity of metaphysical inquiries in the epistemic continuum that includes both metaphysics and the sciences. The objects of metaphysical inquiries need not be non-sensible. A position that attributes the ontology of sense data to the empirical sciences involves the relevant idea of the unconditioned. It is, therefore, metaphysical in Kant’s sense. The reasons that are the basis for the acceptance of such a position cannot merely consist of observational data that involve the application of concepts to sense impressions. As noted above, in Kant’s theory, a priori criteria distinguish between empirical knowledge and pretensions of knowledge of the unconditioned. Although I suggested that we need to suspend this claim, this does not entail that one cannot extract ideas from Kant’s theory that are useful for a descriptive account of metaphysics. I suggest replacing Kant’s principle of demarcation with a weaker principle that concerns the type of evidence that can be provided and the type of experiments that can be conducted in metaphysics. Although science and metaphysics are connected by an explanatory link, it does not follow that the evidence used in metaphysics is of the same kind as that used in the ordinary sciences.19 The metaphysician may be assisted by empirical evidence. Yet, empirical observations are almost never the data that warrants a metaphysical position. 17. The current metaphysical controversy regarding subjectivity and qualia is probably the most notable example in this respect. I will discuss this issue in Part Two. 18. As Popper notes (1963a: 199), although philosophical and metaphysical theories are irrefutable in the logical sense “...if we look upon a theory as a proposed solution to a set of problems, then the theory immediately lends itself to critical discussion – even if it is nonempirical and irrefutable.” 19. I am not suggesting that this is explicitly acknowledged by current metaphysicians. Many of them regard metaphysics as a proper extension of the so-called “hard” sciences. The fact that the only type of experiments used in metaphysics are metaphysical thought experiments, and that the only type of evidence are intuitions (not in the Kantian technical sense of this term) is not regarded by them as entailing the distinction between metaphysics and the sciences. But it is questionable whether the role of thought experiments in physics is the same as that in metaphysics. I address this point in Chapter 3.




Metaphysical inquiries are conceptual investigations connected to the rational demand for complete explanations of observable phenomena. These explanations inherently involve the idea of the unconditioned. But metaphysical investigations are “merely” conceptual, and the experiments used in metaphysics are thought experiments that aim to uncover one’s intuitions regarding one’s conceptual scheme.20 Regarding the justification of metaphysical propositions, the only available notion of justification is that of internal coherence: But pure reason is a sphere so separate and self-contained that we cannot touch a part without affecting all the rest. We can do nothing without first determining the position of each part and its relation to the rest; for, as our judgment within this sphere cannot be corrected by anything without, the validity and use of every part depends upon the relation in which it stands to all the rest within the domain of reason. (4: 263)

Coherentism was recently attacked by many philosophers.21 Yet given that perceptual evidence is almost never the evidence that supports the acceptance or the denial of metaphysical propositions, the only available type of justification for metaphysical propositions is internal coherence. Below are the propositions that constitute the modified version of the Kantian framework: 1. Controversies in metaphysics cannot be isolated and disconnected from disputes regarding logical issues, meaning, truth, the nature of explanation, etc. 2. Metaphysical claims should not conflict with scientific claims. They must cohere with them. 3. There are no a priori criteria for distinguishing between science and metaphysics. 4. It should not be assumed that there is no progress in metaphysics. 5. An explanatory continuum connects science and metaphysics. It affects the concept formation of the empirical concepts used in empirical explanations of scientific phenomena. 6. The type of evidence and the type of arguments that are characteristically used in metaphysics and metaphysical controversies are different from those used in the empirical sciences. 7. The canonical type of justification relevant to metaphysics is that of internal coherence.

20. I will address this issue in detail in Chapter 3. 21. See Bonjour & Sosa (2003).

Chapter 1. The Kantian framework 

5. The first antinomy: The explicit argument Kant’s official account of the role of controversies in metaphysics involves his transcendental theory. The question that I will address now is whether his relevant claims allow one to uncover an account of the role of controversies in metaphysics that does not entail his transcendental theory. By analyzing the first antinomy in the Critique of Pure Reason, I will show that such an account is indeed feasible. Making the complex structure of Kant’s argument explicit is important not merely in order to elucidate Kant’s hidden view, but also to clarify the role of metaphysical controversies in metaphysics. The first antinomy addresses the question of whether the world has a beginning in time and is limited in space. The thesis (hereafter T) is that the world has a beginning in time, and is limited in space. The antithesis (hereafter A) denies that it has a beginning in time, or is limited in space. Below is the argument for the temporal part of the thesis.22 For if one assumes that the world has no beginning in time, then up to every given point in time an eternity has elapsed, and hence an infinite series of states of things in the world, each following another, has passed away. But now the infinity of a series consists precisely in the fact that it can never be completed through a successive synthesis. Therefore an infinitely elapsed world-series is impossible, so a beginning of the world is a necessary condition of its existence; which was the first point to be proved. (CPR A 426/B 454)

The argument for the antithesis is as follows: For suppose that it has a beginning. Since the beginning is an existence preceded by a time in which the thing is not, there must be a preceding time in which the world was not, i.e., an empty time. But now no arising of any sort of thing is possible in an empty time, because no part of such a time has, in itself, prior to another part, any distinguishing condition of its existence rather than its nonexistence (whether one assumes that it comes to be of itself or through another cause). Thus many series of things may begin in the world, but the world itself cannot have any beginning, and so in past time it is infinite. (CPR A 427/B 455)

Kant’s argument implicitly addresses the question of whether the notion of a temporal series of objects and events can satisfy the following claims: (a) that actual infinity is impossible, and (b) that the coming to be of an event or an object, including the world conceived as a whole, must have a sufficient reason. The 22. The argument for the spatial part for the thesis and the antithesis depends on the argument for the respective temporal part. Since my intention is to analyze the argumentative structure of the first antinomy, and since the spatial part of the argument does not involve any feature that is not involved in the temporal part, I will not address the spatial part of the argument here.




argument presupposes that it is generally agreed that these two claims are true. Yet, although the notion of a temporal series must be compatible with both claims, Kant’s argument aims to demonstrate that (T) and (A) are compatible either with (a) or with (b), but not with both. Now, (T) and (A) argue for the claim they defend by showing that a temporal series that has the property assigned to it by the proposition that is central to the rival position, say (T), entails that one of these features (a) or (b) cannot be instantiated. In the case of (A), it is the denial of actual infinity, and in the case of (T), it is having a sufficient reason. In other words, the overall argument of the first antinomy aims to demonstrate that the above two features of a temporal series of objects and events – a temporal series the reality of which is assumed to be certain – cannot be true together of the temporal series according to any of the two conflicting positions. In this sense, Kant’s argument addresses the conflict from the point of view of an outside observer, that is, a third, apparently neutral position. Kant’s argument does not end at this point but also includes his solution for the dispute. The solution consists in two steps. The first step is to identify a general principle that also seems to be part of the background of beliefs:

(P) If the condition is given (a temporal event), the whole series of conditions must also be given.

It should be noted that this principle neither entails the thesis of the first antinomy nor the antithesis. Its role in the argument is rather to provide reasons to believe that only one of the conflicting possibilities could be true, or that only one is false. For assuming that the givenness of the conditioned (an empirical event) entails the existence of the complete series of conditions, then either the beginning of the world is entailed by any temporal event, or an infinite series of events is entailed. Yet since neither (T) nor (A) is compatible both with (a) and (b), (P) cannot cohere with the idea that both are true. The goal of the argument is to reveal the need for a conceptual change required for the coherence of the rational grounds of the concept of a temporal series. Kant’s solution to the first antinomy – indeed his solution to all the antinomies – requires P to be abandoned. The solution first identifies a link between (P) and another claim: (Q) The objects and events that are given must be things as they are in themselves. Kant assumes that (Q) is presupposed by (P). However, let us note that (P) and (Q) do not have the same epistemic status. As noted above, (P) is a principle of reason. As Kant painstakingly aims to convince the reader in the first Critique, a direct or indirect perceptual experience that warrants the acceptance of (P) is

Chapter 1. The Kantian framework 

inconceivable. On the other hand, (Q) at least appears to be based on common sense. But according to Kant, indeed according to many metaphysicians, judgments that are based on common sense or common practice are to be respected only up to a certain point. In the present context, Kant’s main idea is that (Q) could be abandoned. It could be replaced by a principle that entails its negation without being incompatible with the evidence that we have at our disposal: (R) The sensible things given to us are appearances and not things in themselves. 6. The first antinomy: The implicit argument Kant’s argument seems to address a well-defined subject. Its explicit goal is to resolve a conflict of reason within itself. Yet its additional goal, which is connected to its explicit goal, is to warrant Kant’s doctrine of Transcendental Idealism. The explicit goal is to provide a solution to a rational problem. We may assume that if the argument that establishes Kant’s solution is known to be sound, the argument does indeed solve the problem it is supposed to solve. But the argument is sound only if all of its premises are true. Yet as Kant himself attests, where metaphysical propositions are concerned, there is no epistemic ground other than coherence with the other propositions that may warrant them. What appears to provide an Archimedean point is the alleged incontestable certainty of (a) and (b). Yet, it should be noted that neither (a) nor (b) possess absolute certainty. Nor does Kant claim that they do. They are assumed to be better justified than either (P) or (Q). Kant does not argue for this claim explicitly. The argument implicitly presupposes that (a) and (b) are incontestable. Let it be noted that if the negation of (a) and the negation of (b) were impossible, there would be no room for an implicit argument. But arguments in metaphysics do not “prove” their conclusion in the mathematical or formal sense. This is central to the type of rationality that Kant attributes to this intellectual field. According to Kant, it consists of synthetic a-priori judgments, or at least of pretensions of such judgments. Metaphysical arguments are different from mathematical arguments.23 Clearly, the difference between mathematical and metaphysical arguments cannot consist in the claim that nothing can really be known in metaphysics. For this would have prevented Kant’s own position, indeed his resolution of the antinomies. There is no reason to suppose that the argument examined above, which points out the conceptual chasm associated with the 23. See in particular CPR A 712–713/B 740–741.




rational conflict and the resolution of this conflict that introduces novel distinctions and novel concepts – Kant’s doctrine of transcendental idealism – belongs to the logical-mathematical type. In order to clarify Kant’s implicit argument, one must first spell out the main ideas that underlie Kant’s approach to metaphysics.

(1) The metaphysical concepts and propositions that articulate metaphysical intuitions (not in Kant’s technical sense) that form the core of our common sense and common practice at a given historical time need not constitute a self-consistent whole.

This is, in fact, Kant’s initial step. The conceivability of conflicting metaphysical conceptions is inherent to what is pre-critically conceivable.

(2) Metaphysical controversies are the locus of the conflict between these incompatible articulations of intuitions. They uncover the fact that the set of conceivable metaphysical possibilities at a given time need not be a consistent set.

The metaphysician cannot avoid the need to respond to the positions that conflict with her own. There are various types of available responses that result from the significance that the metaphysician attributes to the opposite position.

(3) The response inherently involves one’s own representation of the metaphysical controversy.

(3) is a blind spot in the rhetoric of the argument; that is, it is a part of the argument but does not have an explicit role in it.

(4) A key feature of the way in which metaphysical controversies are represented concerns the weight attributed by proponents to each of the beliefs that are part of the background of the debated propositions.

The attribution of weight is mirrored by the arguments that Kant used. It expresses what is believed to be rationally acceptable. The attribution of the degrees of weight to the background beliefs may coincide with the opponent’s assignment of degrees of weight to the background beliefs. Yet this is rarely the case where two conflicting metaphysical positions are concerned. Clearly, the assignment of degrees of weight to metaphysical beliefs is not disconnected from what is known in the mathematical and empirical sciences. In the case of the first antinomy, the denial of actual infinity was part of the mathematical knowledge of Kant’s time, that is, part of the background knowledge. Two other implicit assumptions that are parts of Kant’s argument are:

Chapter 1. The Kantian framework

(5) The weight of a metaphysical belief is partly determined by its connection to science.

(6) The attribution of degrees of weight consists in the way it figures in the overall system of metaphysical concepts and beliefs.

According to Kant, metaphysics is not a self-enclosed moving whole disconnected from what is known in the empirical and the exact sciences.24 Metaphysical contentions must cohere with warranted scientific theories. Yet even this supposition, which concerns the way in which knowledge is structured, involves the attribution of degrees of weight, which is in principle debatable.

(7) The attribution of degrees of weight is inherently connected to the interest that one has in the controversy, that is, to what one believes to be important.

Kant thought that the rationality of metaphysical controversies could not be disconnected from the interest that the parties involved have in the debate.25 This interest is connected to the human kind of desire and the human kind of (rational) will. This idea is sometimes expressed in Kant’s exegesis as the “primacy of the ethical.” It may therefore be assumed that the attribution of epistemic values to metaphysical beliefs is determined by two types of reasons. One type of reason connects the metaphysical beliefs, individually or collectively, to what is considered to be scientifically warranted. But in most cases, the connection to scientifically corroborated knowledge does not settle metaphysical controversies. The other type of reason connects the metaphysical beliefs to other metaphysical and ontological beliefs. The interest in metaphysical claims to knowledge affects the latter type of reasons. But normally, one is blind to the extent to which it affects one’s judgment. This sort of blindness is one of the main reasons why controversies in metaphysics and the attribution of epistemic values are indispensable for criticism in this field. A fundamental point that Kant’s argument reveals is the following:

(8) The assignment of weight to metaphysical beliefs can change.

Again, this is not an explicit part of the argument. Yet the conceivability of Kant’s own claim to a revision in our notion of objectivity, which is required in view of 24. It should be noted, however, that these commitments that reflect the primacy of science in determining the epistemic value of a proposition need not be shared by all proponents. For example, some regard religious beliefs to be more important than scientific beliefs. Yet, the Kantian model is more suited to the present practice of metaphysics in analytical metaphysics. 25. See in particular the chapter “The Antinomy of Pure Reason, Third Section, On the interest of Reason in these Conflicts” CPR A 462/B 490, and the chapter “The Canon of Pure Reason” in the “Transcendental Doctrine of Method” CPR A 795/B 823.






the conceptual problems revealed by the metaphysical controversies, can be explained only if (8) is true. It may be assumed that each metaphysical position is bound to attribute degrees of warrant to the various metaphysical propositions that differ from the conflicting positions. The weakening of the epistemic value of some beliefs, and the strengthening of the value of other beliefs, can be accomplished by means of thought experiments, or by other means that reveal new possibilities or new conceptual connections. Yet, this must be done by considering the possibility of rival and incompatible positions that cannot merely be dismissed as contradictory.

(9) In some cases it is not possible to entirely eliminate the metaphysical intuition that grounds the conflict that one aims to resolve by dismissing it.26 The recalcitrant intuition is the basis for an illusion of reason.

This principle is central to the resolution of the conflicts of pure reason. As we shall see in Chapter 5, it plays also a significant role in current metaphysical debates. On the basis of Kant’s example of how innovations in metaphysics should be pursued, one can spell out one of the roles of controversies in metaphysics. The arguments that are acknowledged by metaphysicians as having the normative status of real metaphysical innovations, involve the recognition of a metaphysical problem for which there is no solution within available conceptual resources. The conflict is exhibited by controversies which the new position intends to resolve.27 The motives for constructing a metaphysical position cannot be disconnected from the problem that the relevant controversies reveal. I would like to suggest that this is central to Kant’s concept of metaphysics. Actual and imagined metaphysical controversies are inherently involved in the creative activity of constructing metaphysical positions and arguing for them. Kant’s argument in the first antinomy could be regarded as a paradigmatic example of one powerful type of argument that is used in metaphysical controversies. The explicit representation of the conflict misleadingly presents Kant as an outside observer. The conflict addressed by the argument in fact involves three distinct positions. The argument is intended to shift the justificatory degree of (Q) in order to explain away the metaphysical conflict.

26. In the case of the antinomies of pure reason, the intuition that one cannot eliminate is explicated as the supposition that objects are given to us as they are in themselves and not as they appear. 27. As Bennett (1974) notes, Kant’s arguments in the antinomies were apparently a response to the Leibniz-Clarke correspondence.

Chapter 1. The Kantian framework 

Kant believed that change in metaphysics was possible, but only at the price of ruling out all further changes.28 We need not accept the elimination of the possibility of further changes in order to follow his example of how changes in metaphysics are feasible. Not all changes in metaphysics are similar to the radical type of change exemplified by Kant’s own critical philosophy. However, I suggest that the features that his implicit argument reveals could serve as a basis for the account of the rational role of metaphysical controversies in metaphysics, and of why they are both unavoidable and fruitful.

28. With regard to the explicit argument, Dascal’s claim that “...metaphysical controversies do not play any constitutive role in the actual progress of knowledge for Kant” seems to be true. See Dascal (2005: 4–5). Yet, as I have tried to demonstrate here, the explicit argument is in fact part of a far more complex implicit argument in which controversies have a constitutive role in actual epistemic changes in metaphysics.

chapter 2

The idea of controversy and metaphysics 1. The limits of Kant’s theory Kant’s theory of metaphysical controversies contains valuable insights regarding the role of controversies in metaphysics. Yet, as it stands, the theory is incomplete. It leaves out the features of real polemical exchanges. The need to supplement Kant’s theory with an account of the role of polemical exchanges requires some clarification. It might be argued that properties related to polemical exchanges are irrelevant to metaphysics. Features of real communicative exchanges seem to be irrelevant to what metaphysical contentions are about. The methods used by metaphysicians involve reflectively conceiving possibilities, presenting thought experiments, etc.1 Warranting metaphysical contentions seems to be based mainly on grasping how concepts are applied to possible cases, and how a proposition fits other propositions. That is, it is a type of warrant that seems to rule out the relevance of features related to real communicative exchanges. If polemical and other communicative exchanges have any role in determining the rational character of metaphysics, they must be added to the context of discovery together with other psychological and sociological features that seem to be irrelevant to the method of metaphysics and to what metaphysics is about.2 Kant certainly admits that they are relevant to metaphysics in this particular sense. However, a historical examination of the actual conduct of metaphysicians reveals that this objection is flawed. The methods used by metaphysicians do indeed partly abstract from features of polemical exchanges. Features of such exchanges may be irrelevant to what metaphysical propositions are about. Yet, as I hope to persuade the reader in Parts Two and Three of this book, a close examination of the actual way in which metaphysical theories are constructed, distributed, received, or replaced by other theories reveals that in arguing metaphysically, metaphysicians seem to be bound to argue dialectically, that is, to respond to real arguments presented by real philosophers with whom they disagree. The arguments presented by these philosophers exploit the range of layers of the contexts 1.

See for example Jackson 1998.

2. Cf. Rescher 1977: Chapter 8.




of the exchange and of the background that determines the relevant community of researchers. These dialectical arguments are anything but a superficial layer of the theories presented and defended within this discourse. Traces of the polemical discourse can be detected in the formation of novel concepts, in the reasons that warrant the reception or denunciation of metaphysical theories, and in various other levels of metaphysical texts. How can the features of polemical discourse be involved in metaphysics qua rational discipline without entailing that metaphysical propositions are reducible to features of human nature or human discourse? The response I wish to defend here is that the issues with which metaphysics is concerned are bound to generate metaphysical polemics by virtue of the way they are presented to the human mind. Metaphysical polemics result from attempts to articulate the topics or issues that metaphysics is concerned with in the same way that they are presented to particular individuals in specific discursive contexts. They are anchored in the fact that concept formation and the presentation of metaphysical theories articulate the modes in which the topics of metaphysics are presented to us in particular discursive contexts. When confronted with the diversity of conflicting ways in which metaphysical intuitions are expressed, the positions based on them, and their histories, one is tempted to either endorse metaphysical or ontological relativism, or to be satisfied with one’s own position, and to dismiss the positions that conflict with one’s own as irrelevant. The significance of controversies to metaphysics is revealed when one realizes why an incompatible position cannot be dismissed as inherently irrelevant to one’s own position even if the warrant of a metaphysical contention must also have an intuitive basis. One cannot merely rely on how things are intuitively given since intuitions are defeasible sources of knowledge. Controversies are the discursive means through which other people’s contentions that conflict with one’s intuitions become relevant to one’s own conceptions. In this sense, they have an indispensible critical role in the pursuit of metaphysical knowledge. Paraphrasing Kant, it could be said that metaphysics without metaphysical intuitions is empty. However, metaphysics is to a certain extent blind without real polemical exchanges in which conflicting intuitions are articulated. In the present chapter, I will point out some features of polemical arguments relevant to my present concern. I will then clarify the features of the main type of polemical argument, the epistemic relevance of which to current metaphysics will be examined in detail throughout this book. The following chapter is dedicated to an account of the concepts of metaphysical intuition and metaphysical thought experiments. The study of arguments is a complex and an expanding field of inquiry. Since my intention here will be to spell out the features of a Kantian antinomy conceived

Chapter 2. The idea of controversy and metaphysics 

as the form of real polemical exchanges, I will only discuss the issues related to this type of argument. In particular, my concern here will be with features of arguments presented in theoretical and epistemic contexts that are relevant to metaphysical polemics. Also, since my concern here will mainly be with metaphysical controversies between experts in this field, I will focus primarily on features of this type of controversy. My account will leave out many other important issues. Yet, I hope that what I say here will clarify the epistemic significance of polemical dialogues to metaphysics. 2. The nature of controversies as polemical dialogues What distinguishes controversies from other polemical exchanges? Following Dascal, there are three types of polemical dialogues. Dascal distinguishes between two macro-levels of polemical dialogues, namely, the ‘strategical’ level and the ‘tactical’ level (1998a: 19). At the strategical level, he differentiates between three ideal types of polemical dialogues: ‘discussion,’ ‘dispute,’ and ‘controversy.’ The criteria for this typology are: (a) the scope of disagreement, (b) the kind of content involved, (c) the presumed means for resolving this disagreement, (d) the ends pursued by the contenders, (e) the typical way in which the exchange ends, and (f) the cognitive gain (1998a: 21). According to Dascal, at the tactical level, three types of moves are employed in polemical exchanges: ‘proof,’ ‘stratagem,’ and ‘argument.’ Although each type of dialogue allows for the occasional use of the above three types of moves, each type of dialogue “has an inherent affinity with one of these types of moves.”(ibid.) ‘Discussion’ has an inherent affinity with ‘proof,’ ‘dispute’ with ‘stratagem,’ and ‘controversy’ is related to ‘argument.’ According to Dascal, a fundamental property that distinguishes controversies from discussions is that the former allow no solution. Arguments are the tactical means used by contenders involved in a controversy to persuade the addressee to believe that a proposition is true, while a proof purports to establish the truth of a proposition (1998a: 25).3 Controversies are also distinct from disputes. The stratagem, which is the typical tactical means used in disputes, “purports to cause a relevant audience to react in a certain way by inducing the audience to believe that a proposition is true,” (1998a: 25) while arguments “achieve their effect by providing recognizable reasons.” (ibid.) In other words, recognizable reasons are used in 3. It is important to note that the category of proof, identified as a kind of move that establishes the truth in the above sense, is not identical to the category of proof found in axiomatic systems, or exact sciences. If the truth of a proposition is established beyond reasonable doubt, the proposition might very well be false. ‘Beyond reasonable doubt’ is not equivalent to ‘beyond any possible doubt.’




arguments (in contrast to stratagems) in order to persuade without providing a solution (in contrast to proofs). Like proofs and unlike stratagems, arguments use reasons in order to persuade.4 The difference between proofs and arguments lies in the kind of relation that the reasons bear to the truth of the proposition. Although I basically accept Dascal’s typology, it raises some questions that are of interest to the present study. The main problem concerns his distinction between the types of reasons used in arguments and in proofs. Dascal’s view seems to be that proofs are truth-directed, while arguments are belief-directed. Yet, this distinction appears to collapse in epistemic contexts. If there is a distinction between purporting to establish that a proposition is true and persuading one to believe that a proposition is true, it seems that there should be a distinction between epistemic reasons for believing a proposition, and epistemic reasons for holding it to be true. But in epistemic contexts, in which the pursuit of knowledge is the goal of the exchange, no such distinction is available. Given the standard account of knowledge, epistemic reasons are reasons to believe that the proposition is true. Assuming that one ought to accept only the propositions that one has sufficient reasons to believe to be true, the recognizable reasons used in epistemic contexts to persuade one to believe that p are reasons related to the truth of p. Also, the distinction between the two types of reasons cannot be based on differences between types of epistemic warrant. Defeasibility cannot constitute the difference between arguments and proofs. In most empirical cases, the recognizable reasons are defeasible. Not all reasons to believe are epistemic reasons, since epistemic motives are often inextricably intertwined with non-epistemic motives. This is an important issue that deserves some attention. Non-epistemic motives are relevant to the account of the nature of controversies not merely where the fundamental layer of metaphysical intuitions is concerned. Entering a debate or ceasing to be involved in it is a deliberate act guided by several kinds of non-epistemic motives: one’s assessment of the personal gain of entering the exchange or terminating it, of the prospects of the sociological and institutional gain to the status of the position or discipline that one advocates, etc. There are cases in which it is difficult to distinguish between the epistemic and the non-epistemic motives. Nevertheless, this does not entail that the distinction between epistemic and non-epistemic reasons to believe is lost. This distinction grounds an important type of critical argument used in epistemic contexts. These arguments criticize reasons that express commitment to non-epistemic motives. Arguments that utilize the distinction between epistemic and non-epistemic motives are powerful critical tools. 4. Controversies therefore belong to the category of persuasion dialogues. See Walton 1998: Chapter 2.

Chapter 2. The idea of controversy and metaphysics 

If one wishes to keep the distinction between the two types of use of reasons denoted by Dascal by ‘argument’ and ‘proof,’ then what distinguishes between them? My response to this question is limited to the metaphysical case. It contains two parts. First, although the idea that in epistemic contexts arguments aim to persuade the addressees merely to believe that p should be rejected, there is a type of acceptance of propositions that lies between acceptance as a mere hypothesis or assumption, and acceptance of a proposition as being outright true. Following Rescher (1977, 2006), this type of acceptance can be characterized as accepting p as a presumption. Accepting p as a presumption is not, properly speaking, a claim to knowledge, and yet it involves a provisional commitment to the truth of p and is related to the pursuit of knowledge. Arguments presented in controversies aim to persuade the audience to accept p as a presumption. Although accepting p as a presumption is not the same as accepting it as true, presumptions are not disconnected from truth. This seems to be close enough to Dascal’s formulation, as far as controversies regarding matters of knowledge are concerned. I will discuss the concept of presumption and its relevance to metaphysical controversies in the following section.5 Next, we should note that according to Dascal, one feature related to the distinction between arguments and proofs is the scope of the disagreement, that is, the way in which the parties involved conceive the disputed matter. I suggest that polemical exchanges that are extensionally equivalent to Dascal’s ‘discussion’ are polemical exchanges in which it is possible to address only the explicit topic of the exchange due to the features of the context and of the subjects discussed. In contrast, in another type of debate that is extensionally equivalent to Dascal’s ‘controversy,’ the overt disputed matter is conceived of as the tip of the iceberg, which is related to many other issues that cannot be separated from the explicit topic of the dispute. A polemic exchange may begin as one type of dialogue, and then become another type. At the initial point, the subject matter disputed explicitly is a particular topic. However, in contrast to discussions, the interlocutors’ choice of reasons in a controversy will be sensitive to the implications of these reasons for their entire position. By contrast, this type of sensitivity is not to be expected in the case of discussions. This difference between controversies and discussions also manifests itself in the motives of the interlocutors to continue the exchange, or to end it.

5. Presumptions have a key role in the project that Dascal calls ‘soft rationality,’ an account of rationality that he detected in Leibniz’s relevant writings. For the significance of Leibniz’s soft rationality to the pragmatic theory of controversies see Dascal 2005b, 2008b, 2008c.




3. Presumptions and the burden of proof According to Rescher (2006: 1), “[to] presume...is to accept something in the absence of the further relevant information that would ordinarily be deemed necessary to establishing it.” Similarly, in Ulmann-Margalit’s view (1983: 143), “[presumptions] have to do with assumptions made ahead of time...a supposition not fully justified, yet not quite rash either.” The notion of presumption has legal roots and is nevertheless significant for a variety of cognitive and epistemic affairs. It is traditionally associated with the allocation of burden of proof. This seems to be a point of agreement between theorists exploring presumption. Many contemporary theorists agree that presumption has an important role in argumentation theory. Yet, as Godden & Walton note (2007: 313), “argumentation theory lacks a unified, robust account of the nature of presumptions.” My intention in this section will not be to pursue such an account. It is plausible to maintain that it is not possible to assign one concept of presumption to the various uses of ‘presumption’.6 In pursuing an account of the relevant concept of presumption, I will mainly refer to Ullmann-Margalit’s (1983) and to Rescher’s (1977, 2006) illuminating accounts. I will mainly attempt to clarify the characteristics of presumption that are relevant to the epistemic significance of metaphysical controversies. My main claim is that the acceptance of metaphysical propositions has certain features which render them similar to presumptions, though usually they are not conceived as such. The dialectical character of presumptions (to be explicated below) can clarify the motivation to integrate an account of metaphysical polemics in the account of the nature and content of metaphysical contentions. In the article referred to above, Ullmann-Margalit’s main intention is to clarify the role that presumptions have in a theory of action. According to UllmannMargalit (1983: 149), presumption formula should be represented by the string ‘pres (P, Q). P stands for the presumption-raising fact whereas Q stands for the presumed fact. For example, consider (1) below: 1. There is a presumption that a child born in wedlock is legitimate. In (1), the presumption-raising fact is ‘being born in wedlock’, while the presumed fact is ‘being fathered by the mother’s husband’. Ullmann-Margalit next interprets the presumption formula as the following rule:

6. As Freeman notes (2005: 23), there is an obvious difference between presumptions in the legal sense and presumptions in the epistemic sense. Presumptions in the legal sense are stipulative, which seems to distinguish them from presumptions in the epistemic sense. This may support the supposition that there is more than one concept of presumption.

Chapter 2. The idea of controversy and metaphysics 

Given that p is the case, you (=the rule subject) shall proceed as if q were true, unless or until you have (sufficient) reason to believe that q is not the case. (1983: 147)

As Ullmann-Margalit notes (ibid.), the presumption formula is propositional. Nevertheless, in her view, presumptions are concerned “not so much with ascertaining the facts as with proceeding on them...”(ibid.) Although presumption rules are associated with inferences, they should be distinguished from them: To avoid confusion, let me draw attention to a distinction. Quite apart from whether or not presumption rules are about inference, there is associated with them a rule of inference. It is the formal rule governing the operation of the ‘pres’ operator: from (i) pres(P,Q) and (ii) p, it follows the (iii) pres q...(1983: 148)

Interestingly, Ullmann-Margalit’s view is that the presumption is not associated with any commitment to the truth value of the presumed fact: So the conclusion of the rule of inference associated with presumption rules is to the effect that a certain fact is presumed. But this is not to be confused with an inference to the presumed fact. And the question with which we started was whether or not what presumption rules are about is inference to the presumed fact(s). In answer to this question I submit that the presumption rule involves no commitment to, nor guarantee of, the truth value of the presumed fact q. It makes no claim upon its subjects’ cognitive or epistemic systems. The rule entitles one to hold q as true for the purpose of concluding one’s practical deliberation on the impending issue; it neither requires nor entitles one to believe that q. (1983: 149)

The distinction between presumption rules and any sort of commitment to the truth of the presumed fact, that is, of any claim upon the subject’s cognitive system, is somewhat surprising. This claim is plausible with regard to cases that are similar to the legal presumption of innocence. Yet it is questionable where other presumptions are concerned; for example, presumptions similar to (1). Clearly, presumption rules do not sanction outright commitment to the truth of the presumed fact. But the claim that the presumed fact makes upon our cognitive system apparently differs from that of a hypothesis or an assumption. In contrast to Ullmann-Margalit, Rescher places the type of commitment related to presumptions between outright commitment to truth and no commitment. He distinguishes between presumptions and assumptions as follows: Presumption is not the same as assumption – let alone presupposition. To be sure, in presuming something we assume (or suppose) it to be so: a presumption is an assumption (or supposition). But the reverse is not the case: not any old assumption (or supposition) is presumption. We can assume something to be so for the sake of discussion or argument or deliberation. But here the result is a mere hypothesis while a presumption is something we actually accept. (2006: 29)




Presumption is linked here to differences in types of acceptance concerning one’s commitment to the truth of the proposition: ... acceptance does not lie along a one-dimensional spectrum which ranges from ‘uncertainty’ to ‘certainty.’ There are not only degrees of acceptance but also kinds of acceptance. And presumption represents such a kind: it is sui generis, and not just an attenuated version of “acceptance as certain.” (1977: 42)7

As Rescher notes (2006: 6), “[...] presumption is certainly not knowledge: we do not know what we merely presume.” And yet, “[to] class a proposition as a presumption is to take a definite and committal position with respect to it, so as to say “I propose to accept it as true insofar as no difficulties arise from doing so.”” (2006: 23) Hence, in Rescher’s view (2006: 23), presumption is an epistemic category. A cognitive presumption stakes a claim that outruns the substance of actually available information; it is a proposition that, in suitably favorable circumstances, is accepted as true in the absence of any counterindications. (2006: 27)

In Rescher’s view (2006: 14), presumption and burden of proof represent correlative concepts. As he adds (2006: 19), the idea of burden of proof is not a strict logical concept: “[...] rather than being a logical concept, burden of proof is a methodological one. It has to do not with valid or invalid reasoning, but with probative argumentation in dialectical situations.”(ibid.) Presumption is a dialectical concept, and is central to Rescher’s (1977) account of the epistemic significance of controversies to ordinary and scientific knowledge. Arguments used in dialectical situations, that is, in controversies, aim to shift the burden of proof by providing reasons to accept p as a plausible presumption, and not to persuade one to hold it as being true beyond reasonable doubt. The weight of a presumption is determined by its plausibility or reasonability. However, plausibility should be distinguished from a measure of probability. Various features determine the plausibility of a presumption. As Rescher notes (2006: 40), the probative strength of confirming evidence substantiates the plausibility or reasonability of a presumption. In addition (ibid.), the plausibility of a contention may be based on “thesis-warranting principles.” In this context, inductive considerations such as simplicity, uniformity, specificity, definiteness, and determinativeness come into play. But according to Rescher (2006: 42), the plausibility of a presumption is not merely determined by inductive or probative features. Presumption and the correlative concept of burden of proof are, in Rescher’s view (2006: 22), pragmatic devices that pertain to language users and not to language as such. 7.

See also Rescher 2006: 23.

Chapter 2. The idea of controversy and metaphysics 

In Rescher’s view, the main feature that validates a presumption is its usefulness or pragmatic value: ... the justification of presumption is, in the final analysis, purposive – a matter of future efficacy within the context of purpose in which the practical presumption at issue figures. (2006: 48) A cognitive presumption as such is not a thesis or theory: it represents a practical policy – a modus operandi. (ibid.) ... what justifies one’s making presumptions about factual matters, in the wake of acknowledging that presumptions are not established and certifiable truth? The answer is that presumption is not so much a matter of evidentially probative considerations as of procedurally practical ones. Presumption is a functional, purposeserving device...epistemic presumptions serve as tools to facilitate the objections for which such cognitive practices as communication and inquiry are instituted. And the validation of a mode of practice is always ultimately pragmatic. (2006: 49)

In Rescher’s view, the cognitive role of presumptions is not limited to a particular epistemic field. Cognitive presumptions shape Rescher’s epistemology as a whole. His examples include presumptions such as that regarding the evidence of the senses, and he presents no metaphysical example. Yet, there is an apparent similarity between his account of presumption and the nature of acceptance of metaphysical propositions. The acceptance of metaphysical propositions share many features that are assigned by Rescher to presumptions in general: a. Metaphysical propositions are accepted as true in the absence of the further relevant information that would be deemed necessary to establishing it. Probative and inductive evidence contribute to the plausibility of a metaphysical proposition. But these are not sufficient for accepting them as being outright true. b. Metaphysical propositions are not mere hypothesis or assumptions. They are accepted as true though not as being outright true. c. The acceptance of a metaphysical proposition is linked to the allocation of the burden of proof to those that disavow them. d. Metaphysical propositions are defeasible. Arguments can be produced for accepting them or for disavowing them. Although the acceptance of a metaphysical proposition seems to be similar to the acceptance of a presumption, there also seem to be some differences between the two. It might be objected that, in contrast to Rescher’s account of the nature of presumptions, the acceptance of metaphysical propositions cannot merely be identified with pragmatic value even if their acceptance is similar to the acceptance of a presumption. What is accepted when a metaphysical proposition is accepted seems to extend far beyond human needs and practices. If the claim that the acceptance




of a metaphysical proposition means acceptance of it as a presumption, which entails that the reasons for accepting a metaphysical proposition consist merely in its pragmatic value, this claim begs the question. As a response, I submit that the acceptance of a metaphysical proposition cannot be justified merely by pointing out its pragmatic value. And yet, the acceptance of a metaphysical proposition cannot be disconnected from the practical dimension. The practical role of metaphysical propositions in cognitive matters is discernable, for example, in issues such as the unity of science. Needless to say, the acceptance of metaphysical propositions is a source of motivation for a variety of ethical and religious concerns. Note that granting the apparent similarity between metaphysical propositions and presumptions need not be linked to the ubiquitous epistemic role assigned by Rescher to presumptions. Rescher’s standpoint is, in this sense, opposed to Kant’s position, which distinguishes between reason and the understanding. We may suspend judgment on the question of whether presumptions indeed have the ubiquitous role that Rescher assigns them. How could metaphysical propositions be analyzed as presumptions? Consider the following example. Materialists hold that all the properties and objects that our world contains are physical properties and objects. This claim is sometimes stated in terms of possible worlds: Every world that is minimally a physical duplicate of our world is a duplicate simpliciter of our world.8 In this case, the presumption-raising fact is the main thesis of physicalism: All the objects and properties in our world are physical. Clearly, this claim is not an empirical generalization. It is indeed supported by the success of the physical sciences in explaining physical phenomena without violating the closure thesis (according to which all physical phenomena are explainable merely on the basis of physical properties and relations), but this evidence does not in itself suffice to establish its outright truth. The main thesis of physicalism is a defeasible proposition. The presumed fact in this case is, for example, the contention that conscious phenomena are physical states, properties, and processes. Both the presumed fact and the presumption-raising fact could be challenged. Evidence could be raised against the presumed facts, and, given that enough presumed facts were defeated by contrary evidence, it is more plausible to abandon the presumption-raising fact. Presumptions are defeasible. They are accepted as true unless there are plausible reasons that rebut them. What, however, is the nature of the reasons that can be used to rebut a presumption? This question has an easy answer with regard to ordinary presumptions such as the above (1). In order to rebut the presumption that a child born in wedlock is legitimate one could use empirical evidence, 8. A minimal physical duplicate of a world is a world that contains only physical objects and properties. See Jackson 1998; Chalmers & Jackson 2001.

Chapter 2. The idea of controversy and metaphysics 

observations, etc. If it is established, say by means of DNA tests, that a child that is born in wedlock is not legitimate, it would be reasonable to abandon the presumed fact that the examined child is legitimate. If in most cases in which children born in wedlock are examined and revealed to be not legitimate, it would be reasonable to abandon the presumption-raising fact. What, however, are the types of reasons that could be used in order to defeat a metaphysical presumed fact or a metaphysical presumption-raising fact? It should first be noted that the above type of empirical counterevidence clearly will not suffice in order to rebut metaphysical presumptions. There are, I suggest, three types of defeaters of metaphysical presumptions. A metaphysical presumption is disavowed if one reveals its inconsistency with a stronger presumption, that is, a presumption that is more plausible than the first one. A metaphysical presumption could also be questioned on the basis of the metaphysical intuitions invoked, for example, by metaphysical thought experiments, or by other means. Finally, a metaphysical presumption could be undermined by means of arguments introduced by opponents representing competing and incompatible metaphysical positions in metaphysical controversies. The third type of counterevidence is related to the fact that presumptions are not merely defeasible but may also conflict with each other. However, in cases in which there are two or more conflicting presumptions, which one should be accepted? Given Rescher’s account, the answer seems quite simple: the presumption that is more reasonable and plausible. Yet this response is not always helpful when the conflicting presumptions of proponents involved in a controversy are at stake. One cannot assume that plausibility is an objective, quasi-objective, or intersubjective feature of presumptions that is shared by all interlocutors. Each of the proponents may view some of his presumptions as being more reasonable and plausible. In the metaphysical case, the weight of the presumption is determined by empirical information, metaphysical intuitions, as well as one’s practical motives. Opponents involved in controversies could therefore differ in how they assess the weight of their presumptions. The problem raised by such cases is not merely that we do not possess a neutral objective or quasi-objective basis for determining the plausibility of the competing presumptions. The idea of a presumption is correlated with that of a burden of proof, and it is not clear in this case how the burden of proof should be allocated. I will discuss metaphysical intuitions in Chapter 3. In the final section of this chapter, I will clarify how the burden of proof is allocated in cases of metaphysical controversies. 4. Radically opposed views and chains of controversies Controversies are discursive phenomena. They are always carried out by proponents and opponents by means of oral and written acts of communication.




However, who are the proponents and opponents involved in metaphysical controversies between professional philosophers? Surely, these are individual philosophers presenting arguments for and against some theses in particular discursive contexts. Their aim is apparently to persuade the opponent or the audience to accept the theses they support and to abandon the theses they criticize. Yet, this account does not suffice as a description of what such controversies involve. Contenders may, indeed, defend their own presumptions and attack the presumptions of their opponents. But they may be conceived as representatives of philosophical or metaphysical positions qua something that is shared by other individuals in different historical phases, and not merely their own particular view. In controversies, according to Dascal, contenders hold radically opposed views. With regard to controversies between representatives of opposed metaphysical positions, I suggest that commitment to radically opposed views is commitment to incompatible metaphysical positions. Positions are social, cultural, and historical entities, but they cannot be identified with any set of statements explicitly held by their representatives at a given time; they involve unarticulated intuitions that await articulation, explicit beliefs, and implicit commitment to ethical and religious values, preferences, and principles of relevance.9 Where metaphysics or philosophy are concerned, the titles of such positions are terms such as ‘Platonism,’ ‘Aristotelianism,’ ‘Cartesianism,’ Kantianism,’ ‘Hegelianism,’ ‘Neo-Lockeanism,’ that is, positions that are named after the authors of philosophical systems. Alternatively, they are referred to by means of terms such as 9. Hamblin (1970, 1971) devised the concept of ‘commitment store’ or ‘commitment set’ that allegedly includes a store of statements of the totality of an arguer’s commitments during the sequence of a dialogue. As Walton notes (1998: 43), Hamblin’s notion of commitment was meant for dialogues in which information is exchanged. These dialogues differ significantly from persuasion dialogues, which also include disputes. According to Walton (1984, 1998), where persuasion dialogues are concerned, the relevant concept of commitment differs from Hamblin’s concept. For example, according to his loophole or way-out principle (1984: 138), a respondent should reject or withdraw from any statement that together with other statements implies the opponent’s thesis, either directly or indirectly. I suggest that this violates the sincerity condition, which should preferably or at least ideally be posited in epistemic contexts. Yet, in my view, the idea that one’s commitments may entail statements that are incompatible with one’s own position is significant. Given that one’s commitments do not merely include some statements that one formulates but also every statement that is entailed by one’s commitments, the notion of a well-defined set of commitments is groundless. In particular, assuming that rational creatures are not committed to contradictions when they explicitly conceive them as such, and assuming that some of the statements to which one is committed might entail a contradiction, the strict notion of commitment cannot have clear boundaries. I suggest that commitment should be replaced by presumption, which means commitment to statements that inherently depends on the information extracted from the argumentative-dialogical context that shifts with the evolution of the dialogue as new arguments are presented in it.

Chapter 2. The idea of controversy and metaphysics 

‘Empiricism,’ Rationalism,’ ‘Nominalism’, ‘Realism,’ ‘Logicism,’ ‘Representationalism’, or ‘Disjunctivism.’ In this case, the title of the position originates from the propositions or properties that identify the positions, and distinguishes them from other positions. These propositions are normally associated with a certain theoretical approach to a given set of philosophical or metaphysical problems and questions. Positions can evolve out of positions by means of some distinguishing features. The variety of Kantian positions and Wittgensteinian positions may illustrate these types of positions. Similarly, the varieties of functionalism, representationalism, or HOT theories of consciousness in the philosophy of mind illustrate the same type of phenomenon. Positions in metaphysics are expressed by means of statements that belong to a given natural language. Although technical language is often used, the purpose of introducing new terms is to refine and improve existing vocabulary. Metaphysical positions usually have a claim to what we all mean by what we say, whether we know explicitly what we mean or not. Some of the relevant characteristics of ‘position’ are the following: 1. A position does not normally have clear boundaries, but can be distinguished from rival positions due to the following facts: (a) Each position has a set of propositions expressed in some natural language that are conceived by the supporters of the position as being particularly significant and important. I will refer to this set as the ‘core set’. (b) Some propositions that belong to the core set are incompatible with propositions that belong to the core set of the rival positions, or imply propositions that are either incompatible with a proposition that belongs to the core set of the rival positions, or with propositions entailed by propositions that are in the core set of the rival positions. (c) The propositions that belong to the core set are linked to other propositions that can be expressed in some natural language, and may also include observational statements. 2. A position need not be introduced to the public as a comprehensive worldview. In metaphysical controversies, the explicit topics of the exchange are normally propositions that are explicitly disputed. I will refer to them as ‘symptoms.’ Symptoms indicate the differences between positions.10 The symptomatic propositions need not be part of the core set of the position, but are linked to the core set. 10. These could either be statements acknowledged to be such symptoms at the beginning of the controversy, or statements that are revealed to be a matter of disagreement as the controversy develops. The notion of a symptom is akin to that of ‘thesis.’ See van Eemeren & Grootendorst (1984: 85) and Walton (1998: 39). I prefer the term ‘symptom’ for the following reason: It implies the link between the explicit topic of the controversy at the beginning of the controversy, and the position that lies in the horizon of the intention involved in the act of




Metaphysical propositions do not face reality in isolation. They receive support from each other. What representatives of a given position accept as their presumptions is not the same as what the rival position or positions accept as their presumptions. Arguments are used either in order to establish the symptomatic propositions (or propositions that are entailed by them), or to undermine them. Nevertheless, in certain cases, one must distinguish between the explicit and implicit goals of arguments used in a controversy. The explicit goal is to advocate acceptance or denial of the symptomatic proposition. The implicit goal is to undermine one position and to establish the competing position.11 If the explicit goal is accomplished, this does not entail that the implicit goal has been achieved. The implicit goal lies in the horizon of the intention involved in the act of persuasion. Controversies between representatives of incompatible positions involve the positions themselves. This fact has significant results regarding the way in which controversies belonging to this type reach their closing point, as well as the nature of this closing point. Consider, for example, the account of this issue by the representatives of the pragma-dialectical theory. According to the theory defended by van Eemeren & Grootendorst (2004), there are 15 rules of critical discussion. Rule 14 is the following: a. The protagonist is obliged to retract the initial standpoint if the antagonist has conclusively attacked it (in the manner prescribed in rule 9) in the argumentation stage (and has also observed the other discussion rules). b. The antagonist is obliged to retract the calling into question of the initial standpoint if the protagonist has conclusively defended it (in the manner prescribed in rule 9) in the argumentation stage (and has also observed the other discussion rules). (2004: 154)

According to van Eemeren & Grootendorst (2004: 151), the protagonist has conclusively defended an initial standpoint if she has successfully defended both the propositional content called into question as well as the force of its justification. The rules suggested by van Eemeren & Grootendorst are useful in ordinary cases of critical discussions. They are, however, inapplicable to the way in which controversies between representatives of metaphysical positions end, and they are unable to clarify how such controversies evolve from each other. Let us note that it is not clear what ‘conclusive’ means in this context. It may be assumed that the conclusive defense of a position consists in the presentation of persuasive arguments arguing. In other words, the term is sensitive to the fact that controversies tend to develop from one given topic to many other topics that are related to the positions taking part in the controversy. 11. Arguments used in controversies usually comprise both goals.

Chapter 2. The idea of controversy and metaphysics 

that the other party is currently unable to counterattack. However, in many cases, the mutual recognition of the participants that the antagonist or the protagonist have presented persuasive arguments for the position they support does not result in the attacked party’s retraction of her initial position. Nevertheless, this should not be depicted as an irrational move on the part of the attacked proponent. Rather, this move may reflect her presumption that the current absence of good reasons that undermine the persuasive power of the arguments presented to attack her position does not entail that such reasons will not be available at a later time, and that the current non-availability of such reasons does not suffice to undermine the strength of the presumption in favor of the attacked proposition. The controversy between the parties involved may be stopped but not terminated. It may be stoped without the losing side retracting the proposition she defends. This situation can be compared to the differences between winning a battle and winning a war. The battle may have been won, but the war lingers on. A new controversy addressing the same issues between different representatives of the same positions, in which different presumptions are challenged, may emerge out of the first controversy. The new controversy may also spread to other topics that were not explicitly discussed in the initial controversy. Yet the explicit topics discussed in the new controversies are inherently related to the results of the first round of controversies in a way that allows a rational reconstruction of this development. I will use the phrase ‘chain of controversies’ to refer to a class of controversies that evolve from each other and are linked to each other. Normally, in metaphysics these are controversies between incompatible positions represented by different representatives of the positions in each stage of the development of the chain. Let us sum up this section by noting that one way of pointing out why arguments used in controversies do not provide solutions consists in the distinction between their explicit goals and their implicit goals. A successful argument cannot be evaluated without noting how it enhances or diminishes the plausibility of the whole position that is part of the horizon of the intentional act. Nevertheless, in assessing the results of a controversy, particular attention should be paid to the way in which the arguments used affect the plausibility of the positions involved, and motivate the contenders or the non-active audience to either accept or reject them.12

12. According to van Eemeren & Grootendorst 1992: 34, a conflict is resolved “only if somebody retracts his doubt because he has been convinced by the other party’s argumentation or if he withdraws his standpoint because he has realized that his argumentation cannot stand up to the other party’s criticism.” This, I suggested, is a rather narrow conception of resolution. In fact, where metaphysical controversies are concerned, complete resolution is very rare. Nevertheless, as I hope to show in what follows, controversies motivate significant changes even when they are not completely resolved. On this issue see also Walton (1998) Section 7.




5. The addressee According to Perelman (1982: 11), “argumentation is intended to act upon an audience, to modify an audience’s conviction or dispositions through discourse...” In controversies, one party endeavors to persuade another party by means of reasons. However, who is the addressee that one is trying to persuade? To what audience is she speaking? The addressee might be the opponent herself, that is, the person criticized by the contender. But this is not always the case in public controversies in which messages are publicly distributed. Indeed, in public controversies, contenders directly address and interpret what their opponents say by exploiting the variety of layers of meaning of their opponent’s symbolic message. But in addition to the contenders themselves, controversies held in public contexts also involve an audience, comprised of observers that are not actively involved in the exchange but are exposed to it. In many if not most such controversies, the agent that one intends to persuade is not one of the contenders but rather a competent audience of observers.13 As Perelman & Olbrechts-Tyteca note (1969), an audience is constituted by agreements. The agreements that constitute audiences naturally differ between audiences. Perelman & Olbrechts-Tyteca (ibid.: 28) distinguish between the type of argument addressed to a specific audience, that is, arguments that aim to persuade, and the type of argument addressed to an audience that transcends all specific audiences – the universal audience. The latter type of argument includes arguments that aim to convince rather than persuade, that is, “to gain the adherence of every rational being.” (ibid.) Which audience is addressed by the arguments presented by contenders involved in metaphysical controversies? In Perelman & Olbrechts-Tyteca’s terms, it might be supposed that the audience is the universal audience that consists of all rational beings. But what is a universal audience? One way to interpret this term is to claim that arguments that are meant to convince the universal audience should be arguments that address any rational being whatever her presumptions might be. Another way is to maintain that the arguments that are meant to persuade every rational being hold with respect to all such beings even if not all rational being would necessarily be persuaded by them. The use of arguments in metaphysical controversies is clearly not merely intended to persuade a particular audience but to convince all rational beings. Yet the arguments presented in real controversies address the universal audience if at all, in the second of the above suggested interpretations. These are arguments that are meant to persuade a particular audience although they are meant to be valid for all rational being. These two apparently conflicting features presumably reflect the arguers’ presumptions that there is no point trying to convince all possible rational beings whatever their 13. See van Eemeren & Grootendorst 2004: 99.

Chapter 2. The idea of controversy and metaphysics

presumptions might be in order to establish what they consider to be the most plausible resolution of a conflict for any rational being.14 Particular audiences are constituted by particular agreements, and the agreements relevant to the audiences that speakers and writers address in metaphysical controversies is constituted by agreements of what they accept as their presumptions. The audience may consist of supporters of one of the contenders, intellectually mature ordinary agents who are not familiar with the matter at hand, professional philosophers from a specific philosophical genre, agents who share the same religious background, scientist from a certain discipline (physicists, cognitive scientists, and neuroscientists), students beginning to form their intellectual identity, and so forth. The nature of arguments is naturally also determined by the identity of the intended addressee. The choice of rhetorical means, the types of moves used, and the type of information exploited and distributed are adapted to the intended addressee – the agent that the interlocutor intends to persuade by means of her argument. For example, if the opponent’s arguments criticize a religious position held by the contender, the goal of her response need not be to persuade the opponent, but rather to enhance the self-assurance of the audience of observers who share her religious views. Alternatively, her goal might be to persuade an intellectual audience that supports neither her position nor her opponent’s position. The identity of the intended addressee can be revealed by examining the means used in the communicative exchange of symbolic messages. For example, a contender who wishes to persuade her opponent will avoid insulting her opponent and will refrain from invoking arguments that ridicule her opponent’s position. On the other hand, such rhetorical means might be used for an audience consisting of non-active neutral or supportive observers. 6. Persuasion and the rational support of reasons In a controversy, the parties involved often tend to articulate their differences in terms of the propositions that each one accepts or denies. However, their differences also consist in what each contender accepts as good reasons. The contender that recognizes the differences in how she and her opponent conceive reasons to be good or bad does not normally think that the reasons she provides would only be valid for her. It might be supposed that a rational dialogue is not possible in 14. According to Frogel (2005: 47), the speaker of a philosophical text is obliged to present herself “as one whose objective truth alone or...fear of illusion is what motivates and drives...” Interestingly, this characteristic may hold with respect to the above two suggested interpretations of the universal audience.






contexts in which the disagreement concerns the adequacy of the available reasons. In these contexts, there is clearly a temptation to depict the recalcitrant opponent as irrational, so that one would lose the motivation to try to persuade her in a rational way, that is, by using reasons. But in many cases, attempts to persuade the opponent do not stop when contenders realize that the reasons considered to be “highly convincing” are not regarded as such by the opponent. This is probably the most baffling feature of controversies. The differences between contenders are not merely limited to the matter under dispute. They also include the rational support for or against the issue at hand. How can the dialogue continue when this type of difference is revealed? How can one persuade one’s opponent in these contexts? On the one hand, contenders should attempt to persuade their opponent by means of reasons, that is, in a rational way. But, if the rational support provided by the contender’s reasons is interpreted differently by the contender and the addressee, it seems that arguments cannot be used for persuasive purposes. My response to this query will be limited to the case of metaphysical controversies. One source of the above puzzle is presumably the tendency to view conflicting metaphysical positions as incommensurable. But this view is misguided. Metaphysical controversies are not based on theoretical solipsism. The language of metaphysics that combines scientific, technical, and ordinary discourse makes it possible for there to be radically opposed positions without entailing that there are no points of intersection between them. Radically opposed metaphysical positions cannot fail to have points of intersection if they have a claim to the most basic features of being (there is only one being!), thoughts about being, and the language in which they are expressed. The common background suffices to assure the possibility of persuasion in contexts of disagreement such as those of radically opposed metaphysical positions. The goal of persuasion, in this case, is not merely to validate a statement by using reasons, but rather to lead the opponent or the audience to also accept these reasons qua good reasons. Validating the proposition and establishing the reasons qua reasons are not two distinct acts. By using reasons in particular dialogical contexts, contenders attempt to establish the matter at hand and to persuade the opponent or the audience to accept them as good reasons. 7. Epistemic function Controversies do not allow a solution. Yet given that controversies motivate epistemic changes, whether or not they allow solutions may change over time as a result of these epistemic changes. The responses that are available within a controversy and to a controversy are context-bound, and depend on the actual state

Chapter 2. The idea of controversy and metaphysics 

of knowledge, and the theories, concepts, methods, and procedures with which contenders are familiar. If persuasion as rational activity is feasible, there must be some recognizable reasons available to advocates of the opposite position that are linked to reasons provided by the contender’s argument. Ordinary knowledge of meaning does not merely allow incompatible positions but also makes it possible for a person to criticize a position from another person’s standpoint. It is a truism that ordinary knowledge of the meaning of a statement does not involve the explicit knowledge of all the statements implied by it. Arguably, contenders do not know some of the implications of the (symptomatic) statements of the positions they hold. In the ideal case, a position is internally complete if it can account for all the facts, phenomena, problems, and questions it purports to deal with. We may assume that no position is internally complete at any given point of time relative to what is conceivable by experts who advocate it and those experts who dispute it. The concept of conceivability-at-a-given-point-of-time is distinct from timelesslyconceivable.15 The rational support one could provide for a position on the basis of what is conceivable-at-a-given-point-of-time t differs from the timeless concept of a rational support. Conceivable-at-a-given-point-of-time indicates what is explicitly and reflectively available to contenders, that is, what is really conceivable to them. Being conceivable-at-a-given-point-of-time t and having the respective rational support at a time t may converge with timeless-conceivability and the respective timeless rational support. At least, there is no reason to suppose that they will never converge. Yet, what is conceivable-at-a-given-point-of-time t may differ from what is conceivable at an earlier time t-x or later point of time t+x and, therefore, also from timeless-conceivability. Naturally, the rational support for a given statement will change due to the changes in what may be really conceived. By contrast, timeless-conceivability is unaffected by the time something is conceived. Conceivable-at-a-given-point-of-time also makes it possible to distinguish between what is conceivable from within a given position, that is, from within the core set that makes it the position that it is, and what is conceivable from within the position that is incompatible with it. Where metaphysics is concerned, it is questionable whether timeless-conceivability allows conflicting positions, that is, if one assumes that all the relevant information is available to all contenders. It is possible to add that given conceivability at t may differ from conceivability at a later time t+x, contenders may not be aware of the scope and force of the internal difficulties, problems, and questions before the controversy begins.16 15. For a similar distinction, see Block 1990: 57. 16. For a closely related claim, see Freudenthal (1998: 158). As Freudenthal claims, “the opponent ... does not necessarily know all the statements which are or can be integrated into this system.”




By ‘real-possibility,’ I mean ‘conceivable-at-a given-point-of-time.’ ‘Real-possibility’ as used here is an epistemic modality. It stands for the conceptual capacities actually possessed by experts that advocate or oppose a given position; that is, the concepts they can explicitly conceive and employ by making judgments and by presenting arguments, and which are accessible to cognitive employment. Propositions that employ concepts and distinctions that are not available “yet” to the experts but can nevertheless be added to the position without conflicting with the core set that identifies it do not represent really possible states of affairs. Clearly, what is really possible may change in the course of time. In order to have some conception of an actual case of epistemic change that might be motivated by a controversy, let us first introduce a distinction between two kinds of arguments that might be used by contenders in controversies: (A) Arguments that aim to reveal internal difficulties taken by the arguer to be unexplainable by the position X that she is attacking. However, the reasons used in her argument do not use the presumptions related to her own position Y (presumptions that are incompatible with the position attacked). Such an argument might either use statements that are presumed to be compatible with the shared background (as represented by the contender) or with the core set of attacked position itself.

Type (A) could be regarded as providing direct support for the arguer’s position only if one assumes that either position X or position Y is well-founded, but not both; that is, if there is no third alternative.17 Such cases are clearly very rare. A second type of argument is the following: (B) Two contenders S1 and S2 argue for two incompatible positions P1 and P2. Both acknowledge the need to provide an explanation of a given fact f stated in their common language. An argument used by S1 has the following features: (a) it points out the fact that f is unexplainable within P2, (b) it demonstrates how f is explainable by using reasons related to propositions that are either (i) presumed by S1 to be part of the core set of P1 (and are not part of P2), or (ii) reasons related to propositions presumed by adherents of P1 and P2 that together with some commonly accepted novel reasons support the plausibility of a presumption that is part of the core set of P1 (and not of P2) and that explains f.

Clearly, successful arguments that belong to type (B) undermine P2 by providing support to S1’s position P1. But this is not the only significant feature of this type of argument. Representatives of incompatible positions naturally differ in the propositions that they accept as presumptions. They will naturally tend to view the propositions accepted as presumptions that are part of the core set of their 17. For a similar view, see van Eemeren & Grootendorst 2004: 173.

Chapter 2. The idea of controversy and metaphysics

own position as more plausible than those that are part of the core set of the rival position(s). Yet, this does not entail that all presumptions have equal normative status and that the contenders involved in the controversy are not aware of the differences in the normative status between the presumptions related to each position. To take one example, the presumptions shared by property dualists do not have equal normative status and equal plausibility as those of the supporters of physicalist/functionalist positions in analytic philosophy of mind. Although dualists may be convinced that what they presume is more plausible than what physicalists and functionalists presume, this does not entail that they are not aware of the fact that their position is the minority position. This difference is significant to the allocation of the burden of proof. Arguments of type B are particularly effective in shifting the burden of proof to the majority side. They do this by attempting to persuade their target audience that the propositions that both parties presume to be true cohere better with the propositions that are part of their core set than with those that are part of the opponent’s core set. The arguments that aim to shift the burden of proof to the majority side must therefore be sensitive to the agreements that constitute their target audience (in Perelman & Olbrechts-Tyteca’s sense). Positions are constituted by the conceptual links between the variety of presumptions (the presumption-raising facts) that form their core set and also by the links between the presumed facts. Hence, by connecting the propositions that are part of the core set of their own position to those that are presumed by both parties the users of type B arguments render their position relevant to that of the rival and incompatible position. There are at least three different types of ‘epistemic change’ within a controversy: (α) A phenomenon that has not been explained by S1 by using the rules, principles, available concepts and theories of her position P1 (that constitutes what is really-possible for her) is explained by applying the already established principles, available concepts and theories, types of arguments and existing evidence as a result of a challenge posed by the opponent.

(α) is probably the weakest case of growth of knowledge. It requires only the application of existing types of reasons, procedures, methods, and theoretical knowledge. (β) Certain phenomenon that has not been explained by S1 by means of the rules, principles, and theories that are really-possible within her position P1 is explained by using new concepts and principles, new types of arguments, or even by employing new methods as a result of a challenge posed by her opponent S2. In providing these novel explanations, S1 acquires new beliefs without having to revise her old beliefs, in particular those beliefs that belong to the core set of her position, which comprises the statements that she accepts and her opponent rejects.






Type (β) represents a case of what might be called ‘growth of knowledge within the position.’ The opponent’s challenge initiates the discovery of new explanations, and new reasons without having to give up the propositions that are questioned by the opponent. (γ) Suppose that the positions P1 and P2 held by two experts S1 and S2 respectively are distinguished among other things by the fact that the core set of P1 includes {p, q, r...} that S2 does not accept. Difficulties and unanswered questions revealed by the arguments presented by S2 are within the scope of what S1 regards as the desired goals of her position, of what from her own standpoint she has to address in order to defend her position. It is not really possible to resolve these difficulties without eliminating some of the propositions that belong to the core set of the position.

The epistemic relevance of controversies is most notable in the epistemic changes of type (β) and type (γ) that are motivated by arguments belonging to type (B). The most interesting class of such changes includes changes that are motivated by arguments that belong to type (B); that is, arguments that render a position internally relevant to an opposite position. A unique kind of epistemic change is generated by arguments which the persuasive resolutions they offer to problems recognized by supporters of a given position as internal problems of their positions provide direct support for the opposite position. However, if the arguments used by S2 belong to type (A), the result will be different. These arguments point out the difficulties involved in S1’s position without rendering S2’s position relevant to S1’s position, and therefore, without providing direct rational support for S2’s position. Equipped with the above distinctions, it is now possible to point out the interesting feature exhibited by a modified concept of a Kantian antinomy, that is, an antinomy conceived as a polemical phenomenon. A modified Kantian antinomy is a case in which arguments used by representatives of P1 against P2 and by representatives of P2 against P1 both belong to type B. In other words, an antinomy is a case in which (distinct) arguments used by S1 and S2 challenge some of the presumptions that are part of the respective core set of each of the respective two incompatible positions P2 and P1. By virtue of the arguments used by each contender, they are recognized to be internally relevant to the contrary position. The burden of proof lies, in this case, with both sides. This type of antinomy could be viewed as a pragmatic antinomy. As we shall see in the following chapters, interpreting a polemical exchange on a given metaphysical issue as an antinomy in the above sense is a useful rhetorical move that motivates the introduction of significant epistemic changes by means of an argument that aims to resolve the antinomy. There is more than one type of response to an antinomy, depending on how the antinomy is interpreted.

Chapter 2. The idea of controversy and metaphysics 

The most modest type consists in the attempt of one of the contenders to reinterpret the contentions used in her opponent’s argument that challenges her presumptions in a way that coheres with her conceptual framework. A second type of response introduces new concepts and distinctions that provide better support for one of the existing positions. A third type of response consists in the introduction of a new position P3 that differs both from P1 and P2. The new position includes some new core propositions that are not acceptable to S1 or S2. This kind of response, which could be called ‘the revisionary response,’ characterizes Kant’s argument for Transcendental Idealism. In Part Two of this book, I will examine controversies in which the response to conflicts interpreted as antinomies belongs to the first and to the second type. In Part Three, I will examine the revisionary type of response. However, before addressing these interesting issues, we must examine the notion of intuition in metaphysics and metaphysical controversies. The next chapter is dedicated to this task.

chapter 3

Intuitions, thought experiments, and controversies 1. Introduction Intuitions are indispensable to metaphysics. Yet metaphysical intuitions often conflict, and are prone to generate metaphysical controversies. Surprisingly, the nature of conflicts between intuitions and their relevance to metaphysics has been overlooked by current metaphysicians.1 The minor attention paid to the role of conflicting intuitions in metaphysics is presumably due to the belief that such conflicts merely indicate the fallibility of intuitions. Indeed, most of the topics with which metaphysics is concerned do not belong to the contingent kind. The theories regarding ‘substance,’ ‘property,’ their connection to ‘possibility,’ ‘necessity,’ ‘actuality,’ ‘existence,’ ‘being,’ ‘causation,’ ‘identity,’ and so forth, do not merely hold in this or that possible world.2 But intuitions may be the grounds for false contentions even when the judgments that they motivate are not about contingent states of affairs. One of the conflicting statements or beliefs that are based on intuitions may be fallible in the sense that it does not express a possibility. The fallibility of metaphysical intuitions is indeed significant for an account of their role in metaphysical controversies. Yet, fallibility does not suffice to clarify the role of conflicting intuitions in controversies. In order to clarify what is at stake let us begin by noting that, in many cases, conflicts between intuitions are conflicts between the articulation of intuitions by metaphysicians that have equal claim to the supposedly shared ordinary conception.3 Given that intuitions constitute what seems to us most obvious and central, 1. There has been much philosophical discussion regarding the role of intuitions in philosophy and in the epistemology of modality. See in particular the collection of papers edited by M. De Paul and W. Ramsey (1998), and Levine (2004). However, there has been almost no discussion of the nature of conflicting intuitions and their role in metaphysics. Yablo (1993) is an important exception. 2. It could be added that these terms are ‘semantically stable’ in Bealer’s sense (1999: 23). An expression is semantically unstable if the external environment makes some contribution to its meaning. It is semantically stable, if the external environment makes no such contribution. 3. In some cases, these are conflicts between the intuitions of one and the same subject.




conflicts between philosophers that have equal claim to the supposedly shared ordinary conception may lead to radical skepticism regarding the very possibility of metaphysics. Conflicting metaphysical intuitions seem to jeopardize the very foundations of metaphysics as rational and objective enterprise.4 How can one know that one’s own intuitions are reliable if one lacks the means to rule out the reliability of the contrary intuitions that have equal claim to the (allegedly shared) ordinary conception, that is, the reliability of what seems most obvious and central to someone else? Given this type of conflict, if intuitive seeming is not open to critical and rational arguments, metaphysical knowledge is not possible. In the next sections, I will attempt to clarify how the above puzzle could be resolved by examining the similarities and differences between intuitions and beliefs. I will assume that cognitive presumptions are belief-like states. I will then examine the role of a unique kind of metaphysical thought experiment, with which I will be concerned in other parts of this book. I will conclude this chapter by spelling out the nature of one type of progress exhibited in metaphysics, which is inherently linked to metaphysical controversies. 2. Intuitions and beliefs According to Bealer (1996a, 1996b), intuition is propositional seeming.5 Regardless of whether or not all intuitions are states that have propositional content, it is clear, I suppose, that the intuitions relevant to our present concern motivate assertions that have propositional content.6 How are intuitions related to beliefs that have propositional content? According to Jackson, the purpose of conceptual analysis in serious metaphysics is to reveal our shared ordinary conception. “But how should we identify our ordinary conception?,” he asks, and goes on to answer: 4. Cf. Yablo 1993: 37–38. 5. Not all metaphysical intuitions are propositional. Some of them concern entailments. For example, the knowledge intuition related to the knowledge argument, which will be discussed in detail in Part Two, consists in the entailment of the statement that Mary does not know what it feels like to see red tomatoes by the description of Jackson’s imagined situation. But the notion of propositional seeming suffices to clarify the distinctions with which I am concerned in the present section. 6. Some philosophers distinguish between intuitions of objects and intuitions of facts or propositions. For example, Kant’s explicit notion of intuition seems to be of the former kind. In the Critique of Pure Reason, the term ‘intuition’ (Anschauung) is defined as a singular and immediate representation of an object (A 320/B 377). I will not discuss this distinction in the present context.

Chapter 3. Intuitions, thought experiments, and controversies 

The only possible answer... is by appeal to what seems to us most obvious and central about free action, determinism, belief, or whatever, as revealed by our intuitions about possible cases. Intuitions about how various cases, including various merely possible cases, are correctly described in terms of free action, determinism, and belief, or, as it is often put nowadays, our folk theory of them. (Jackson 1998: 31)

In Jackson’s view, intuitions are my intuitions or yours: “...my intuitions about possible cases reveal my theory...likewise, your intuitions reveal your theory”. (1998: 32) But how can they form the basis of an objective endeavor, as Jackson believes they do, if intuitions reveal my theory or yours? His answer is: To the extent that our intuitions coincide, they reveal our shared theory. To the extent that our intuitions coincide with those of the folk, they reveal the folk theory. (1998: 32)

This characterization apparently allows for the possibility that different people’s intuitions may either coincide or fail to coincide. My intuitions or your intuitions may reveal our shared conception if they happen to coincide, and if they coincide with the folk conception. But there is no assurance that they will coincide. Conflicts between intuitions are conflicts between what “seems to us most obvious and central.” It might be suggested that such conflicts must result from differences between the explicit responses and the intuitions themselves. This appears to be a promising suggestion, though one must carefully examine what it involves. There are indeed cases in which it is possible to detect differences between a person’s intuitions and her responses. For example, according to Jackson, A person’s first-up response as to whether something counts as a K may well need to be discounted. One or more of: the theoretical role they give K-hood, evidence concerning other cases they count as instances of K, signs of confused thinking on their part, cases where their classification is, on examination, a derivative one...their readiness to back off under questioning, and the like, can justify rejecting a subject’s first-up classifications as revealing their concept of K-hood. (1998: 35)

Some of the overt responses of a given subject may conflict with her overall responses. In Jackson’s view, ...extracting the cases that count as K’s from a person’s responses to possible cases is an exercise in hypothetico-deduction. We are seeking the hypothesis that best makes sense of their responses taking into account all the evidence. (1998: 36)

Clearly, a person may fail to grasp how her responses are connected. It is questionable, however, whether intuitions can be uncovered by conducting the above




type of investigations. Such investigations seem to presuppose that intuitions can be revealed by examining a given person’s responses up to a given time. But the above sort of tests cannot distinguish between one’s immediate responses that are expressions of beliefs or habits and the articulation of intuitions.7 Intuitions, however, should be distinguished from beliefs and habits.8 Consider, for example, Gettier cases. Before being exposed to Gettier examples, most epistemologists believed (or presumed) that the necessary and sufficient conditions for knowledge were justification, truth, and belief. After Gettier introduced his examples, most epistemologists changed their beliefs regarding the conditions that constitute knowledge. Clearly, before Gettier, epistemologist did not entertain the belief that the traditional account of knowledge does not suffice for knowledge. Before they were exposed to the Gettier examples, no such dispositional state, belief, or belief like mental states was part of their cognitive system. No such state influenced their deliberations and actions, or was expressed in their utterances. However, before Gettier introduced his examples, did they have the intuition that eventually led them to judge that the traditional account of knowledge does not suffice for knowledge? The question is a bit puzzling, and yet, it is reasonable to suppose that they had the intuition all along.9 Being exposed to Gettier’s examples did not generate the intuition. In this case it is reasonable to suppose that their previous beliefs had changed due to the intuition that they had all along, that is, the intuition invoked or made explicit by Gettier’s argument and

7. It should be added that it is difficult to explain modal errors. As Yablo notes (1993: 32), “What makes us hesitate is not that conceiving can sometimes lead astray, but that we have so little idea how this happens.” 8. Bealer also distinguishes between intuitions and beliefs. But in Bealer’s view, the ground for the distinction consists in the fact that “... belief is not aseeming; intuition is. For example, there are many mathematical theorems that I believe (because I have seen the proofs) but that do not seem to me to be true and that do not seem to me to be false; I do not have intuitions about them either way. Conversely, I have an intuition – it still seems to me – that the naive comprehension axiom of set theory is true; this is so despite the fact that I do not believe that it is true (because I know of the set-theoretical paradoxes).” (1996a: 123) In other words, according to Bealer, the feature that distinguishes intuitions and beliefs is the plasticity of beliefs. My reasons for distinguishing intuitions and beliefs differ from Bealer’s reasons. Intuitions, I maintain, are not cognitive states similar to beliefs, and they are open to change. 9. I am not suggesting that whenever a change in philosophical or metaphysical beliefs is discernable, the intuition was there all along. As I will clarify later, there are cases in which the intuition is missing before the argument.

Chapter 3. Intuitions, thought experiments, and controversies 

the examples he employed in it.10 Clearly, if they had the intuitions and did not have the belief, then beliefs should be distinguished from intuitions.11 3. The fallibility of intuitions The temptation to confuse intuitions and beliefs is presumably due to the features they share. The articulation of intuitions consists in judgments that are naturally accompanied by the formation of beliefs that have the same content. Like beliefs, they seem to have mind-to-world satisfaction conditions, and they are as fallible as beliefs are. Nevertheless, intuitions differ from beliefs. In order to pursue an account of the feasibility of resolutions of metaphysical conflicts, let us first attempt to grasp what constitutes the fallibility of a metaphysical intuition. We might first note that intuitions do not face the tribunal of reality in isolation. Intuitions manifest hierarchical order and complex relations of dependencies that allow one to rule out an intuition that conflicts with other intuitions that seem more central and important. In this sense, they parallel presumptions. Some intuitions are context-dependent in the following sense: Given a particular epistemic and discursive context, subjects are prone to intuitively respond in certain ways. Yet, when one changes the context by conceiving or imagining new possibilities, or by adding new empirical information, the context-dependent intuitions lose their hold upon us. Clearly, context-dependent intuitions are fallible. Context-dependent intuitions differ from intuitions that do not vary from context to context. There is no reason to suppose that all intuitions are contextdependent, even if we arguably lack reasons that prove that some of our intuitions do not vary from context to context. Some intuitions may be stable and will not change when the context changes. But which intuitions belong to the former type, and which belong to the latter type? In some cases, there may be reasons to distinguish between the two types. Yet, given that at any given point of time we do not possess the comprehensive evidence required to determine which intuition is context-dependent and which is context-independent, we might fail to know which of our intuitions is context-dependent and which is context-independent.

10. It was on the basis of this intuition, not on the basis of a definition of knowledge that the Gettier examples were conceived as counterexamples to the traditional analysis of knowledge. 11. Let it be stressed that the fact that we must assume that the intuition was there all along in cases that are similar to the Gettier case does not entail that all changes in what one believes or presumes are changes in which one is bound to assume that the intuition was there all along. I will address this point in what follows.




Discursive contexts also include polemical exchanges. By revealing the inconsistency between one’s intuitive responses that have equal claim to ordinary conception, by pointing out the relevant features of contexts, and by providing conceptual information relevant to the disputed issues, polemical arguments can rationally motivate changes in ways one responds to possibilities. In particular, as the conflicts between metaphysicians’ intuitions that have equal claim to ordinary shared conception indicate – the kind of conflicts with which I am concerned in the present book – we cannot suppose that the allegedly shared conception is coherent. If two conflicting contentions articulate conflicting intuitions, which one is the reliable one? Let us assume that we lack the empirical means to determine which of the conflicting contentions is true. Whether conscious states are identical to physical or functional states can hardly be determined merely by appealing to empirical considerations. Similarly, whether personal identity consists in psychological continuity, bodily continuity, or the continuity of one’s life as a human being is not determined merely by means of empirical information. One may contend that what may determine the matter at stake in such polemical exchanges are possibilities that have not yet been conceived. The introduction of such new possibilities (by means of thought experiments, or by other means) may reveal the connection between one of the disputed intuitions and the shared intuitions. The appeal to new, as yet unconceived possibilities is indeed a significant move in the attempt to resolve conflicts between intuitions. It may support one of the conflicting intuitions and uncover the unreliability or emptiness of the opposed one. But, as many real cases testify, such a move almost never suffices to reveal the unreliability of opposed intuition. The new conceivable possibilities could be interpreted in conflicting ways by also invoking issues that pertain to the deepest level of thought and language. I will address this important issue in detail in Sections 4–6 below. 4. The emendation of intuitions The distinction between context-dependent and context-independent intuitions, on the one hand, and intuitions and beliefs or acquired habits, on the other hand, might suggest that the ordinary shared conception contains the resources to provide a rational resolution to conflicts that do not allow for empirical resolution. This type of resolution shows that the responses of one of the interlocutors express beliefs and/or intuitions that could be ruled out in the face of other intuitions that are already at our disposal. However, some of the most interesting metaphysical conflicts cannot be resolved in this way. As we shall see in the following chapters,

Chapter 3. Intuitions, thought experiments, and controversies 

in these contexts, intuitions seem to pull in conflicting directions due to features of the current scientific and ordinary conception. Such conflicts indicate gaps in the shared ordinary conception, that is, the sense in which it either lacks rational unity or is indeterminate. The articulation of many philosophical or metaphysical problems involves the recognition of such gaps. They are grasped as philosophical or metaphysical problems when one realizes that it is not possible to reconcile the conflict by using the available established means. The recognition of such gaps often (though not always) results from relevant polemical exchanges. A rational resolution of such problems requires a unique type of creative act that goes beyond the state of the ordinary conceptual framework at a given point of time. It often involves the introduction of new concepts and distinctions. It seems that being persuaded or dissuaded by means of arguments involves appealing to one’s intuitions and, yet, in these cases there are no relevant intuitions. Nevertheless, even in these cases, arguments that attempt to resolve the conflict are available. These arguments are not merely based on the appeal to intuitions. They involve the introduction of new concepts and distinctions that are, properly speaking, not part of the shared ordinary conception. The recognition of the conflict that indicates the gap is the first move in such arguments. The other move aims to persuade the addressee that a resolution of the conflict is feasible by means of the new concepts and distinctions that the argument introduces. This type of argument seems to have the features of an argument to the best explanation.12 As noted above, intuitions are not identical to beliefs. Yet, when they are made explicit, they may generate beliefs and presumptions. Moreover, because they are connected to non-epistemic motives and practical concerns, that is, to the variety of institutions and practices that characterize one’s own form of life, intuitions may generate patterns of thought that preclude change despite the persuasive rational force of the argument that advocates it. Although we possess the rational capacity to pursue an impartial viewpoint on a disputed matter, we do not possess a viewpoint that can escape from our dispositions. The rational amendment of intuitions must be carried out from within one’s cognitive system. A single critical persuasive argument rarely suffices to amend intuitions. Nevertheless, we possess the capacity to amend our intuitive responses on the basis of enough persuasive arguments assessed as such from an impartial viewpoint, which attempts to leave out all non-epistemic motives, subjective goals, and so forth. If a person is presented with a sufficient number of persuasive arguments, she is expected to rationally amend her responses. Following Kant, one may assume that awareness of this rational ought is part of our first-personal capacities. Nevertheless, it is in a sense also third-personal: consciousness of 12. See Swoyer 1999.




oneself from “outside” one’s desires and goals. This unique view possessed by rational creatures (and which other animals presumably lack) attempts to abstract from all sorts of particular dispositions without losing hold of the world. This is a point of view from which one may criticize others or even oneself. Yet, there is no assurance that a person would necessarily amend her psychological dispositions as a response to persuasive arguments. The rational power of persuasive arguments is not equivalent to causal power. It motivates or inclines one to freely abandon the old dispositions by operating from within one’s cognitive system by means of rational intuitions, and not by means of imposition. The possibility of modifying intuitions is significant to the controversy-based approach regarding progress in metaphysics. The reality of such changes is corroborated by the history of this discipline, and there is no need to examine whether or not any changes are motivated by metaphysical controversies. Rather, it is important to search for the role of controversies in pursuing metaphysical knowledge. There are two questions that arise in this context. The first one concerns the features of the epistemic contexts within which metaphysical investigations are carried out. The second question is whether the unique features of these epistemic contexts entail that metaphysics has unique subject matter. 5. Intuitions and the role of imagined ideal cases One significant feature of the epistemic contexts that are unique to metaphysics – in particular to the metaphysics of mind – is that investigations that belong to this type often appeal to intuitions aroused by thought experiments that involve imagined ideal or technically impossible cases.13 Clearly, metaphysical investigations that aim to reveal the theories implicit in our ordinary conceptions are not limited to these types of cases. They also make use of ordinary examples. Nevertheless, in many cases, some of which will be examined in what follows, the imagined ideal or technically impossible cases serve as paradigms around which a complex discourse evolves.14 Where metaphysical intuitions are concerned, the imagined case, the assertions that it motivates and the assertions it entails seem 13. By ‘technically impossible cases’ I mean cases the possibility of which is conceivable though we do not possess the technology to realize them. 14. Consider, for example, Descartes’s dream argument, Locke’s example of the cobbler and the prince, Shoemaker’s Brown-Brownson case, and the imagined ideal case involved in Jackson’s knowledge argument. All of these represent imagined ideal or technically impossible cases around which a whole discourse evolves. In Parts Two and Three I will examine the controversies that evolved around some of these cases.

Chapter 3. Intuitions, thought experiments, and controversies 

to comprise one whole.15 Yet, although the intuitive response to an imagined ideal or technically impossible case is linked to the assertion that it motivates and to the entailment of other assertions, each appears to be logically independent of the other. The imagined case is represented with the intention of rendering some interpretation of an ordinary sentences true in the imagined situation. But the imagined situation seems to contain “more” than what is graspable merely on the basis of the concepts of the intended interpretation that guide the act of representing of it. Though contenders may agree that the given sentence is true in the imagined case, they may disagree as to which sentences are entailed. Leaving out confused thoughts, or the inaccuracy of slapdash responses, these differences reflect conflicting interpretations that cannot all be true. Nevertheless, it is not possible to determine merely on the basis of the imagined case which propositions are expressed by the ordinary sentences. For example, consider Jackson’s knowledge argument, which will be discussed in detail in Part Two. Jackson invites his readers to imagine that Mary, an exceptionally brilliant scientist who is locked in a black and white room in which she has spent her whole life until the day she is released, already knows all the facts about the physical information related to the perception of objects.16 Knowledge of all the relevant physical facts is apparently an ideal possibility. Given this imagined ideal case, one is asked to consider the question whether Mary learned anything new when, upon her release, she saw the red color of a tomato. Imagining this case in the context depicted by Jackson seems to motivate the acceptance of the following statement:

15. As I noted above, in Bealer’s view (1996b: 4–5, 2002: 74) a philosophical intuition is a propositional seeming. Seeming is understood “not in its use as a cautionary or “hedging” term, but in its use as a term for a genuine kind of conscious episode. For example, when you first consider one of de Morgan’s laws, often it neither seems to be true nor seems to be false. After a moment’s reflection, however, something happens: It now seems true; you suddenly “just see” that it is true. Of course, this is intellectual seeming, not sensory or introspective seeming.” (1996b: 5) He distinguishes between physical intuitions (When the house is undermined, it will fall) and a priori intuitions (Something cannot be true and false at the same time). In Bealer’s view, “a priori intuitions about hypothetical cases are often being erroneously called thought experiments.” (ibid.) He seems to believe that thought experiments involve physical intuitions (intuitions about laws of nature and about the physical behavior of objects) and that the former should be distinguished from a priori intuitions. As I hope to convince the reader in the following chapters, at least where metaphysical thought experiments are concerned, it is not possible to separate physical intuitions from a priori intuitions. 16. See Jackson 1982, 1986.




(1) A person may know all the relevant facts about the physical features of the perceptual mechanisms involved in the perception of physical objects, but will nonetheless learn something new upon seeing red objects for the first time.

Many philosophers agree that (1) is true. Yet, as the controversies that evolved around the knowledge argument clarify, they disagree on their interpretation of (1). Philosophers tend to provide conflicting answers regarding whether (2) is entailed by (1).

(2) Experiences instantiate non-physical (qualitative) features.

Some philosophers believe that (2) is entailed by (1) while others deny this. Jackson’s thought experiment is in fact contrived to motivate one to entail (2) from (1). But the majority of philosophers deny that (2) is entailed by (1). Physicalists deny this entailment, even if they accept that (1) is true. Yet even they disagree on the interpretation of (1). For example, they may disagree as to whether the qualitative features of experience are representational (physical) features of experienced objects and states of affairs, or intrinsic non-representational (physical) features. Assuming that contenders have a claim to the meaning of ordinary sentences, such conflicts between interpretations at least indicate differences in their beliefs or presumptions regarding the meaning of (1). Clearly, knowledge of meanings must include the ability to correctly apply terms to objects and properties in a variety of contexts. Yet, such knowledge does not suffice to rule out conflicting interpretations.17 We do not seem to lose touch with ordinary language. Yet, as I noted, philosophers may agree that some interpretation of a given sentence is true in the idealized, imagined context, that is, an interpretation that seems to be the ordinary one, and nevertheless disagree as to which statements are entailed. Their conflicting assessments of (1) and (2) indicate that differences regarding the use of words in these contexts are not dichotomous. Although they seem to agree on the use of some sentences (namely, that the intended sentences that depict the imagined case are true), they do not agree on what is entailed by it, and, in many cases, cannot really state precisely how their conceptions differ. Their attempts to produce arguments for and against (2) are meant to rule out the conflicting interpretations and to explicate what (1) involves, not by means of an arbitrary 17. Some philosophers, most notably Kathleen Wilkes (1988), believe that many metaphysical thought experiments are flawed due to the fact that the relevant context of the thought experiment is not determined. In contrast, it could be argued that good metaphysical thought experiments are as determined as they could possibly be. They represent possible situations by using the relevant available conceptual means, even if they do not suffice to rule out conflicting interpretations.

Chapter 3. Intuitions, thought experiments, and controversies 

stipulation of what it should mean but, rather, by means of the painstaking and creative act of analysis. As we shall see in what follows, this is highly significant for grasping the role of polemical dialogues in metaphysics. 6. Objectual imagining, propositional imagining, and imagining ideal cases Two interconnected questions about intuitions arise in these epistemic contexts. The first question concerns the type of epistemic role of the imagined ideal case in the thought experiment with respect to the judgments and entailments that it motivates. The second question concerns the fact that metaphysical investigations revolve around ideal possibilities or technically impossible circumstances that can be merely imagined, that is, that leave ordinary use behind. I will consider these questions in turn. In imagining ideal possibilities, one is guided by the use of sentences of a particular natural language. But as is clear from what I have said so far, the imagined case does not merely serve to make what is conceivable independently of the act of imagining more vivid.18As Yablo notes (1993: 26), we may distinguish between propositional imagining and objectual imagining; that is, between imagining, say, that the tiger is behind the curtain and imagining the tiger itself. As Yablo adds, only the former has alethic content – the kind that can be evaluated as true or false –, while only the latter has referential content – the kind that purports to depict an object (ibid.). I suggest that imagining a situation is objectual imagining, which is the type of imagining relevant to modal and metaphysical inquiries. But it is guided by propositional imagining. In imagining situations, we may imagine them as determined by the concepts that guide the construction of the situation without imagining all the properties instantiated by the objects involved in the imagined situation. The fact that the understanding guides the imagination in representing the imagined case might blur the gap between the imagined situation and propositional imagining. The imagined case is represented with the intention of rendering a given proposition true in that context. Nevertheless, the imagined case seems to allow for a gap between one’s beliefs as to what the sentences that one is using mean and what these sentences actually mean. The gap could be explained by supposing that the ordinary conception implicit in ordinary knowledge of 18. Let it be noted that imagination should not be confused with a faculty of our mind that inherently involves sensible presentations. Though it may also involve such presentations, it doesn’t have to involve them. It should rather be conceived as the capacity to represent what is neither remembered nor perceived “as if” real.




meaning is not fully determined, even if it may contain clues or restrictions of how one has to pursue its completion. 7. Imagined ideal cases and perception The differences between imagining ideal or technically impossible cases and propositional imagining are helpful in explaining the clash between philosophers’ responses to possibilities. The imagined situation seems to provide a shared background that is similar to that provided by the perceptions of objects and states of affairs. This background allows different interlocutors to rationally debate issues that are of a most abstract nature. Nevertheless, it is important not to confuse the role of imagination with that of perception. To imagine, even objectual imagining, is not equivalent to perceiving “as if.” When one imagines a brilliant scientist who knows every physical fact about the perception of red objects, the imagining person does not “as if” perceive anything. The gap that separates the imagined case and the concepts that guide the imagining of the case seem to suggest that imagining ideal or technically impossible cases has the same features of perception. The gap appears to be similar to the gap that, according to some philosophers, separates the conceptual capacities and the sensible capacities involved in perception. But a gap between concepts and perceptual information, if such a gap exists, is not the same as a gap between intended interpretations of sentences and imagined ideal cases. Confusing imagination and perception blurs the epistemic role of imagined ideal cases and the nature of controversies that evolve around them.19 Supporters of underdetermination arguments maintain that perceptual information is open to conflicting interpretations. But even if one grants the claim that the imagined ideal case allows the conflicting interpretations by virtue of its being underdetermined, in these contexts, underdetermination cannot be the same as that involved in perception. In the perceptual case, sensible data, which is what words are about, is underdetermined. In contrast, the imagined ideal cases may allow conflicting interpretations due to the fact that meaning is partly indeterminate. The fact that the choice between the conflicting interpretations is not “merely arbitrary” indicates the differences between the articulated beliefs involved in the intention to represent an ideal imagined case and the content of what one represents. Though these differences seem to be similar to the differences between beliefs about the content of perception and the content of perception, it is more reasonable to suppose that these differences are related to the fact that in imagining 19. Modal mistakes are different from perceptual mistakes. See Yablo (1993: 32–33).

Chapter 3. Intuitions, thought experiments, and controversies 

an ideal case one represents situations that are intentionally represented as involving concrete individuals and properties though, properly speaking, such individuals are not given to one as they are given to us in perception. The appeal to imagined ideal cases, contemplating their features by linking them to ordinary states of affairs, seems to be a source of information that guides the attempts to look for a better defense of a given position. But reflective contemplation of ideal possibilities may be informative, sometimes contrary to one’s expectations. It may reach beyond the intended sense, even beyond shared ordinary knowledge of meaning. It allows conflicting interpretations without rendering the choice between them arbitrary. And yet, one cannot decide on a disputed matter merely by contemplating the features of the imagined case. These reflections suggest that the interplay between language and the capacity to imagine ideal cases operates in two opposite directions. Imagination is guided by concepts, but it may also inform our beliefs regarding the meanings of words. The interplay between imagination and conceptual capacities seems to be consistent with a central feature of meaning. Yet, the apparent similarity between cases in which the appeal to imagined ideal cases informs our beliefs and cases in which beliefs regarding meanings change in the face of new perceptual and scientific information might be a source of confusion. I assume that it is widely agreed that explicit beliefs regarding meanings are open to various kinds of changes. Yet this feature is usually depicted as linked to progress in the empirical sciences, to etymological investigations, and so forth. Here again one must be cautious not to confuse the empirical and the metaphysical context. Assuming that perceptual objects are fully determined, one can maintain that ‘water’ now refers to the same liquid that filled the oceans and lakes 2000 years ago, even if our beliefs about the meaning of ‘water’ differ from Aristotle’s (false) beliefs. Yet, the type of changes relevant to the differences between our beliefs about the meaning of a natural kind term such as ‘water’ and Aristotle’s beliefs is not of the type of changes that are relevant to metaphysics. Metaphysics is not unconnected to empirical knowledge (rather, as I suggested, it presupposes empirical knowledge). But to the extent that changes in explicit beliefs regarding meanings result from controversies that address idealized cases, the metaphysical case differs from the empirical case. The role of intuitions regarding imagined ideal or technically impossible cases seems to support the view that this type of investigation is an a priori type of investigation. Idealized cases are naturally linked to ordinary cases given that ideal cases are idealizations of ordinary cases. Conflicting intuitions regarding such cases are naturally also relevant to ordinary cases. But the attempt to resolve such conflicts merely by appealing to intuitions regarding ordinary cases might be hindered by the gap that separates idealizations and ordinary cases. Idealizations and technically impossible cases are not empirically testable. Another way




to respond to such conflicts is by means of dialectical arguments, and it is interesting to note that the relevant dialectical arguments usually do not make use of the methods that are used in the natural sciences. The nature of dialectical arguments encourages the view that metaphysics is the a priori layer of science. As we shall see in the next section, there are additional reasons to hold this view.20 8. Logical matters and “matters of fact” The differences between the role of perception in empirical knowledge and the role of imagined situations in metaphysical thought experiments may lead to the conclusion that metaphysics is not a genuine branch of science. In the present context, I will not attempt to argue for the scientific adequacy of metaphysics. Rather, like those that practice metaphysics, I assume that it is indeed adequate. In the present section, I will clarify the sense in which the themes of metaphysics are unique, as revealed by the conflicting intuitions that motivate metaphysical controversies. Does metaphysics deal with unique issues? Consider the issues involved in the metaphysics of mind. Metaphysical inquiries appear to be about the same topics studied in psychology, the cognitive sciences, and the brain sciences, which are questions regarding the ontological character of the human mind, the nature of experiences, representations, and personhood, and other related issues. The uniqueness of metaphysics cannot merely consist in the claim that it is concerned with the most basic features of the entities investigated by psychologist, cognitive scientists, and neuroscientist. There seems to be no reason to suppose that the empirical sciences are not concerned with the deepest layers of the domain that they investigate.

20. Let it be noted that in order to endorse this view, one need not hold the assumption that conflicts between interpretations of ideal cases are in principle not resolvable by empirical means. An interpretation that is linked to empirical explanations may be favored over a conflicting one. We have no a priori assurance that such empirical resolution will not be available. Nevertheless, we have no assurance that a link to empirical explanations will be available, and in some cases it even seems highly difficult if not impossible to conceive of the feasibility of such a resolution. At any event, the actual practice of metaphysicians cannot really be described as the search for empirical explanations that may resolve metaphysical conflicts. The progress manifested by metaphysician’s intellectual activities is mainly the result of contemplation and of the dialectical exchange with other philosophers.

Chapter 3. Intuitions, thought experiments, and controversies 

There is, however, one feature that is most notable in the chains of controversies that have unfolded in the philosophy of mind, which seems to be particularly important in clarifying the sense in which metaphysical investigations differ from empirical investigations. In the first section of this chapter, I noted that metaphysics is not concerned merely with what is true in a particular possible world. Yet, given the variety of available dualist and physicalist positions, if any of these positions is true in our world, then it is contingently true, that is, true in our world without ruling out the possibility of its being false in other worlds.21 This is, to be sure, a controversial position. Nevertheless, it seems to be consistent with the supposition that if metaphysics involves questions such as those related to the mind-body problem, it does not have a unique field of inquiry. The only difference between metaphysical investigations and relevant scientific investigations, if there is such a difference, is a methodological one. Some features of this field do not allow one to determine which of the competing positions has better chances of being true on the basis of empirical data. Metaphysical investigations about the nature of our minds are, therefore, investigations about factual matters. Yet, as is familiar to anyone acquainted with metaphysics, it must be noted that arguments for or against a metaphysical position inherently involve the appeal to logical matters, such as issues that address our modal concepts and logical concepts. To be sure, the logical, semantic, and modal theories are debatable on their own right. The fact that they motivate the acceptance of a given metaphysical position is considered in some cases as a reason to argue for or against the logical theory itself. This mixture of “matters of fact” (Are persons identical to their bodies?) and logical matters (What is the nature of identity?) is unique to metaphysical investigations and to the metaphysical intuitions invoked in them. I suggest that the uniqueness of the metaphysical intuitions related to the metaphysics of mind consists in, among other things, the inherent connection between matters of fact represented as imagined ideal or technically impossible cases and implicit or explicit commitment to a debatable logical theory. This explains why conflicts between attempts to articulate metaphysical intuitions are not merely conflicts about possibilities that could be classified as belonging to the realm of “matters of facts,” but are also conflicts that tend to expand to one’s logical intuitions.22 The differences between conflicting metaphysical intuitions cannot be explained by noting that intuitive responses to imagined cases express possibilities that are true in some world but are false in our world. Here the concept of possibility faces its limits of intelligibility. This further supports the view that metaphysics is not just an empirical discipline. The problem is not merely that we 21. In particular, see Chalmers 1996; Jackson 1998. 22. As shown in Chapter 1, this feature is central to Kant’s implicit argument.




lack the proper empirical means to resolve the disputed matter. If arguments for and against the nature of some “matters of fact” involve conflicting attempts to articulate intuitions regarding our logical concepts, they cannot be merely founded on empirical means.23 9. Types of responses to conflicts of intuitions Let us attempt to relate the issues discussed here to those discussed in the previous chapter. Recognizing a conflict between intuitions may lead to at least four types of responses: a. One of the contenders presents arguments that aim to disavow an intuition by refusing to endorse some features of the thought experiments, or even by presenting the whole activity of conducting such experiments as groundless, that is, by rendering the possibility it allegedly represents and the link between the sentences that guide its depiction incoherent or empty. An example of this type of response is Dennett’s (1991, 2005) refusal to accept that arguments such as Jackson’s knowledge argument or Chalmers’s zombie argument are significant for the account of conscious states.24 The arguments used by such contenders are usually arguments of type (A) discussed in Chapter 2. b. The second type of response endorses the presumption that the ordinary interpretation of the sentence that describes the imagined situation is true. It nevertheless denies the controversial conclusion drawn by the opponent. This is achieved primarily by reinterpreting the relevant sentences with or without adding novel distinctions or forming novel concepts. In the first case, the arguments to which the contenders respond are arguments of type (A) discussed in Chapter 2, and the type of epistemic change involved fits type (α). In the second case, they are arguments of type (B), which fit type (β) discussed in Chapter 2. c. A third type of response also maintains that the sentences that describe the imagined possibility have a true interpretation which is incompatible with that involved in the argument being disputed. Yet, according to this view, the intuitions of the opponent require a deeper explanation than merely refusing to accept it. According to this view, these intuitions are natural though deceptive. They are the cases of a unique type of cognitive illusions. Kant believed that reason is a faculty that naturally involves such illusions which 23. Cf. Janet Levine 2004. 24. I will discuss Dennett’s response to the knowledge argument in Chapter 5.

Chapter 3. Intuitions, thought experiments, and controversies 

we cannot avoid even if we can avoid the mistakes to which such illusions lead. Current examples of this type of response are Papineau’s (2002) notion ‘intuition of distinctness,’ which involves an illusion, Tye’s claim (2000) that the explanatory gap is a cognitive illusion based on our natural cognitive capacities, and Jackson’s similar claim (2004).25 This type of response usually involves arguments of type (A). The epistemic change related to the discovery of the illusion is that of type (β). d. A fourth type of response denies that we can establish one of two conflicting interpretation as the true one. Arguments that establish this conclusion are of type (B). This type of argument is close to Kant’s notion of an antinomy. Arguments belonging to this type of response aim to show that each of the conflicting intuitions is reliable. Nevertheless, given the available conceptual repertoire, it is not possible to provide better support for one of the conflicting views. Regarding the problem of consciousness, the position that maintains that we are unable to know how to resolve the conflict is called ‘mysterianism’.26 Yet mysterianism is not the only option here. As we shall see, if this move is just one step in an argument that motivates one to pursue an alternative theory, it involves the introduction of novel concepts and distinctions; that is, it involves significant epistemic changes. 10. Controversial relevant alternatives I will conclude this chapter by readdressing the notion of epistemic progress discussed in Chapter 2. I will attempt to clarify the notion of progress that suits metaphysics and metaphysical controversies. Arguments used in metaphysical controversies may involve or motivate epistemic changes. What could ‘epistemic change’ mean with regard to the articulation of metaphysical intuitions? A response should consider the type of change actually exhibited in metaphysics. At first sight, it appears that the notion of progress is hardly applicable to metaphysics. No attempt to establish a substantial metaphysical claim, to resolve a disputed contention, or to provide a solution to a metaphysical problem enjoys the status of being generally accepted among metaphysicians. This seems to be as true now as it was in Kant’s time. Metaphysicians more often than not dispute each other’s contentions and theories, and the solutions they offer for metaphysical problems. Nevertheless, the absence of general agreement among metaphysicians does not necessarily indicate a lack of progress. Upon a close examination of the discourse of metaphysics, it 25. I discuss these positions in Chapter 5. 26. See McGinn 1991, 1999.




seems that even though metaphysicians seem to argue about almost everything, they nevertheless seem to agree, either implicitly or explicitly, on the importance and significance of a novel argument that introduces a novel problem, a novel way of presenting a problem, or a novel resolution of a metaphysical problem. They seem to agree that a given argument establishes a position as a relevant alternative to their own position, even if they deny the intuitions, do not share the view that the revealed problem is a genuine one, or do not agree with the resolution offered. Hence, it is important to distinguish between the significance attributed to such arguments and the acceptance of what they aim to establish. Such a distinction may serve to clarify a notion of progress that is relevant to metaphysics. Progress does not mean differences between statements generally assumed to be true by the members of a given research community at a given point of time t and statements that are generally accepted as true by members of the same community at a later time t’. Progress in this context does not involve changes in the general acceptance of a statement or a theory, but rather changes in the general acceptance that a given position enjoys the status of a relevant alternative. A controversial relevant alternative is not a generally accepted statement or theory, but rather a position supported by an argument or set of arguments that articulate metaphysical intuitions. It is generally recognized in the relevant community as being relevant and significant for the rational evaluation of the rational support that is provided by the articulation of metaphysical intuitions by adherents of other positions. A controversial relevant alternative is relevant to the account of the topics at hand, regardless of whether the conclusions of the arguments are accepted or denied, or even of whether one disapproves that the argument is sound. Controversial relevant alternatives effectively transform the horizon of real possibility. They are therefore recognized as important, even though they are nevertheless disputable. Clearly, a relevant alternative must have a certain degree of coherence. It must be able to shed light on the phenomena addressed, and it should be consistent with the background of the research community. The relevant alternative has competing alternatives. What renders the alternative relevant is that one has to consider it when establishing one’s own position, even if one’s position conflicts with the conclusion provided by the relevant alternative, or with its premises. Accepting a position as a relevant alternative at least implies recognizing that its intuitive appeal indicates real features of our ordinary conception, whether or not one accepts its articulated contentions. I suggest that establishing a novel position as a relevant alternative or an existing position in a novel way by means of persuasive arguments constitutes epistemic progress. The arguments that can account for the role of a relevant alternative in metaphysics are successful arguments that fall under type (B), which was discussed in Chapter 2. Contenders that respond to what they regard as relevant alternatives

Chapter 3. Intuitions, thought experiments, and controversies 

(at least implicitly) represent them as presenting a type (B) argument, which is an argument that persuasively connects some features or states of affairs the existence and importance of which they accept as uncontroversial, and a substantial proposition that they intuitively reject. Not all arguments invoked in metaphysical controversies address an opposed position as a relevant alternative. Arguments can be used in order to undermine the position’s implicit claim to being a relevant alternative. But as we shall see below, this type of argument differs in its results from the type of argument that recognizes a position as a relevant alternative. Clearly, no philosopher merely aims to establish her metaphysical position as a relevant alternative. She normally aspires to establish it as the only true position. The ironic relationship between what the metaphysician aspires to achieve by means of successful arguments and what she actually achieves seems to be crucial to this type of inquiry. These ironic relations reveal the gap between one’s intuitions and the rational grounds of one’s presumptions. When one’s intuitions are at stake, one does not allow the opposed alternative to have equal claim to truth. But as far as the normative notion of progress is concerned, progress should also be evaluated in terms of the changes in the status of a position as a relevant alternative for the research community. 11. Conclusion In the following chapters, I will mainly address two types of epistemically effective arguments in metaphysics, namely, arguments that establish a position as a relevant alternative, and arguments that effectively respond to relevant alternatives. To be sure, an effective argument involves persuasive reasons that also motivate the acceptance of a certain position without addressing relevant alternatives, real alternatives or merely conceivable alternatives. But its persuasive force also depends on whether the argument can establish a relevant alternative by responding to competing relevant alternatives. To the extent that only one metaphysical theory is true, we cannot avoid the need to respond to competing alternatives. Yet, there is usually more than one effective response to a relevant alternative. Each of these effective responses is incompatible with the others.27

27. Though Hegel underscored the importance of the dialectical and historical characteristics of reason, the role of arguments in establishing and reestablishing positions, and the fact that there is more than one way to respond to a controversial relevant alternative, seems to rule out his position as a possible candidate for explaining the role of controversies in metaphysics.




The issues discussed in the present chapter are rather abstract. However, I hope that my examination of some actual debates will help persuade the reader that the position defended here is a relevant alternative to the current conception of the nature of metaphysical controversies, conflicting articulations of intuitions, and their relevance to an account of the nature of this intellectual field. The following chapters are devoted to this task.

part two

The knowledge argument

chapter 4

The polemical character of the knowledge arguments 1. Introduction In Part Two and Part Three of this book, I will attempt to confirm the ideas presented in Part One by examining some debates in the current metaphysics of mind. In Part Two, I examine the controversies related to the knowledge arguments. Controversies pertaining to the knowledge arguments explicitly address the nature of conscious phenomenal qualities, but, as we shall see, they are not limited to this topic. One disputed issue is the nature of phenomenal properties. Another disputed issue is that of the nature and availability of psychophysical explanations. My concern in Chapter 4 and Chapter 5 will be with debates that address the first topic. In Chapter 6, I will deal with the questions related to mental causation and psychophysical explanations. The term ‘knowledge argument’ is usually associated with the argument in Jackson’s paper “Epiphenomenal qualia”(1982) (EQ). Knowledge arguments in the sense of the term used in this book are arguments that draw ontological conclusions from epistemic properties. Debates on knowledge arguments have a long history that extends beyond the last five decades. Nevertheless, though participants in the recent debate in analytic philosophy of mind sometimes refer to earlier positions in the history of metaphysics, the earliest arguments addressed are arguments that were first presented about 50 years ago. In other words, the recent debate on the mind-body problem is dissociated from the earlier historical phases of debates on the same topics. The recent knowledge arguments are arguments that dispute the feasibility of the most promising philosophical approaches to the mind-body problem presented in the 1950s and the 1960s: the identity theory and functionalism in the philosophy of mind. These were arguments that rendered the old ideas relevant to the main themes of contemporary physicalist positions. These positions were explicitly receptive to the novel scientific ideas that emerged in the 20th century in the computer sciences, cognitive sciences, physics, biology, and related disciplines. Though the issues involved in the arguments that challenged physicalism were not new, the reasons used to support them were novel




reasons that were able to challenge physicalism from within the physicalist’s scientific and philosophical framework. Jackson’s knowledge argument was preceded by at least two famous arguments that challenged the main presumptions of the physicalist and functionalist-reductive approaches to conscious mental states: Kripke’s modal argument presented in “Identity and necessity” and in Chapter 3 of Naming and Necessity, and Nagel’s argument presented in “What is it like to be a bat?”. These arguments seemed to present a new agenda of problems that metaphysicians and philosophers of mind needed to deal with in addressing the ontological features of the mind-body problem. The knowledge intuition – a feature of conscious mental states that was either dismissed or overlooked by most accounts of the mind in the 1950s and 1960s – became the main feature of conscious mental states that any account of consciousness had to address. As I hope to show in the present chapter, the persuasive power of the knowledge arguments can also be traced to their polemical structure. This chapter is dedicated to a short presentation of these arguments. There is a vast literature on each of them. As will be shown in the chapters that follow, they serve as the foundations for the later development of the philosophical discourse on consciousness. I do not intend to cover the complex issues, arguments, and conceptual changes related to these arguments in detail. Rather, I will focus on features that are relevant to the account of the nature of polemical arguments and polemical dialogues. 2. Kripke’s modal argument for dualism as a polemical argument In Naming and Necessity, Kripke’s main concern is with issues related to the meaning of proper names, natural kind terms, and the distinction between epistemic modalities and metaphysical modalities. The first two lectures include arguments that challenge a set of descriptivist semantic theories of proper names, and present an alternative account of the meaning of such terms. The third lecture extends this account to natural kind terms, and presents an argument for mind-body dualism using the results of the first two lectures. Though Kripke’s views on meaning and modality were immediately recognized as an outstanding philosophical achievement, his argument for mind-body dualism that is based on these views was debated by many philosophers. My intention in the present section will be to take a fresh look at his argument for mind-body dualism. I will claim that it establishes his position as a relevant alternative to a large number of physicalist positions that accept Kripke’s semantic and modal theory, and nevertheless wish to reject the dualistic conclusion of his argument.

Chapter 4. The polemical character of the knowledge arguments 

As several responses to Kripke’s account attest, many physicalists tend to view his account of meaning and modality not merely as compatible with their physicalist and scientific positions, but as actually supporting them. For them, the fact that Kripke’s argument supports the opposed metaphysical position is therefore baffling. As I will attempt to show, Kripke’s argument exemplifies the features of the type B argument discussed in Chapter 2. Kripke’s argument is a polemical argument that addresses the identity theory defended by Place (1956), Feigl (1958), and in particular by Smart (1959). Kripke’s argument led to much discussion and attempts to refute it. Although the majority of philosophers of mind do not accept Kripke’s dualistic conclusion, many of them recognized its importance. They recognized the fact that if they wished to defend their position, among other things, this required a response to his argument. I will begin with a very brief review of Kripke’s account of meaning and modality. As Kripke notes, identity statements – statements such as ‘Hesperus is Phosphorus’ – express necessary propositions. Given that identity is a relation between an object and that object itself (and not a relation between the symbols that flank the identity sign), one cannot analyze proper names as having the same meaning as an identifying description, or a cluster of such descriptions. Say that Hesperus is the heavenly body that rises in the evening at XYZ, and Phosphorus is the star that rises in the morning at X’Y’Z’. If Hesperus and Phosphorus are different names of the same star, Hesperus is necessarily identical to Phosphorus (Hesperus is necessarily identical to itself). As Kripke clarifies, if Hesperus and Phosphorus are two names of one and the same heavenly body, one cannot substitute ‘Hesperus’ with an identifying description such as, say, ‘the star that rises in the morning at X’Y’Z’ or with a class of such descriptions. Note the four statements below: 1. Hesperus is Phosphorus. 2. Hesperus is the star that rises in the evening at XYZ. 3. Phosphorus is the star that rises in the morning at X’Y’Z’. 4. The star that rises in the evening at XYZ is the star that rises in the morning at X’Y’Z’. In contrast to the necessity of (1), statements (2), (3), and (4) express contingent propositions. A name is a rigid designator. It refers to the same object in all possible worlds. The fact that in contrast to (1), the sentences in (2) and (3) are not, properly speaking, identity statements, does not entail that identifying descriptions have no role in the account of meaning of proper names. According to one interpretation of Kripke’s views, the role of the descriptions involved in (2) and (3) is to fix the reference of ‘Hesperus’ and ‘Phosphorus,’ not to provide synonyms for these names.




The distinction between the meaning of proper names and the way in which their reference is fixed is linked to another distinction that Kripke discusses in Naming and Necessity, namely, that between apriority and necessity. As Kripke clarifies, the content of a priori knowledge need not be a necessary proposition. Similarly, a proposition could be necessary even if it is not known a priori. (2) can serve as an example of the former possibility. If ‘the star that rises in the evening at XYZ’ is used in (2) in order to baptize the referent of the name ‘Hesperus,’ (2) does not express a posteriori knowledge. Although the identifying description ‘the star that rises in the evening at XYZ’ by which the reference of the name ‘Hesperus’ is fixed expresses a contingent property of the object that the name rigidly designates, the knowledge expressed by (2) is a priori knowledge. By contrast, (1) is an example of the necessary a posteriori. The differences in the ways in which the reference of ‘Hesperus’ and ‘Phosphorus’ are fixed clearly indicate that one could be ignorant of the fact that (1) is necessarily true, and hence not know it a priori. The discovery that Hesperus is the same star as Phosphorus is, therefore, a significant empirical discovery. In Kripke’s view, identity statements could express a posteriori necessary knowledge. This claim appears to support the philosophical theories that maintain that knowledge of the identity of mental states and processes and of physical states and processes could be the result of scientific empirical investigations. Nevertheless, as Kripke aimed to demonstrate in the third lecture of Naming and Necessity, the same ideas that the distinction between a priority and necessity is based on could be used to challenge the identity theory. His argument addresses the identity theory explicitly. One version of the identity theory was presented by Smart in his paper “Sensations and brain processes.” (1959)1 In his defense of the identity theory, Smart begins by admitting that a correlation between mental states and processes, and, on the other hand, physical states and processes does not suffice to establish identity. An object or a process cannot be correlated with itself. He presents the main idea of his version of the identity theory below: Let me first try to state more accurately the thesis that sensations are brain processes. It is not the thesis that, for example, “after-image” or “ache” means the same as “brain process of sort X” (where “X” is replaced by a description of a certain sort of brain process). It is that, in so far as “after-image” or “ache” is a report of a process, it is a report of a process that happens to be a brain process. It follows that the thesis does not claim that sensation statements can be translated into statements about brain processes. Nor does it claim that the logic of a sensation statement is the same as that of a brain-process statement. All it claims is 1.

Large parts of Smart’s paper have the dialectical structure of objections and replies.

Chapter 4. The polemical character of the knowledge arguments 

that in so far as a sensation statement is a report of something, that something is in fact a brain-process. Sensations are nothing over and above brain processes. (1959: 144–145)

The identity theory does not maintain that terms that refer to phenomenal features of experiences are translatable to terms referring to physical processes of the brain. Nevertheless, according to the identity theory, they refer to the same thing. Here one might contend that the epistemic differences between knowledge of, say, the phenomenal features of pain, and one’s ignorance regarding the nature of the physical processes with which pain is supposedly identical provides reason to distinguish between mental states and physical states. Smart responds to this sort of objection as follows: In short, the reply to objection i is that there can be contingent statements of the form ‘A is identical with B,’ and a person may well know that something is an A without knowing that it is a B. An illiterate peasant might well be able to talk about his sensations without knowing about his brain processes, just as he can talk about lightning though he knows nothing of electricity. (1959: 147)

Identity statements similar to ‘pain = C-fiber stimulation’ are contingent statements. Though Smart does not draw the distinction between epistemic contingency and metaphysical contingency here, ‘contingent’ apparently means epistemic contingency here: a person may know that something is A without knowing that it is a B. Hence, according to this position, although ‘C-fiber stimulation’ does not translate ‘pain,’ and although ‘pain = C-fiber stimulation’ is contingent, ‘pain’ and ‘C-fiber stimulation’ necessarily refer to the same features or properties of mental states. One supposition of Smart’s arguments, namely, that there are contingent identity statements, is refuted in the first two chapters of Naming and Necessity. But on first approximation, this seems to be insubstantial to the physicalist’s goals. For Kripke provides convincing reasons to suppose that there are a posteriori necessities, and it seems that this is precisely what the physicalist requires. For if it can be maintained that the necessity of ‘pain = C-fiber stimulation’ is compatible with its being an empirical discovery that ‘pain = C-fiber stimulation,’ then Kripke’s account of the meaning of names and natural kind terms and his distinction between epistemic modalities and metaphysical modalities seems to meet the physicalist’s goals. The apparent contingency of ‘pain = C-fiber stimulation’ appears to be insubstantial to her main thesis. Moreover, the physicalist clearly wishes to hold the view that ‘pain = C-fiber stimulation’ expresses a necessary proposition. It is in this context that Kripke’s argument takes a surprising twist. The theory of the modal and epistemic features of statements of identity is used against the




physicalist’s goal. Kripke first notes that the account of proper names qua rigid designators also applies to natural kind terms. Consider the following statements: 5. Heat = mean molecular motion 6. Light = a stream of photons 7. Water = H2O2 If each of these statements is true, the statement is necessarily true. Nevertheless, each of them expresses empirical (a posteriori) discovery. In each of these cases, the terms to the right of the identity sign are fixed in a different way from the way in which the terms to the left of the identity sign are fixed. ‘Heat’ is presumably fixed by ‘the physical phenomena that causes the sensation of heat’ while ‘mean molecular motion’ is fixed (qua theoretical concept) in a different way. It cannot be known a priori that ‘heat’ and ‘mean molecular motion’ express one and the same property. But, if (5)–(7) are true, then they are necessarily true. Consider the following statement: 8. pain = C-fiber stimulation It might be supposed that statements such as (8), in which the terms to the right of the identity sign refer to physical states, whereas the terms to the left of the identity sign refer to mental states, also express (if true) necessary a posteriori propositions. But according to Kripke, this supposition is false. For in contrast to (5)–(7), in which the reference of at least one of the terms is fixed by an identifying description that contingently applies to the object or property referred to, terms such as ‘pain’ in (8) are not determined by means of identifying descriptions that assign a contingent property to the referent: In the case of the identity of heat with molecular motion the important consideration was that although ‘heat’ is a rigid designator, the reference of that designator was determined by an accidental property of the referent, namely the property of producing in us the sensation S. It is thus possible that a phenomenon should have been rigidly designated in the same way as the phenomenon of heat, with its reference also picked out by means of the sensation S, without that phenomenon being heat and therefore without its being molecular motion. Pain, on the other hand, is not picked out by one of its accidental properties; rather it is picked out by the property of being pain itself, by its immediate phenomenological quality. Thus pain, unlike heat, is not only rigidly designated by ‘pain’ but the reference of the designator is determined by an essential property of the referent. (1980: 152–153)

‘Pain’ is picked out by the property of being pain that, according to Kripke, is ‘its immediate phenomenological quality’. By contrast, the referents of the terms to the 2. Kripke 1980: 129.

Chapter 4. The polemical character of the knowledge arguments 

right of the identity sign in (5)–(7) are not picked out by the essential properties of the referents, but rather by descriptions that fix the reference (‘heat’ is fixed by means of the sensation of heat). This difference motivates an argument for dualism: i. “To create this phenomenon [C-fiber stimulation, Y.S.], it would seem that God need only create beings with C-fibers capable of the appropriate type of physical stimulation; whether the beings are conscious or not is irrelevant here.” (1980: 153) ii. “...to make the C-fibers stimulation correspond to pain, or be felt as pain, God must do something in addition to the mere creation of the C-fiber stimulation; He must let the creatures feel the C-fiber stimulation as pain...” (1980: 153–154) iii. “[Therefore] the correspondence between a brain state and a mental state seems to have a certain obvious element of contingency.” (ibid.) Hence, by using his account of the meaning of proper names and natural kind terms and the reasons for the differences between epistemic and metaphysical modalities, Kripke’s argument draws ontological conclusions regarding the relations between minds and physical entities on pure philosophical grounds. It is important to note that Kripke’s argument could also be represented as a polemical argument against the identity theorist. As a polemical argument, its first premises express presumptions that he shares with identity theorists such as Smart: iv. ‘Pain’ is not translatable to ‘C-fiber stimulation.’ v. Given that ‘pain = C-fiber stimulation’ is an identity statement, if it is true it is necessarily true. vi. The knowledge that ‘pain = C-fiber stimulation’ is a posteriori. If true, it stands for an empirical discovery. But in addition, Kripke claims that the very reasons related to the intuitions that are the basis for the acceptance of (iv-vi) also require one to accept (vii), which is manifestly incompatible with the identity theorist’s goals: vii. The features of necessary a posteriori identity statements between natural kind terms, statements that express empirical discoveries, cannot be satisfied by psychophysical identity statements similar to ‘pain = C-fiber stimulation.’ Hence Kripke’s argument has all the features of a type B argument. By entailing a dualistic conclusion from presumptions that are acceptable to the physicalist – of course, a physicalist sympathetic to his account of meaning and modality – he renders the Cartesian dualist’s position directly relevant to the identity theorist’s position which is incompatible with it.3 Kripke’s argument at least 3. Clearly, the argument is ineffective with respect to identity theorists that are unmoved by Kripke’s modal intuitions and by his modal theory, such as Smart 2007.




reveals the tension between the identity theorist’s modal intuitions and her metaphysical intuitions. As we shall see in what follows, some philosophers who accept the ideas involved in Kripke’s semantic theory nevertheless deny the soundness of his argument for mind-body dualism for a variety of reasons. At the present point, let us note that the argument is open to some objections. First, it appears that according to Kripke the term ‘pain’ designates only an ‘immediate phenomenological quality.’ Nevertheless, ‘pain’ also seems to refer to a functional property defined by the causal role it has in the cognitive system of the creature that experiences it, that is, by its role in producing a certain kind of behavior given certain perceptual stimulations and given the creature’s other mental states. In addition, even if the term ‘pain’ represents an immediate phenomenological quality, in normal cases, it also represents physical features of the bodies and nervous systems of those that experience pain. To be sure, this does not rule out the possibility that ‘pain’ represents a functional property, physical features of the bodies of those that experience it, and a certain type of immediate phenomenological property. But if ‘pain’ designates all of them, one has to face the task of responding to the question of whether the functional property and the immediate phenomenological property are one and the same property, or whether the term ‘pain’ as it is ordinarily used, is in fact ambiguous. It should be noted that in Naming and Necessity, Kripke did not present any argument to the effect that the immediate phenomenological quality designated by ‘pain’ could not be identical to a functional property. His argument for mental-physical dualism presumes rather than demonstrates the conceivability of the claim that C-fibers stimulation could exist even if the immediate (conscious) phenomenological property of pain does not exist and vice versa. What seems to be missing in Kripke’s account is an answer to the question of why the fact that ‘pain’ is picked out by an essential property of its referent, which is the immediate phenomenological quality, rules out the possibility that this phenomenological quality is in fact identical to a physical property of the brain, be it a functional property, or another type of property. Another feature that seems to be missing in Kripke’s argument is the link between the phenomenological quality designated by ‘pain’ and its inherent firstperson perspective. ‘Immediate’ in ‘immediate phenomenological quality’ seems to intimate ‘first-person access.’ But Kripke’s argument leaves the precise role of the first-person access to conscious mental states (such as pain) somewhat unclear. 3. Nagel’s argument Kripke’s argument for mind-body dualism in Naming and Necessity presupposes a background of commitments. Two features of this background need to be

Chapter 4. The polemical character of the knowledge arguments 

pointed out: the realist (in contrast to anti-realist) stance regarding meaning, truth, and objectivity, and the alleged privileged role that the physical sciences have in one’s claims to truth and objectivity. It seems almost pointless to state these preconditions of his argument, since they are apparently shared by all those interested in it. However, they are not shared by philosophers from other traditions: idealists of various kinds, Hegelians, phenomenologists, and even by some philosophers from the analytic tradition. The background of agreements of his argument for dualism therefore effectively delineates a relevant addressee. The background of agreements of the discourse on dualism in recent analytic philosophy of mind is further clarified when one examines Nagel’s argument in “What is it like to be a bat?.” (1974) Nagel’s argument is not based on a complex account of proper names and natural kind terms. Also, in contrast to Kripke, he does not leave out the fact that mental states are also entities involved in the causation of behavior: I do not deny that conscious mental states and events cause behavior, nor that they may be given functional characterization. I deny only that this kind of thing exhausts their analysis. (1974: 437)

Two facts are left in the margins of Kripke’s argument, that immediate phenomenological qualities are properties of conscious mental states (and not of mental states as such) and that the awareness of these features involves the first-person perspective. These facts become the explicit topics of Nagel’s argument. The immediate first-person awareness of the qualitative character of experience that, in Nagel’s view, determines the subjectivity of experiences is claimed to be inexplicable within the conceptual resources of the physical and biological sciences. But according to Nagel, subjectivity is at least in some sense a feature of creatures that are also regarded as biological creatures: While an account of the physical basis of mind must explain many things, this appears to be the most difficult. It is impossible to exclude the phenomenological features of experience from a reduction in the same way that one excludes the phenomenal features of an ordinary substance from a physical or chemical reduction of it – namely, by explaining them as effects on the mind of human observes. If physicalism is to be defended, the phenomenological features must themselves be given a physical account. But when we examine their subjective character it seems that such a result is impossible. The reason is that every subjective phenomenon is essentially connected with a single point of view, and it seems inevitable that an objective, physical theory will abandon that point of view. (1974: 437)

As the details of Nagel’s argument clarify, the single viewpoint is one of a certain type of biological or physical creature. Interestingly, although Nagel’s argument

 Controversies and the Metaphysics of Mind

aims to persuade his readers that a physical reduction of the immediate phenomenological qualities of conscious mental states (that have the status of a brute fact revealed by experience) is inexplicable within the available conception of physical reality, he does not abandon the link between the conscious subjective phenomena and the physical and biological character of the creature that has its own subjective single viewpoint. In fact, Nagel’s argument is as follows: 1. On the basis of our first personal reflective capacities, we can immediately know that our sense experiences have a phenomenological qualitative character. 2. Conscious mental states are also functional states. (1974: 437) 3. The physical and biological sciences can reveal the functional structure of the nervous system of bats, including the features of their sonar system. 4. We do not immediately know what it feels like to perceive objects and states of affairs by means of the sonar sense unique to bats. 5. We cannot know what sonar sensations feel like by knowing the functionalphysical structure of sonar system of bats. 6. Assuming that sonar sense perceptions have unique phenomenological qualities (as is the case with visual, acoustic, tactual sensations, etc.), we cannot know what it feels like to have these sensations, neither by imagination nor by any other first person cognitive capacity that we possess. (1974: 439) 7. Realism about the subjective domain in all its forms implies a belief in the existence of facts beyond the reach of human concepts. (1974: 441) 8. There are features of reality that are beyond human knowledge. We have reasons to suppose that such features are probably instantiated by creatures that have sense organs that we do not have. (1974: 441) Nagel’s argument is a metaphysical argument. It aims to establish a significant contention regarding the facts involved in conscious experiences by invoking intuitions regarding cases the nature of which cannot be determined empirically. Attempting to determine the qualitative features of the sense-perception of bats is similar to examining imagined idealized or technically impossible cases. Whether bats have unique sense experiences cannot be conclusively determined on the basis of the available empirical evidence. At first sight, Nagel’s argument appears to involve an inherent tension. One could ask how it is possible to know that there are features of reality beyond human knowledge. The fact that we have a subjective viewpoint as well as thirdperson scientific (physical) knowledge allows us to claim that we know from our own case that the immediate phenomenological qualities are somehow correlated with the functional-physical properties. But a central premise of the argument for the existence of features that we have no way of knowing is that the correlation between, say, a certain type of brain state and visual phenomenological qualities

Chapter 4. The polemical character of the knowledge arguments 

is not equivalent to knowing that they must be correlated. Since the argument assumes that we do not know how to integrate the subjective viewpoint in physical reality, it seems that we are not allowed to claim that we know that a creature that is biologically different from us has sense experiences the qualitative nature of which we are incapable of grasping. The realistic standpoint regarding the subjective domain expressed in premise (7) does not suffice in order to claim that one knows that bats have sense experiences that are unknown to us. For it appears that in order to know that they have such experiences, we must know that the immediate qualitative character must be a feature of the experiential life of creatures endowed with an organ that functions as a sense organ. And since we do not know how to integrate the immediate phenomenological qualities of the subjective standpoint within the functional-physical account of the mind of any creature, we do not know that such features exist if we are not aware of them in the first-person subjective viewpoint. All that follows from Nagel’s argument is the mere conceivability of the existence of such unknowable features, not knowledge of their existence. This might be intriguing enough for the physicalist who shares Nagel’s background assumptions, but not as intriguing as the claim that such features exist. 4. Jackson’s knowledge argument as a polemical argument Nagel’s knowledge argument in “What is it like to be a bat?” was not meant to be an argument for the ontological distinction between the qualitative features of conscious mental sates and physical or functional states. By contrast, in EQ, Jackson’s intention was to present an argument for epiphenomenal property dualism, that is, to support the view that any conception of physical reality is bound to leave out the qualitative character of conscious phenomena that are immediately known to us by being reflectively aware of our own experiences. EQ in fact contains three knowledge arguments which I will name: ‘the simple argument,’ ‘Fred’s argument,’ and the “What Mary didn’t know” argument. In addition, it contains one argument for epiphenomenalism, in which the former arguments serve as parts. Jackson’s knowledge argument is usually presented as a deductive argument that entails the non-physical character of qualia from two highly plausible premises. Yet, as I hope to convince the reader, the argument for epiphenomenalism in EQ is in fact a polemical argument that also implies some views on the nature of the philosophical method that should be favored in this type of metaphysical investigation. The official knowledge argument, the “what Mary didn’t know” argument (the third knowledge argument in EQ) does indeed present a problem for the functional-physical reductive approaches to conscious




phenomena, but it does not suffice as an argument for epiphenomenalism. Epiphenomenalism was meant to be the solution to the problem presented by the second and the third knowledge arguments, and its recommended endorsement was based on additional implicit premises presupposed by the argument. That Jackson’s argument had this character was, as we shall see, significant for the complex polemical discourse that developed around it. Jackson does not base his argument on a complex semantic theory, or on an elaborated account of the meaning of ‘reduction,’ ‘functional role,’ or ‘functional reduction.’ ‘Physical information,’ which plays an important role in the second and the third arguments, is contextually described by Jackson as “the kind of information provided by the physical, chemical, and biological sciences.” (1982: 127) This sketchy characterization is not meant to provide a definition of ‘physical information.’ In fact, Jackson believes that his argument in EQ does not require a definition of the terms used in the premises: I do not mean these sketchy remarks to constitute a definition of ‘physical information’, and of the correlative notions of physical property, process, and so on, but to indicate what I have in mind here. It is well known that there are problems with giving a precise definition of these notions, and so of the thesis of physicalism that all (correct) information is physical information. But – unlike some – I take the question of definition to cut across the central problems I want to discuss in this paper. (1982: 127)

Why does Jackson believe that definitions are not needed? A clue could be found in his reasons for why the second and the third knowledge arguments are required. According to Jackson, the simple argument which qualia freaks base their position on is a perfectly good argument: There are many qualia freaks, and some of them say that their rejection of physicalism is an unargued intuition. I think that they are being unfair to themselves. They have the following argument. Nothing you could tell of a physical sort captures the smell of a rose, for instance. Therefore, physicalism is false. By our light this is a perfectly good argument. It is obviously not to the point to question its validity, and the premise is intuitively true both to them and to me. (1982: 127–128)

The task of the knowledge argument is not to convince qualia freaks that entail (a) non-physical qualia exist from (b) “nothing you could tell of a physical sort captures the smell of a rose...” Although, in Jackson’s view, the argument that entails (a) from (b), is “a perfectly good argument,” it has the following deficiency: I must, however, admit that it is weak from a polemical point of view. There are, unfortunately for us, many who do not find the premise intuitively obvious. The task then is to present an argument whose premises are obvious to all, or at least to as many as possible. (1982: 128 my italics, Y.S.)

Chapter 4. The polemical character of the knowledge arguments 

In other words, the argument presented in EQ is meant to persuade those for whom (a) is not obviously (intuitively) true. Moreover, it has to persuade those who believe that (a) is obviously (intuitively) false that the conclusion of the argument must be accepted on the basis of what they regard as being obviously true. Presumably, definitions are not required by virtue of the fact that the unanalyzed premises of the argument that meet the requirements of the polemical context are accepted as true by all relevant disputants. Jackson’s second argument presents a thought experiment in which a person named Fred is assumed to be able to perceptually discriminate between two colors Red1 and Red2, which we are unable to distinguish. As Jackson notes (1982: 129), the intuitions that we apparently share is that Fred can really see at least one color that we are unable to see. After conducting the appropriate physical and biological investigations, we may discover the physical mechanism that underlies Fred’s perceptual ability, but we will still not be able to see the color that he sees. If our neural system is transformed by some kind of operation, we might be able to discriminate between R1 and R 2 as he does, that is, to see the colors that he sees. But once we are equipped with the ability to discern these colors, we have acquired knowledge of a new property. In the third argument in EQ, the reader is asked to imagine a brilliant scientist, Mary, who has been locked in a black-and-white room all her life. Any visual information that she receives from the outside world is provided by means of a black-and-white monitor. While she is locked in her black–and-white room, Mary learns all the physical information about the physical and biological nature of other people, including the physical information related to the perception of red objects. But Mary has never had visual experiences of red objects. After spending long and painful years in the black-and-white room, she is finally released. According to Jackson, when she sees red objects for the first time, she learns something new about how other people see red objects.4 Given that before her release Mary knew all the physical information about the physical objects and physical properties involved in seeing red objects, the property she discovered after her release, that is, the property linked to the experience of seeing red objects, is not a physical property.

4. Some philosophers have complained that Jackson’s thought experiment is poorly conceived. Mary may see color when she dreams, rubs her eyes, or has afterimages produced as a consequence of brightness or lightness perception (Thompson 1995: 264). Yet as Graham & Horgan suggest (2000: 61–62), this defect could be remedied by supposing that Mary has been completely monochromatic from birth and that her ability to see afterimages in color has atrophied (or by means of other revisions that are insubstantial to the point of the thought experiment).




Jackson’s thought experiment could be viewed as a paradigm of the use of ideal imagined cases in metaphysics. The third knowledge argument is clearly an idealization of Fred’s argument. It is supposed to invoke intuitions that Jackson believes entail that experiences instantiate non-physical properties, though, as we shall see in what follows, these intuitions are apparently open to conflicting interpretations. Jackson’s intended interpretation is expressed in the following argument: 1. Mary (before her release) has all the physical information there is to know about what goes on in the brains of subjects that see red tomatoes. 2. When she is released from her room, Mary learns something new about the world and our visual experiences, that is, what it feels like to see red things. 3. Therefore, there are non-physical properties instantiated by experiencing red things. Her knowledge of all the physical information about people’s perceptual experiences is incomplete knowledge of these experiences. (1982: 130)5 It should be noted that the second and the third arguments are not sufficient for establishing epiphenomenalism. If the premises of the third argument entail its conclusion, what Mary learns upon her release about other people’s visual experiences is not physical information. But even if one accepts the conclusion, it does not follow that the properties instantiated by other people’s visual experiences are epiphenomenal properties. Epiphenomenalism is entailed if one accepts at least two additional premises, both of which are not explicitly defended in EQ: the closure thesis, according to which physical changes allow only physical explanations, and the claim that phenomenal properties are systematically linked to physicalfunctional states and are “somehow caused” by these physical-functional states. Let it be noted that these contentions, in particular the second implicit premise, are controversial in their own right. These additional implicit premises are located at the margins of the knowledge argument, but are significant for elucidating its rhetorical character. They do not indicate missing steps in Jackson’s argument. Let us remember that the argument was intended to achieve a polemical goal. Rather, these premises indicate that the argument is intended to persuade a relevant audience that accepts these additional premises as being “obviously” true. The task of Jackson’s argument is, therefore, to persuade his relevant audience that given the unavoidability of the conclusion of the third argument, epiphenomenalism seems to be the only viable possibility open to philosophers committed to some form of a functional-physical account of the mind. It would, therefore, 5. Byrne (2002) argued that the knowledge argument is a nonstarter, since it “illegitimately draws a metaphysical conclusion – that physicalism is false – from an epistemic premise – that physically omniscient Mary would not know everything.” Yet, as Stalnaker notes (2008: 25– 26), “there is no mystery about how epistemological premises can have metaphysical consequences, since knowledge implies truth.”

Chapter 4. The polemical character of the knowledge arguments 

be mistaken to suppose that the knowledge argument is meant to prove in the formal sense that non-physical qualia exist.6 Consequently, my suggestion is that the knowledge argument is a polemical argument presented in the context of a conflict between metaphysical intuitions which are meant to persuade the disputant, namely, the physicalist, that the epiphenomenalist’s disputed contention (that the qualitative properties of experience are epiphenomenal non-physical properties) must be accepted if its implicit and explicit premises are accepted. A precondition of the argument is that the premises of the argument are presumptions that are acceptable by all parties involved.7 At least one of these premises has the status of a perceptual or perceptual-like judgment.8 The other premise – that Mary knows all the physical information about the perception of red objects – is an idealization of ordinary cases. It has the status of a regulative idea that guides scientific investigations assumed to be shared by the disputants. The argument as a whole challenges the main presumption of the physicalist, and shifts the burden of proof to her side by rendering the dualist’s presumption directly relevant to the physicalist position on the basis of the presumptions shared by both parties. In light of the above clarifications, it can be seen that Jackson’s intended polemical argument against the physicalist has the character of the type B argument discussed in chapter 2. The premises of the argument are propositions accepted both by S1 (Jackson) and S2 (the reductive physicalist), and the conclusion is a proposition that belongs to the core set of P1 (Jackson’s epiphenomenalism). The argument established Jackson’s position as a controversial relevant alternative in the field. As I will show in the next chapter, the variety of responses to the argument indicates its success in establishing what some philosophers call ‘the knowledge intuition’ and, consequently, Jackson’s epiphenomenalism as a controversial relevant alternative.

6. As Jackson (1997: 569) acknowledges in his response to Churchland, a proof is not tenable in this context. 7. As Jackson notes (1982: 131), this is in fact what the modal argument lacks. Its premises are disputable. 8. The new information that Mary acquires is assumed to be connected to her perceptual capacities. As we shall see in the next chapters, this feature is important to many debates that address the metaphysics of subjectivity.

chapter 5

The antinomies of consciousness and their resolutions Phenomenal concepts, representationalism, and two-dimensional semantics

1. Churchland’s response to the knowledge argument In the present chapter, I will examine the controversies motivated by the knowledge arguments and their epistemic significance. The responses to the knowledge arguments and related issues that concern conscious experiences can be divided into at least three rounds of debates. The first round gradually motivated the view implicitly expressed by several philosophers regarding the extent to which the available conceptual means could deal with the problems of consciousness. As I hope to show, this lack of conceptual means is revealed by exchanges that have the features of the modified Kantian antinomy. The conflicting resolutions suggested for these antinomies generated a second round of debates on the nature of phenomenal consciousness. The third round probes deeper into the grounds of the conflicting intuitions expressed by the philosophers that participated in these debates. They address the basic semantic and modal concepts, and the variety of questions related to the nature of scientific explanations and, in particular, psychological explanations. It is possible to distinguish between at least two types of responses to Jackson’s knowledge argument. One type of response views the argument as valid, but maintains that at least one of their premises is false. A second type of response maintains that the premises express true propositions, but denies that the conclusions follow from these propositions. Each of these two types of responses is linked to a different type of polemical argument. In “Reduction, qualia and the direct introspection of brain states” (1985) (RQ), Paul Churchland criticized Nagel’s and Jackson’s respective arguments by using both types of responses.1 Churchland’s paper includes a significant 1. Though Churchland begins his paper with a short account of his views regarding reduction and elimination, his arguments do not involve his eliminativist approach. See also Churchland (1981). A claim central to his position is that “direct deducibility is an intolerable




contribution to the first round of debates on the ‘problem of consciousness’ pointed out by Jackson’s knowledge argument and by the earlier knowledge arguments. As we shall see, the difficulties revealed to the relevant addressee by this round of controversies motivated the later developments of this debate. According to Churchland, Nagel’s argument either commits an intensional fallacy or one of its premises is false. In defense of the first disjunct, Churchland construes Nagel’s argument as follows: 1. The qualia of my sensations are directly known by me, by introspection, as elements of my conscious self. 2. The properties of my brain states are not directly known by me, by introspection, as elements of my conscious self. 3. Therefore, the qualia of my sensation ≠ the properties of my brain states. (1985: 19)

Being directly known is not a feature of what is directly known, and the fact that in intensional contexts “there is a difference in truth value for two terms ‘a’ and ‘b’ [the “qualia of my sensations” and “property of my brain states” (Y.S.)] is hardly grounds for concluding that ‘a’ and ‘b’ cannot be coreferential or coextensive terms!.” (1985: 20) Nagel’s argument may be construed as having a different form: 1. My mental states are knowable by me by introspection. 2. My brain states are not knowable by me by introspection. 3. Therefore, my mental states ≠ my brain states. (1985: 21)

Here ‘knowable by introspection’ indicates a relational property of what is known.2 This construal eliminates the intensional fallacy, but “the reductionist is in a position to insist that the argument contains a false premise: premise 2.” (1985: 21) Nagel is therefore caught in a dilemma: either the argument is not valid, or one of its premises is false. Jackson’s argument is not open to the same charges. The intensionality of knowledge is irrelevant to his argument, and so is the fact that the qualitative properties are known by introspection. However, according to Churchland, Jackson’s argument involves two other shortcomings. Here is how he construes Jackson’s argument in RQ: 1. Mary knows everything there is to know about brain states and their properties. demand on reducibility.” (1985: 25) As we will see in the following chapter, the question regarding the nature of reduction and reductive explanations will become central to the debate on consciousness. 2. It is interesting to note that the term ‘qualia’ appears explicitly in Churchland’s first construal of Nagel’s argument and is omitted from the second construal.

Chapter 5. The antinomies of consciousness and their resolutions 

2. It is not the case that Mary knows everything there is to know about sensations and their properties. Therefore, by Leibniz’s law, 3. Sensations and their properties ≠ brain states and their properties. (1985: 23)

According to Churchland, the premises express true propositions only if they involve an equivocation. The meaning of ‘know’ in the first premise differs from the meaning of ‘know’ in the second premise. In the first premise, ‘knows’ denotes knowledge of “a set of sentences or propositions, the kind one finds in written neuroscience texts, whereas ‘know’ in (2) seems to be a matter of having a representation in some prelinguistic or sublinguistic representation of sensory variables.”(1985: 23) According to Churchland, this kind of knowledge is ‘knowledge by acquaintance,’ which differs from discursive knowledge.3 In a later article, in which Churchland addresses Jackson’s response to his first contribution to the debate, Churchland (1997: 572) describes the differences between the two types of knowledge as being similar to those between a competent golfer’s motor representation of a golf swing, which constitutes her knowing how to execute a proper swing, and her discursive representation of the same golf swing. The differences between the two types of knowledge do not entail Jackson’s intended conclusion that Mary’s new knowledge constitutes knowledge of a non-physical property. Up to this point, it appears that Churchland’s argument against Jackson begs the question. Does Mary’s ignorance before her release merely consists in the fact that she does not have knowledge by acquaintance of experiences, or does she discover new properties as a result of the experiences she has upon leaving the room? A strong intuition supports the supposition that the latter possibility is the correct one. Upon her release, it seems that Mary had gained knowledge of new facts, and did not merely entertain a new type of knowledge – knowledge by acquaintance – of states of affairs about which she had already known all related facts before she was released. It should be noted, however, that Churchland’s argument does not end at this point. According to Churchland (1985: 24), Jackson’s argument, if it is sound, proves too much. Suppose that Jackson were arguing not against materialism, but rather against dualism. Let Mary now be a brilliant scientist who has learned from books and lectures everything there is to know about ectoplasm, which is the stuff from which non-physical entities are made. According to Churchland (1985: 25), there is still something she doesn’t know: What it is like to see red objects? The counterpart knowledge argument seems to be similar to the original one. It appears that in both cases she fails to know what it is like to 3. The terms ‘knowledge by acquaintance’ originate in Russell’s (1910) “Knowledge by acquaintance and knowledge by description.”

 Controversies and the Metaphysics of Mind

see red objects. Having all the discursive information regarding ectoplasmatic objects, events, and processes leaves her in the same deprived epistemic situation that characterizes her counterpart Mary. It seems to follow that the fact that she fails to know what it is like to see red tomatoes cannot be explained by pointing out that her experiences instantiate non-physical properties. A more reasonable explanation would be to suppose that in both cases the differences between her knowing and not knowing what it is like to see red objects consist in the type of knowledge that she has in each case, and not in the type of states or events that she knows something about. Churchland’s argument is therefore as follows: 1. Mary1 knows (discursively) all the physical facts one could know about the experiences a person has when perceiving red objects, the physical information one could learn from lectures and books, but she does not know what it feels like to experience red objects. 2. Mary2 knows (discursively) all the ectoplasmatic facts one could know about the experiences a person has when perceiving red objects, the information about ectoplasm one could learn from lectures and books, but she does not know what it feels like to experience red objects. 3. The best way to explain what Mary1 and Mary2 both lack is to suppose that in both cases, they do not have immediate and first-person knowledge by direct acquaintance with the relevant entities. 4. If the knowledge that Mary1 lacks is knowledge by acquaintance of the experiences of red objects, then it must be admitted that the premises of Jackson’s argument equivocate the meaning of ‘know about,’ and therefore, the argument does not entail the intended conclusion. Churchland’s argument against Jackson also has the features of type (B) of polemical arguments mentioned in Chapter 2. The premises of the argument include statements that Churchland assumes to be acceptable by Jackson. The conclusion of the argument is a statement compatible with Churchland’s position and incompatible with Jackson’s position and, therefore, rebuts Jackson’s contention that the premises of his argument entail dualism. 2. Jackson’s response to Churchland Churchland’s response to Jackson’s knowledge argument was similar to Lewis’s (1983b, 1988) and Nemirow’s (1980) responses that also construed the argument as involving an equivocation on ‘knowledge.’ Yet, according to Lewis, Mary’s new knowledge should be classified not as knowledge by acquaintance but as ‘knowing

Chapter 5. The antinomies of consciousness and their resolutions 

how’ (knowing how to imagine red objects, how to identify red objects, etc.). There is a difference between the two suggestions, but it should be noted that Churchland’s examples seem to blur the difference between Lewis’s ability hypothesis and his own knowledge by acquaintance hypothesis. Although he construes the difference between discursive knowledge and knowledge by acquaintance as a difference in the type of representations linked to the two kinds of abilities – discursive abilities and motor abilities – the difference appears to be a difference in abilities. One may grant the claim that motor abilities that constitute the golfer’s knowledge on how to execute a good swing involve a unique kind of prelinguistic representations. But this type of functional-representational account of the motor abilities of competent golfers seems to be irrelevant to the new knowledge gained by Mary after her release. If she has all the physical information about seeing red objects, the information must include the functional-representational account of motor abilities. The claim that the new knowledge Mary gained after her release cannot be construed in terms of differences between types of knowledge was central to Jackson’s response to Churchland. According to Jackson (1996: 568), Churchland’s construal of the knowledge argument misquoted the premises of his original argument. In Churchland’s construal, the argument was as follows: (1) Mary knows everything there is to know about brain states and their properties. (2) It is not the case that Mary knows everything there is to know about sensations and their properties. Therefore by Leibniz’s law: (3) Sensations and their properties are not identical to brain states and their properties.

By contrast, Jackson’s argument is the following: (1’) Mary (before her release) knows everything physical there is to know about other people. (2’) Mary (before her release) does not know everything there is to know about other people (because she learns something about them upon her release) Therefore: (3’) There are truths about other people (and herself) which escape the physicalist story. (1997: 568)

There are clearly substantive differences between the two arguments. The conclusion of Jackson’s “what Mary didn’t know” argument was not the claim that brain states and their properties are not identical to sensations and their properties. This was, to be sure, the conclusion of the overall argument in EQ but not of the “What Mary didn’t know” argument. Jackson’s central point in his response was

 Controversies and the Metaphysics of Mind

that after her release, Mary learned new facts, or new truths about other people and herself, something that is not possible according to Churchland’s construal. Given that Mary’s case is more aptly described as a case in which, upon her release, she gains new information about the experiences a person has when perceiving red objects, it is not clear how the new knowledge that she gained could be exhibited in her abilities, though as we shall see below, this was not necessarily one of the advantages of Jackson’s position. Should one prefer Jackson’s construal over Churchland’s reformulation? My suggestion is that from an outside observer’s perspective, there are no conclusive reasons to favor either one of the interpretations. More precisely, there are reasons to adhere to both incompatible positions. First, given the difficulty of depicting the knowledge gained by Mary upon her release as knowledge by acquaintance or knowledge how, it seems reasonable to deny Churchland’s construal. But, as I noted above, Churchland’s overall argument against Jackson’s argument did not merely consist of the claim that the argument equivocates the term ‘know about.’ According to Churchland, no less than the physicalist, the dualist is required to apply the distinction between the two types of knowledge in order to explain what is missed by a person who only has discursive knowledge. In his response to Churchland, Jackson disconnects the charge of equivocation from the counterpart knowledge argument directed against dualism. His reply to this charge (1997: 569) simply denies the possibility involved in the counterpart thought experiment: “Lectures about qualia over black and white television do not tell Mary all there is to know about qualia.” Yet, as Churchland observed (1997: 573–574), this reply begs the question. The most reasonable explanation that clarifies the differences between Jackson’s knowledge argument and Churchland’s counterpart knowledge argument undercuts Jackson’s claim that the differences between the types of knowledge are irrelevant to his argument. Why is it not possible to conceive of a brilliant scientist that has the same type of complete discursive knowledge about non-physical qualia that Mary has about the physical processes involved in perception and cognition? What could explain the contention that she does not know everything there is to know about qualia?4 If qualia qua epiphenomenal properties are unique types of objective properties (though not physical properties), there seems to be no reason why Mary in the counterpart argument cannot learn everything there is to know about them from books and lectures. If one denies that they are objective properties in the required sense, the most reasonable explanation of their being ‘not objective’ is that, in contrast to Mary’s knowledge of all the physical facts, the second type of discursive knowledge inherently involves first-person reflective capacities. But this opens 4. Compare Perry 2001: 163–168.

Chapter 5. The antinomies of consciousness and their resolutions 

Jackson’s argument to the charge that Mary’s ignorance about qualia may after all be due to the differences between types of knowledge. The reflections on the Jackson and Churchland exchange do not necessarily entail that Jackson’s contention that the type of new knowledge which Mary gained after her release cannot be said to be knowledge of new facts, but rather an intriguing question as to how immediate acquaintance and first-person reflective capacities are relevant to the knowledge of these new facts. For obvious reasons, Jackson’s argument minimizes the importance of the first-person reflective viewpoint that dominates Nagel’s argument. But it does not explain the role of the first-person viewpoint in the account of conscious phenomena. If conscious phenomena inextricably involve both qualitative features and a first-person viewpoint, the exchange leaves the outside observe without an answer as to how the two features of conscious phenomena are linked to each other. The exchange between Jackson and Churchland can therefore be described as a case in which each of the contenders is able to present a persuasive type (B) argument against the opponent’s position.5 The exchange can therefore be interpreted as exemplifying an antinomy in Kant’s sense. This was not recognized by the opponents themselves. Yet, as we shall see in the next sections, other philosophers recognized the antinomic nature of the exchange. 3. The causal efficacy of conscious qualitative states The antinomy related to the character of Mary’s new knowledge was not the only antinomy revealed by the debates on Jackson’s knowledge argument. Mental causation and the causal efficacy of the mental were the topic of another antinomy. Supporters of epiphenomenalism willingly admit that epiphenomenalism is an unappealing metaphysical position. By requiring the suspension of the intuition as to the causal efficacy of the qualitative features of conscious states, it blatantly goes against common sense. Endorsing epiphenomenalism on the basis of a persuasive argument seems to parallel conceptual shifts in science that do not agree with our pretheoretical commonsensical conception, such as those required by the spatial and general theory of relativity, quantum mechanics, etc. Yet, the question that was raised quite early in the polemical exchange on the nature of consciousness was not whether epiphenomenalism goes against our commonsensical conception of mental reality, but, rather, whether it is a coherent position.6 Clearly, the accusation that it is incoherent was not entirely new. The nature 5. Jackson’s polemical argument is the argument discussed in Chapter 4. 6. See in particular the exchange between Shoemaker (1975, 1981) and Block (1980).

 Controversies and the Metaphysics of Mind

of mental causation was central to the metaphysical debates on the nature of mind in the 17th and 18th century. Nevertheless, by using ideas included in the novel context in which it was presented, the reformulation of the difficulties linked to mental causation had a significant effect on the particular nature of the historical evolution of the relevant metaphysical concepts and the metaphysical presumptions to knowledge. As we shall see below, the accusation that epiphenomenalism is an incoherent position was made by Dennett (1991) in his response to the knowledge argument. Dennett’s response seems to be similar to Churchland’s. Both philosophers apparently hold similar philosophical views. Yet, Dennett’s response differs from Churchland’s. We can use the ideas developed in Part One of this book to point out this difference. In contrast to Churchland, who addresses Jackson’s position as a relevant alternative that requires a polemical response, that is, as a position that uncovers an important distinction, though in a twisted way, Dennett wishes to undermine Jackson’s position and similar positions qua relevant alternatives. Dennett raises two objections against Jackson’s argument. He first challenges the very idea of a metaphysical thought experiment and the supposition that it is possible to determine what is really possible merely on the basis of what is imaginable or conceivable, by telling stories such as that which the knowledge argument is based on. In Dennett’s view (1991: 400), whether Jackson’s intended conclusion (that Mary learned new facts) is true or false depends on the way Mary’s story is told. Dennett claims that the different ways of telling the story are bound to have conflicting results. It might be supposed that Dennett intends to support a physicalist interpretation of Jackson’s imagined case similar to Churchland’s interpretation. But this way of viewing Dennett’s response misses his main intent. He does not maintain that the other way of telling the story is mistaken. His argument is not meant to establish that Mary didn’t learn anything new, but rather “that the usual way of imagining the story doesn’t prove that she does.” (1991: 400) Thought experiments such as Jackson’s imagined case convey no decidable result. In Dennett’s terms (2005: 103), they are ‘intuition pumps’ that have limited rational import, and they are suspicious as grounds for determinable conclusions. That Dennett’s goal is to undermine the plausibility of his opponent’s position as a relevant alternative is clearly shown by the rhetorical means that he uses. The forms of speech that he uses seem to ridicule the opponent’s contentions. Though his arguments apparently respond to Jackson’s argument, his intended addressee is not a metaphysician or a philosopher, but rather cognitive scientists and philosophers from the anti-metaphysical camp, who view metaphysics as an ill-founded pseudoscience. Here is an alternative way of telling Mary’s story: Say that after her release, as a trick, Mary’s capturers present her with a blue banana. This will be her first

Chapter 5. The antinomies of consciousness and their resolutions 

color experience ever. Given that Mary knows all the physical facts, she will realize that her capturers had tried to deceive her. For before her release, she already knew merely on the basis of her knowledge of physical facts that the physical color of bananas normally causes a particular type of neural processes in normal observers; by analyzing the physical signals emitted from the observed banana and by analyzing the physical processes that occur in her brain she will, therefore, know that the color of the banana she observes is blue and not yellow. This seems to be perfectly compatible with Jackson’s argument. To remind the reader, the argument does not challenge the feasibility of a functional account of thoughts and beliefs. This was crucial for its polemical purposes. But if Mary’s thoughts and beliefs must be realized physically, and if she knows all the physical facts, she will know what thoughts or beliefs a person has when she experiences yellow objects or blue objects. If the knowledge that b is a belief about yellow objects involves knowing what ‘yellow’ means, and if knowing what ‘yellow’ means is impossible without knowing what it is like to see yellow objects, it appears that if Mary knew what thoughts and beliefs a person has before her release, she knew every fact about the experiences of others.7 It seems that Dennett’s objection begs the question. Even if on the basis of her neural states and linguistic behavior, Mary had known what thoughts and beliefs a person has when she perceives red objects, she may fail to know some of the phenomenal facts involved in having such experiences. The physical and behavioral information that she has may suffice to indicate that the person she examines has beliefs about what ‘red’ refers to in that person’s language without fully knowing what experiences of red things involve. Mary could know that she was deceived because she knows what thoughts and beliefs she has. However, this does not entail that she knows all the facts about experiencing colored things. She may both know that she was deceived and nevertheless find out for the first time what it is like to experience blue things. But Dennett’s counter-thought experiment is linked to other, highly implausible, features of Jackson’s epiphenomenalism. These undesired features are stated in his second objection to Jackson. As he notes, if one assumes, as Jackson does, that the new facts Mary learns about color experiences upon her release are epiphenomenal properties of these experiences, 7. According to Robinson (1993), Dennett’s argument contains a confusion that rests on an ambiguity. Mary’s omniscience of physical reality consists in her theoretical knowledge (knowledge of all the relevant physical science). It is not knowledge of all the relevant information regarding physical particulars. Yet it is interesting to note that the relevance of this distinction to the knowledge argument was indirectly denied by Jackson (1998) and Chalmers & Jackson (2001). According to Jackson and Chalmers & Jackson, if physicalism is true about our world, the laws of the physical sciences and indexical information regarding our world entail all the facts about our world. I will address these contentions in Chapter 6.

 Controversies and the Metaphysics of Mind

then the new beliefs she acquires upon her release must also be epiphenomenal. It may seem that this objection does not raise any new problem for Jackson. Yet, if the new beliefs Mary acquired cannot be realized in her brain or be causally relevant to her behavior, it follows that if a person feels pain, the qualitative features of the pain she is introspectively aware of are irrelevant to her physically realized belief system.8 The phenomenal properties of states of awareness cannot cause physically realized phenomenal beliefs. Moreover, the awareness itself cannot be realized physically. It follows that having experiences in the required phenomenal sense is completely irrelevant to the fact that she is physically disposed to react in a certain way. There is no contradiction in this possibility. Yet, this is a highly implausible result of Jackson’s position.9 Should one conclude that Jackson’s knowledge argument contained a mistake? I believe that different conclusions are in order. Dennett’s objections are appealing primarily with regard to Jackson’s epiphenomenalism, that is, with regard to the conclusion of the overall argument of EQ. It is far less appealing with respect to the claim that Mary learned something new upon her release. Philosophers exposed to this conflict seem to be torn between two intuitions related to conscious experiences that seem to pull them in two different directions: the ‘new knowledge’ intuition, and the ‘the causal relevance’ intuition.10 We have no empirical means other than the overt reports of the examined person or her neural states in order to determine whether she has acquired new phenomenal knowledge. But all physical information that is accessible to the third-person viewpoint was known by Mary before her release. The fact that it is impossible to resolve the conflict between intuitions connected to conscious mental states by attending to the third-person realm of information may lead one to suppose that the philosophical question regarding the nature of conscious mental states is a pseudoquestion. Similar reasons seem to motivate Dennett’s dismissive approach to Jackson’s argument and other similar arguments. However, from the realistic standpoint that is shared by most participants in the debate on consciousness, regardless of whether or not the topics involved can be decided empirically, the 8. As Robinson notes (2004: 71), the first version of this type of argument appears in Shoemaker (1975, 1981), that is, before the publication of Jackson’s original paper. As he notes (2004: 72), Dennett’s argument at least demonstrates that one cannot combine a reductivephysicalist account regarding thoughts and beliefs and non-reductive accounts of experience. 9. It should be noted that Jackson himself accepted this criticism when he abandoned his earlier position. See Jackson 2004. 10. As Stoljar and Nagasawa showed (2004), Jackson’s argument was preceded by several arguments that employed the knowledge intuition. Nevertheless, before the knowledge argument was stated by Jackson, the above two widely-shared intuitions were believed to be compatible intuitions.

Chapter 5. The antinomies of consciousness and their resolutions 

problem is not a pseudo-problem. Whether it is a real problem depends on whether the knowledge intuition is as real as the causal efficacy intuition. The conflict between the knowledge intuition and the causal efficacy intuition, therefore, generates another antinomy which is even more intriguing than the first one. Both intuitions seem to be entrenched in our pretheoretical conception of conscious mental states. Yet one intuition seems to entail epiphenomenalism while the other leads to the opposite conclusion. 4. The paradox of phenomenal consciousness In contrast to Dennett, many philosophers of mind conceived of the puzzles related to the qualitative features of conscious mental states as anything but bogus puzzles. Jackson’s knowledge argument persuaded many philosophers that the qualitative features of conscious experiences convey unique type of information. However, if one accepts (a) the closure thesis, (b) that qualia are re-identifiable sensory types, and (c) that physical states of affairs are causally relevant to having qualitative experiences, how can one avoid Jackson’s epiphenomenalism? Epiphenomenalism is not merely an unappealing metaphysical option. As Dennett’s objection and similar objections clarified, epiphenomenalism about qualia entails that beliefs and thoughts about qualia are also epiphenomenal, and therefore, the qualitative features of experiences must be causally irrelevant to beliefs and desires involved in action and behavior. This conflict between equally plausible intuitions has the features of a Kantian antinomy. Concerning the antinomy generated by the knowledge intuition and the causal efficacy intuition, one side consists of the conditional contention that if the qualitative features of conscious experiences involve a unique type of information, they cannot be physical properties of objects, whereas the second side is that beliefs about the qualitative features of conscious experiences, say, that one is in pain, must be causally relevant to pain behavior and action. It appears that one cannot hold both sides of the antinomy. Yet, it is generally agreed not merely that conscious experiences have both features but, in addition, that they must have both. The first round of the polemical exchange on the knowledge arguments revealed the mystery related to the link between these features: how could they together be features of conscious experiences? Needless to say, not all philosophers that addressed the knowledge arguments reached the above conclusion. Dennett and his allies, for example, persistently denied that Jackson’s, Kripke’s, and Nagel’s knowledge arguments are significant to grasping the nature of conscious experiences. Churchland defends a milder position, namely, that Jackson’s knowledge argument uncovers a conceptual

 Controversies and the Metaphysics of Mind

gap. But in his view, the gap could easily be filled by the available conceptual resources without adding conceptual innovations that will revise the physicalist position. Many other philosophers, physicalists and dualists alike, responded differently to the puzzles revealed by the knowledge arguments. Here is how Tye expresses the philosophical problem that he labels ‘the paradox of phenomenal consciousness:’ The cumulative effect of the considerations adduced so far is to put intense pressure on the view that phenomenally conscious states are physical, in any sense of the term ‘physical’, no matter how broad. Suppose now we throw in the towel on behalf of the physicalist and adopt an antiphysicalist stance: experiences and feelings have irreducible, nonphysical, perspectivally subjective features. This position, which is sometime called type dualism or property dualism, relieves the pressure from the problems of perspectival subjectivity, mechanism, duplicates, and the inverted spectrum. But it helps not at all with the problem of phenomenal causation...We seem compelled to admit that they way things feel and look and taste and sound and smell are causally irrelevant. As Frank Jackson (1982, p. 134) notes “[T]hey are an excrescence. They do nothing... But phenomenal qualities are not an excrescence. We know that the phenomenal qualities of our mental states are causally efficacious, just as we know that we have hands and legs, that there are rocks and trees. (1995: 56–57)11

In the above passage, Tye refers to the conflict between the knowledge intuition and the causal efficacy intuition. Clearly, whether the antinomy is real depends on how one articulates the intuitions from which it results. Some philosophers, most notably Colin McGinn (1991, 1999), maintain that the inherent puzzle involved in our notion of conscious phenomena is insolvable. Impressed by the intuitive appeal of the knowledge arguments, Tye and other philosophers believe that the paradox can be resolved. Given that a resolution is feasible, it has the status of a paradox or an antinomy only within the epistemic horizon of real possibilities that precedes the argument that resolves it. That it is genuinely conceived to be a paradox from the perspective of a given epistemic horizon of real possibilities indicates that its resolution required a significant transformation of the relevant conceptual state. It required adding novel conceptual resources and distinctions, and revealing new conceptual connections. Interestingly, although philosophers that share the view that the knowledge argument involves genuine philosophical puzzles normally agree on how they characterize the conceptual abyss it reveals, their resolutions conflict in many 11. Levine explicitly notes that the problem of consciousness has the features of a Kantian antinomy. According to Levine (2001: 9), “[the] mind body problem, at least with respect to the issue of conscious experience, presents us, in a way, with a Kantian antinomy.” I will address Levine’s position in the next chapter.

Chapter 5. The antinomies of consciousness and their resolutions 

cases. As a result of the differences between these responses, new philosophical controversies emerged. As part of their background, these controversies are between positions that share the knowledge intuition that Jackson’s argument rendered relevant to the topics discussed by philosophers of mind. I suggest that the knowledge intuition is assumed in advance by many conflicting physicalist resolutions and the controversies they motivated indicate the epistemic achievement related to Jackson’s argument. Most philosophers of mind who are moved by the knowledge intuition are unmoved by Jackson’s intended conclusion. And yet, the wide agreement regarding the importance of the knowledge intuition and the gap it reveals in the current conception of physical reality indicates the epistemic change that it generated. The asymmetry between agreement regarding the puzzles that need to be resolved, on the one hand, and the conflicting resolutions suggested on the other hand, indicates the rational character of these types of controversies. The epistemic achievement is connected to a problem that is open to conflicting attempts to resolve it, which are allowed by the horizon of real possibilities from which it is conceived. Clearly, these were not Jackson’s intentions when he introduced his argument. Nevertheless, the cunning of critical rationality rendered the position his arguments supported, which entailed the epiphenomenal character of the qualitative features of conscious states, a relevant alternative to the set of positions with which it conflicted. In the next sections of this chapter, I will examine some attempts to resolve the antinomies revealed by the controversies generated by Kripke’s, Nagel’s, and Jackson’s knowledge arguments and the new controversies motivated by these attempts. 5. The phenomenal concepts strategy The first stage of the debates on the knowledge argument resulted in the relatively broad acceptance of the importance of the knowledge intuition to a philosophical account of conscious mental states. One type of physicalist response to the problems revealed by the controversies on the knowledge argument demonstrates how one can “have it both ways” by presenting new distinctions and conceptual innovations. This is one type of resolution of an antinomy. Loar’s response (1990, 1997) and Tye’s response (1995) to the knowledge argument are two significant attempts to present this type of resolution.12 But although their resolutions 12. Tye has significantly changed his position in his recent publications. His earlier resolution of the problem of perspectival subjectivity was based on the concept ‘phenomenal concept.’ He no longer believes that there are phenomenal concepts. See in particular Tye (2009). In the present book, I will discuss his earlier position.

 Controversies and the Metaphysics of Mind

are physicalist, they are incompatible. There are other similar resolutions that differ significantly from theirs. In the present section, I will briefly examine Loar’s response. At the beginning of his (1997) article, Loar explicitly notes that the intention that motivates his arguments is polemical: to rebut the dualist contention.13 In contrast to Dennett and Churchland, Loar maintains that “The antiphysicalists’s phenomenological and internalist intuition are correct.” (1997: 597) However, [t]he idea is to engage them over the central point, that is, whether those aspects of the mental that we both count as phenomenologically compelling raise substantive difficulties for the thesis that phenomenal qualities (thus understood) are physical properties of the brain that lie within the scope of current science. (ibid.)

In Loar’s view, the central supposition that underlies the knowledge argument – the irreducibility of phenomenal concepts to physical-functional concepts – may lead those that are moved by it to endorse three incompatible positions: ...that consciousness and phenomenal qualities are unreal because irreducible; that they are irreducible non-physical-functional facts; that they are forever mysterious, or pose an intellectual problem different from other empirical problems, or require new conception of the physical. (ibid.: 597–598)

It should be recalled that Jackson’s argument in EQ contains four distinct steps, the simple argument, the Fred argument, the “What Mary didn’t know” argument, and the argument for the epiphenomenal character of qualia. The conclusion of the “What Mary didn’t know” argument conflated two claims, namely, that Mary had learned new facts and that these new facts could not be physical facts. The supposition underlying this conflation was that Mary’s knowledge of physical information before she was released is equivalent to all the physical information there is. As Horgan (1984) observed quite early in the dispute, the new facts Mary learned could have been physical facts. In Horgan’s view, the scientific theoretical knowledge attributed to Mary by the first premise of the knowledge argument is not equivalent to knowledge of all the physical facts. It is possible to distinguish between explicit physical information, which is information that is expressed by means of a sentence S “just in case S belongs to, or follows from, a theoretically adequate physical account,” (1984: 150) and ontological physical information, which is information that a sentence S expresses just in case “(i) all the entities referred to or quantified over in S are physical entities, and (ii) all the properties and relations expressed by the predicates in S are physical properties and relation.” (ibid.) This distinction undercuts the contention that Mary’s new information 13. The first version of Loar’s article appeared in 1990.

Chapter 5. The antinomies of consciousness and their resolutions 

must belong to the non-physical type. In other words, according to Horgan (1984: 151), a person who has all the scientific (explicit) physical information could, nevertheless, acquire new physical information, that is, ontological physical information. But Horgan does not explain how the two types of information are related, and the sense in which the information acquired by Mary upon her release is indeed new physical information.14 These queries are addressed by Loar (1990, 1997) in his response to the knowledge argument. Loar (1997: 607) first denies that the puzzle revealed by the knowledge argument is resolvable if one endorses Nemirow’s (1980) and Lewis’s (1983b, 1988) ability hypothesis. Like Horgan, Loar wishes to show that an epistemic change is discernable in Mary’s overall cognitive states, before and after her release. This change concerns information and not mere abilities. Nevertheless, this change could be accounted for without involving the non-physical type of information. According to Loar, the account of the change in Mary’s epistemic states requires a distinction between phenomenal concepts and phenomenal properties. Respectively, he distinguishes between the two functionalist theses “that all concepts of mental states are functional concepts, and that all mental properties are functional properties.” (1997: 613) Loar rejects the first thesis but accepts the second one. The concept ‘phenomenal concept’ is not a functional concept and, nevertheless, the properties that phenomenal concepts refer to are physicalfunctional properties. In order to substantiate the distinction between mental concepts and mental properties, Loar first reveals the common features of Jackson’s knowledge argument and Kripke’s argument for dualism by identifying the implicit semantic premise that they both involve: (Semantic Premise) A statement of property identity that links conceptually independent concepts is true only if at least one concept picks out the property it refers to by connoting a contingent property of that property. (1997: 600)

According to Loar, the implicit semantic premise of both arguments could be denied without harming the unique mode in which phenomenal terms rigidly designate phenomenal properties (Kripke) or the new knowledge gained by someone who has experienced red objects for the first time.

14. It is interesting to note that when Horgan returns to discuss the knowledge argument (see Graham & Horgan (2000, 2005)), Jackson’s thought experiment is described as “...immensely powerful, but alas, currently its force is not sufficiently appreciated.” (2000: 59) Graham & Horgan believe that Jackson’s thought experiment poses a problem to physicalism which cannot be met by resolutions that are based on the concept ‘phenomenal concept’ to be discussed below. See also Raffman (2005).

 Controversies and the Metaphysics of Mind

Loar’s concept ‘phenomenal concept’ does not imply the above semantic premise. It is basically a novel concept, although it does have its historical antecedents. It was not available to philosophers as part of the shared philosophical knowledge before it was presented in the debate on the nature of consciousness. As noted above, Loar assumes that phenomenal properties (qualia) are internal properties of the mind/brain and not features of objects in the world. According to Loar (1997: 601), phenomenal concepts are recognitional concepts that pick out physical-functional properties of the brain. Recognitional concepts are not theoretical concepts. These are concepts that have the form ‘x is of that kind:’ Suppose you go into the California desert and spot a succulent never seen before. You become adept at recognizing instances, and gain a recognitional command of their kind, without a name for it; you are disposed to identify positive and negative instances and thereby pick out a kind. (1997: 600)

As Loar notes, recognitional concepts are generally formed against a further conceptual background, and are perspectival: Suppose you see certain creatures up close and form a recognitional concept – ‘those creatures1’; and suppose you see others at a distance, not being able to tell that they are of the same kind (even when they are), and form another recognitional concept – ’those creatures2’. These concepts will be a priori independent. Now the respect in which they differ is perspectival, in some intuitive sense. A recognitional concept is in part individuated by its constitutive perspective. (1997: 601)

Albeit their perspectival features, recognitional concepts could be applied outside their constitutive perspective in a way that allows for identity statements such as ‘that creature is one of those creatures.’ Phenomenal concepts qua recognitional concepts are also applicable outside their constitutive perspective. In order to rebut the dualistic conclusions of Kripke’s modal argument and Jackson’s knowledge argument, one must explain how phenomenal concepts could differ from physical-functional concepts. It seems intuitively clear that recognitional concepts generally differ from physical-functional concepts. But recognitional concepts other than phenomenal concepts refer to the properties they stand for via the contingent modes of presentation of these properties. The fact that phenomenal concepts do not refer to the properties they designate by means of contingent properties of these properties seems to be precisely the point emphasized by Kripke’s modal argument. In other words, phenomenal concepts seem to create a puzzle precisely because they refer directly to the properties they designate, that is, they capture their physical- functional essence. Nevertheless, they are conceptually independent of the physical-functional concepts that refer to the same properties. But according to Loar, the mutual

Chapter 5. The antinomies of consciousness and their resolutions 

independence of phenomenal (recognitional) concepts and physical-functional concepts does not commit one to the metaphysical independence of the properties that each type of concepts designates. The distinctive cognitive role of each type of concept suffices to determine their independence without involving two distinct kinds of properties (1997: 602). This is, indeed, generally the case with regard to recognitional concepts. They have a different cognitive role than that attributed to physical-functional concepts. Here the anti-reductionist may raise the complaint that the physicalist views “depend on ad hoc assumption and that it is tendentious to suppose that phenomenal concepts differ from all other recognitional concepts in not having contingent modes of presentations.” (1997: 602) Yet, as Loar notes, the anti-physicalist and the dualist are in fact in the same boat. For the anti-physicalist’s account, no less than Loar’s physicalist account, involves phenomenal concepts that directly pick out nonphysical properties rather than properties that they refer to via the contingent modes of presentations of these properties. Given Loar’s account of phenomenal concepts, it is possible to maintain that Mary gained new conceptual knowledge upon her release without learning anything new about the properties involved in experiencing red things. The sentences ‘John feels pain’ or ‘I feel pain’ (uttered by John) and ‘a physical-functional processes of XYZ type occurs in John’s brain’ express the same metaphysical fact though they have different conceptual content. It is not difficult to see that Loar’s resolution of the puzzle echoes Jackson and Churchland’s controversy. Churchland emphasized the unique mode of knowledge involved in conscious experience and denied that Mary’s novel knowledge is conceptually novel, whereas Jackson, who insists that Mary learned new information upon her release attributed no role to the first-personal features of conscious experiences in the account of the phenomenal properties. According to Loar’s resolution, Mary’s novel knowledge is conceptual and, nevertheless, it involves the first personal mode of self-knowledge. Loar’s response to the knowledge argument and his resolution of the puzzle of consciousness therefore involve a significant conceptual change. If phenomenal concepts directly pick out physical-functional properties without involving contingent modes of presentations, the puzzle of phenomenal consciousness is resolved. But it should be noted that it is resolved only if the distinction between concepts and properties is drawn, and only if phenomenal properties involved in perception are non-representational properties of brain states and not of the perceived objects. One can deny that phenomenal concepts are functional concepts, and therefore also explain the first-person/third-person asymmetry only if this distinction is introduced. As we shall see below, these claims are anything but trivial.

 Controversies and the Metaphysics of Mind

Loar’s response to the knowledge argument is not merely an attempt to resolve the puzzle of phenomenal consciousness. It could equally be viewed as an argument for the link between concepts and properties required by physicalism. That the intuitions regarding the nature of phenomenal states were mixed intuitions that equally involved (subjective) “matters of fact” and modal and semantic issues was clear to anyone considering Kripke’s argument, but it was marginal in the explicit form of Jackson’s knowledge argument. By connecting the two arguments, Loar emphasized the relevance of the modal and semantic layers of intuitions regarding conscious experiences to the metaphysical nature of conscious mental states. In order to solve the problem revealed by the controversies about the knowledge argument, the scope of issues addressed needed to be broadened. One had to add semantic, modal, and logical issues to the debate. As we shall see below, this is what characterizes the third round of the controversy on the nature of conscious mental states. 6. Tye’s PANIC account of phenomenal consciousness Before examining the later development of the controversy, let us examine how the knowledge intuition and the puzzles it involves led to conflicting responses. As noted, Loar’s response to the knowledge arguments was not the only response that attempted to integrate the knowledge intuition within a physicalist’s conception.15 Tye’s response to the problems related to the knowledge intuition was no less motivated by the intention to resolve the paradox of phenomenal consciousness. In the present section, I will briefly present his resolution. Tye’s earlier response is probably one of the most sophisticated so far. It consists not only of the introduction of novel concepts and distinctions, but rather integrates ideas gradually developed in the 1980s and early 1990s which opened up new possibilities in approaching consciousness. Tye’s resolution (1995, 2000) shares several things in common with Loar’s response. In particular, his resolution uses phenomenal concepts. Nevertheless, there are substantial differences between the two resolutions that result from the way in which the problem of phenomenal consciousness is depicted. According to Tye, there are in fact ten problems of consciousness, the most difficult of which is perspectival subjectivity.16 Yet in Tye’s view, the most 15. There are, to be sure, several other similar responses. See in particular Perry (2001) and Papineau (2002). I will address the differences between Tye’s and Loar’s responses in order to exemplify the different modes of resolving metaphysical puzzles. 16. See Tye 1995: Chapter 1.

Chapter 5. The antinomies of consciousness and their resolutions 

promising way to resolve this problem is to address it only after providing an account of the other problems; in particular, the problem of phenomenal causation, the problem related to the diaphanousness of phenomenal consciousness, and the problem of ownership. Tye’s main idea is that the content of phenomenal states should be expressed in terms of his PANIC theory: Poised, Abstract, Non-conceptual, Intentional (representational) Content. “Poised” indicates their functional-causal role in mental processes. “Abstract” means “that no particular concrete object enters into these contents.” (1995: 138) Phenomenal states are representational, intentional states. Nevertheless, the content of phenomenal states is not conceptual. Tye’s reason for denying that phenomenal states have conceptual content is the non-doxatic character of their representational content.17 They are symbolic states and non-conceptual states. The non-conceptuality of phenomenal content allows for a response to the problem of phenomenal consciousness revealed by the knowledge argument. The main idea involved in Tye’s PANIC theory is that phenomenal properties are not instantiated properties of mental states, but representational properties that stand for external properties of objects. This was not a novel idea. The transparency of phenomenal states (the fact that the phenomenal character of phenomenal states varies with the way in which external things or physical states of one’s own body are represented) was introduced into the current debate by Harman (1990). According to Harman, phenomenal properties are intentional (representational) properties, not instantiated properties of phenomenal states. By integrating this supposition with other ideas, Tye’s account differs significantly from Harman’s account. The main difference consists in Tye’s claim to the non-conceptuality of sensory (phenomenal) representations of objects. Similarly, though Tye (1995: 165–171) partly endorses Loar’s account of phenomenal concepts, he integrates it into the theory that characterizes their representational content as nonconceptual and as endowed with perspectival subjectivity. Phenomenal properties are not properties of internal states of the brain but are representational properties of the (represented) external objects. This clearly conflicts with Loar’s position.18 Phenomenal concepts are either indexical concepts or predicative concepts.19 In contrast to Loar, they do not refer qua recognitional concepts to physical-functional states of the brain under a unique mode of presentation. Also, they do not 17. See Tye 1995: 104. 18. Tye (1995: 162) rebuts Loar’s (1990: 101) claim according to which “phenomenal qualities are psychofunctional.” Yet his account of phenomenal concepts as recognitional (indexical) concepts is similar to Loar’s account. 19. See Tye 1995: 167.

 Controversies and the Metaphysics of Mind

refer to essential properties of physical-functional states of the brain, but rather to properties of the surfaces of external things. They refer to states characterized as having representational non-conceptual abstract content, that is, to sensory nonconceptual representational states and their unique modes of presentations. Acquiring such concepts requires having these unique types of symbolic (sensory) states. This seems to clarify the cognitive difference between Mary’s mental states before and after her release; that is, this resolves ‘Mary’s puzzle.’ Tye’s theory is a fine example of how a response to problems revealed by the set of controversies that revolve around a given case is based on philosophical ideas that gradually evolved in the previous discourse. Yet, as noted above, Tye’s resolution conflicted with Loar’s resolution. This led to a new controversy the explicit topic of which is the following: Do phenomenal states consist in nonconceptual representations of external objects, or are they non-representational internal states? The controversy regarding this question is also based on apparently conflicting intuitions. On the one hand, phenomenal properties appear to be properties of external objects: The redness appears to be a property of the rose, and pain is a property of my leg. Nevertheless, the redness of the rose feels like something, and pain feels like something. It seems that the fact that they feel like something does not consist in properties of external things or of representational properties. Rather, ‘feel like’ seems to refer to subjective states. And yet it seems that the redness of the rose would not have ‘felt like’ the redness of the rose if it did not somehow involve the representation of the rose. I will not address this issue in the present context.20 7. The knowledge argument and two-dimensional semantics According to Loar, the puzzle involved in phenomenal consciousness is resolvable if one draws a distinction between concepts and properties. As the context clarifies, the introduction of the distinction was motivated by Kripke’s and Jackson’s respective arguments. It was required as a response to the challenge these arguments posed to physicalism. That this was its motivational source does not imply that the distinction between concepts and properties could have been drawn only in the context of a response to these arguments. The distinction is not necessitated by a response to these arguments. Nevertheless, what rendered it epistemically significant and promising was that it could be used to resolve a puzzle that had jeopardized a widely accepted position. Loar’s argument was a type B argument. 20. Block (2007: 533) described the conflict between these two perspectives on consciousness as “the greatest chasm in the philosophy of mind – maybe even of all of philosophy...”

Chapter 5. The antinomies of consciousness and their resolutions 

Its premises were similar to those of the dualistic positions that it rebuts, and, nevertheless, it entailed a physicalist conclusion by means of his account of phenomenal concepts. Clearly, by undermining the rational support for the dualistic positions, his argument rendered his position a relevant alternative to dualism. The attempt to resolve the problem Loar’s arguments posed for the dualist reexamined the wider background of semantic and modal theories. In the present section, I will shortly present another controversy that emerged from the previous one, namely, the controversy between Chalmers and Loar about what Chalmers labels the concept of ‘strong necessity,’ which is, in Chalmers’s view, a problematic type of modality related to positions that distinguish, as Loar does, between concepts and properties. In 1996, David Chalmers published The Conscious Mind (CM), which soon became a classic text in the field. Chalmers’s main aim in CM was to attack materialism. However, Chapter 2 of his book was largely dedicated to issues of logical and natural supervenience, reductive explanations, and to the introduction of two-dimensional semantics, which served as the conceptual background for eliminating the differences between ‘logical supervenience’ and ‘metaphysical supervenience’ and, therefore, also between ‘logical necessity’ and ‘metaphysical necessity.’ Chalmers’s overall argument aimed to disavow physicalism by claiming that conscious mental states do not logically (globally) supervene on physical states. That is, Chalmers claimed that one can conceive of a world that contains no conscious qualitative experiences which is physically identical to our word. Clearly the argument is successful only if conceivability entails possibility. Chalmers’s two-dimensional semantics aimed to clarify why this entailment holds. According to Chalmers’s two-dimensionalism, terms such as ‘water’ involve two dimensions of meaning: primary intension and secondary intension. The primary intension of a concept is a function from worlds to extensions reflecting the way that actual-world reference is fixed. In a given world, it picks out what the referent of the concept would be if that world turned out to be actual. (1996: 57)

For example, the primary intension of ‘water’ picks out in each world what the watery stuff is in that world. In our world, it picks out H2O. But in a different world, it might pick out something other than H2O. In contrast, the secondary intension picks out the same stuff in each counterfactual world as it picks out in the actual world. If ‘water’ picks out H2O in the actual world, it picks out H2O in every counterfactual world.21 21. Jackson (1998) developed a similar semantic view. I will address this issue extensively in the next chapter.

 Controversies and the Metaphysics of Mind

Primary intensions determine the set of conceivable possibilities. According to Chalmers, primary intensions are what matters to the mind/body problem. If one can conceive a world physically identical to our world that nevertheless differs in the way consciousness is distributed in it, consciousness does not logically (globally) supervene on physical states and processes and, therefore, it does not metaphysically supervene on physical states and processes. If so, conscious qualitative properties are distinguishable from all physical properties.22 Clearly, the above claim is opposed to Loar’s main contention. In Loar’s case, failure of logical supervenience will not suffice to refute materialism. The distinction between concepts and properties entails a distinction between logical necessity and metaphysical necessity, and therefore also between logical and metaphysical supervenience. There are several motivational sources for two-dimensionalism: semantical issues, modal issues, issues related to scientific explanation, and other issues as well. Chalmers’s response to Loar in CM on behalf of the anti-materialists suggests that his two-dimensional semantics was at least partly motivated by mixed intuitions expressed by Loar’s argument (intuitions that connect the ontological issues related to the mind-body problem to the semantical and modal issues).23 The debate on the nature of phenomenal consciousness evolved into a debate on the nature of meaning and modality. This was a significant shift in focus, and yet, the debate on the knowledge argument was all along a philosophical debate that involved modal and semantic intuitions that seemed to be combined with intuitions regarding the ontological nature of mind. The controversy between Chalmers and Loar was just one part of the shift that this chain of controversies took to the deeper level of the debate. That deeper level addresses the disputable modal and semantic intuitions that are combined with ontological intuitions regarding the nature of mind. In CM, Chalmers directly addressed Loar’s argument only after introducing the main claims involved in his attack of materialism. The restriction on possible worlds required by strong metaphysical necessity, the type of modality involved in Loar’s and other philosophers’ distinction between concepts and properties, and between ‘logical possibility’ and ‘metaphysical possibility’ (say, in the claim that a zombie world is logically possible and metaphysically impossible), is characterized as “brute and arbitrary.”(1996: 143)24 As Chalmers notes, the distinction between concepts and properties and the related distinction between conceivability and metaphysical possibility is not generally required. It seems to be 22. I will discuss his version of two-dimensionalism in more detail in Chapter 6. 23. See in particular Chalmers 1996: 136–138, 142–143 24. See Chalmers 1996: 137.

Chapter 5. The antinomies of consciousness and their resolutions 

invoked primarily in order to account for the nature of consciousness. It therefore begs the question to introduce the distinction between concepts and properties in the context of the debate between physicalists and dualists. Loar’s response to Chalmers (1999) appeared in a volume of Philosophy and Phenomenological Research that included reviews of Chalmers’s book. In his review, Loar denied the charge that his restriction on possible worlds is “brute and arbitrary.” As he notes, Chalmers uses ‘logical possible world’ “to mean something epistemic or conceptual.” (1999: 470) By contrast, ... if ‘possible world’ means simply a possible world on the standard Kripkean conception, then we take for granted that if a property identity is true in the actual world it is true in all possible worlds. But we cannot take this for granted on an epistemic understanding of ‘possible world.’ (1999: 470)

At the beginning of this paper, Loar addresses Chalmers’s following anti-physicalist arguments in CM: 1. In our world, there are conscious experiences. 2. There is a logically possible world physically identical to ours, in which the positive facts about consciousness in our world do not hold. 3. Therefore, facts about consciousness are farther facts about our world, over and above the physical facts. 4. So materialism is false. (1996: 123)

As Loar notes (1999: 466–467), step (3) does not follow from step (2), even if one accepts Chalmers’s two-dimensionalism. That two concepts C and P have distinct primary intensions does not entail that they must express distinct properties: Let me put the problem with Chalmers’s argument directly: it does not follow (without some further assumptions) from the fact that C and P have distinct primary intensions that they express different properties. Their having distinct primary intensions follows by a set of definitions from their being conceptually independent. But it does not follow (without some further assumptions) from their being conceptually independent that they express distinct properties... (1999: 467)

The principle according to which different primary intensions correspond to distinct properties is an additional principle that is neither explained nor entailed by the distinction between primary and secondary intensions of two-dimensional semantics to which Chalmers is committed. It might be supposed that Loar’s argument begs the question, for his charge clearly involves at least the conceivability of his own distinction between properties and concepts. Yet in my view, this objection misses the main point of his argument. Loar hints that Chalmers’s twodimensional account does not rule out the conceivability of two primary intensions

 Controversies and the Metaphysics of Mind

that pick out the same property. It does not contain an explanation why distinct primary intensions cannot pick out the same property. In his response to Loar, Chalmers (1999) examines Loar’s claims by linking them to the prospects of strong necessity. The first part of his response aims to undermine the alleged positive reasons for characterizing the identity of conscious states and physical states as having the features of strong necessities. In addressing Loar’s claims in “Phenomenal states,” Chalmers contends that there is a tension between the two main points in Loar’s account, namely, that (a) phenomenal (recognitional) concepts are distinct from physical-functional concepts, and (b) phenomenal concepts nevertheless refer not via contingent modes of presentations to the same properties to which physical-functional concepts refer. According to Chalmers: The problem with this is straightforward. The introduction of point (b) undercuts the value of point (a). The significance of point (a) for Loar lies not just in the cognitive distinctness but also the coreference of recognitional and theoretical concepts. But this coreference is explained by the two-dimensional nature of such recognitional concepts: they typically conceive of their referent as ‘the cause of such-and-such experience,’ or under some similar contingent mode of presentation. If we remove this feature of recognitional concepts (as we do in accepting (b)), we no longer have any reason to believe that recognitional concepts and distinct theoretical concepts should corefer. As things stand, in accepting (a) and (b) we are left with the observation that phenomenal concepts and physical concepts (i) are cognitively distinct, and (ii) both express the property that they refer to. It’s clear that nothing here begins to justify the coreference of phenomenal and physical concepts. In fact the situation is the opposite: in every other case of concepts satisfying (i) and (ii), they have distinct referents. One might suppose that recognitionality is doing some extra work here (thus distinguishing this case from other cases involving nonrecognitional concepts), but the only work it does for Loar is in explaining (i) and in providing cases of coreference when (ii) is false. So once (i) and (ii) are granted, there is nothing in Loar’s account to justify coreference, and his explanation of strong necessity fails. (1999: 488)

According to Loar, phenomenal concepts are recognitional concepts that express the same property they refer to. Chalmers contends that the introduction of point (b) undercuts point (a). Phenomenal concepts are cognitively distinct from physical-functional concepts and they corefer. Coreference is explained if they conceive their referent by means of contingent modes of presentation such as ‘thecause-of-such-and-such-experience.’ This is clearly what happens in contexts in which recognitional concepts are introduced (as Loar himself acknowledges). It is therefore not trivial to import this concept to a context in which recognitional

Chapter 5. The antinomies of consciousness and their resolutions 

concepts express the property they refer to. In the first case (the normal context), the two-dimensional account is able to explain coreference. But if it is claimed that recognitional concepts differ from physical-functional concepts even though they express the same property (without involving contingent modes of presentation), one undermines the reasons that explain why recognitional concepts refer to the same properties as physical-functional concepts. In other words, coreference seems to be left without an explanation. As Chalmers notes, Loar’s arguments ... often seem to have the form (1) Given that phenomenal concepts and physicaltheoretical concepts corefer (as is typical with recognitional concepts), and (2) Given that phenomenal concepts express the property that they refer to (as is unusual with recognitional concepts), then (3) phenomenal concepts will express physical theoretical properties. (1999: 488)

Even if one assumes the validity of this argument, the alleged feature of Loar’s phenomenal concepts leave the fact that the property to which C and P refer to is the same property unexplained.25 Interestingly, this is not a very different accusation than that which Loar made against Chalmers. Chalmers is not contending that given Loar’s theory, phenomenal concepts and physical-functional concepts cannot corefer, but rather that the theory lacks the means to maintain that they must corefer. Additional principles are required for that purpose. In other words, it appears that the controversy reveals that each of the two positions suffers from an inadequacy. Chalmers was not unaware of this result. Although he believes there are reasons to favor two-dimensionalism over the modal theory that involves strong necessity, he recognizes the problem revealed by Loar’s arguments and by his response: An opponent could hold that there are strong psychophysical necessities quite unlike the usual a posteriori necessities, that the space of metaphysical possible worlds is smaller than the space of logically possible (or conceivable) worlds, and that zombie worlds are excluded. If my arguments have been successful so far, I have removed many of the positive reasons for believing this. But perhaps the worry is that I have not shown why it couldn’t be true. So why must it be false? That is a deep and interesting question on which I hope to write at length in the future... (1999: 489)

In his 1999 paper, Chalmers’s response to this concern merely identifies the possible source of the problem: the idea of strong necessity involves an erroneous 25. As Levine notes (2001: 58–59), Chalmers does not have a knockdown argument against Loar. Moreover, according to Levine, it is reasonable to accept brute necessities in order to secure materialism, a claim that exemplifies the connection between logical issues and factual issues that characterizes metaphysical intuitions.

 Controversies and the Metaphysics of Mind

conception of modality (1999: 489). A position that operates with four modal notions (instead of two) threatens to undermine the rational basis of our conception of modality. In this paper, Chalmers merely indicated the need to probe deeper into the grounds of our modal notions.26 He carried out this task in a series of papers published in recent years. As we shall see in the next chapter, the queries regarding two-dimensional semantics, strong necessities, and psychophysical explanations motivated another round of debates, this time on Jackson and Chalmers’s two-dimensional semantics and the availability of psychophysical explanations of consciousness. 8. Mind and illusions Near the beginning of the new millennium, almost 20 years after the publication of EP, Jackson abandoned dualism.27 In an article analyzing the differences between his past and present positions Jackson depicts his earlier dualistic stance as grounded upon an illusion that is both natural and deceptive. Jackson is not alone in employing this move in the philosophical debate addressing the nature of consciousness. A similar appeal to illusion is expressed by Tye (2000) and by Papineau (2002). Invoking cognitive illusions as the possible source of the debate is a significant move in philosophical and metaphysical controversies that is far more prevalent as an implicit move than what might be supposed on the basis of the overt meanings of the relevant texts. On the one hand, it expresses the avowed powerlessness of those that employ this move to provide further reasons and arguments with which they will be able to persuade those who are not persuaded by the arguments they can provide. On the other hand, it expresses the belief that one can explain what causes the temptation to endorse the opposed view in the face of arguments that are considered by those that employ this move to be decisively persuasive (in the rational sense of the term). In a book that probably contains the most developed account of the phenomenal concept strategy, Papineau also provides an explanation of the cognitive illusions related to conscious experiences. He attributes an important role to the explanation of these illusions in the resolution that he offers for the philosophical 26. In his (2004) paper “Phenomenal concepts and the knowledge argument” on the topic of phenomenal concepts, Chalmers once again criticized Loar’s position. Although his argument there makes the relevance of two-dimensional semantics to the knowledge argument explicit, the argument reiterates Chalmers’s previous claims: the arbitrariness of strong necessities and the inexplicability of their alleged coreference with physical-functional concepts. 27. See in particular Jackson (2004).

Chapter 5. The antinomies of consciousness and their resolutions 

problems of consciousness. The resolution does not merely require a philosophical account of consciousness, but also a therapy.28 Papineau believes that the physicalist has a comprehensive account of conscious mental states at her disposal that does not leave any real puzzle unresolved, and yet, I recognize, though, that there certainly seems to be a mystery here. But I don’t think that this is because there is something unfathomable about the thesis that conscious states are material. Rather, it is because something prevents us from ever fully accepting this thesis in the first place, and convinces us that conscious states are not material states. And then, of course, everything does seem mysterious. (2002: 2)

The “something” that “prevents us” from fully accepting the physicalist’s contention, that is, to be persuaded by his arguments, is what Papineau terms ‘the intuition of distinctness:’ Let me now focus on the intuition of distinctness itself. In my view this is what makes the mind-body problem seem so intractable. Even given all the arguments, intuition continues to object to mind-brain identity. How can pain (which hurts so) possibly be the same thing as insensate molecules rushing around in nerve fibers? Or, to repeat Colin McGinn’s question, how can our vivid technicolor phenomenology (our experience of reds and purples and so on) possibly be the same as cellular activity in grey matter? ... I shall try to explain this intuitive resistance to materialism about the mind. I think there is indeed something special about the mind-brain relation. It generates this overwhelming intuition of distinctness. Even convinced materialists are likely to feel the pull of this intuition. I know that in my own case it continues to press, despite any amount of immersion in the arguments of the previous chapters.(2002: 161)

As Papineau aims to explain, the intuition of distinctness can be explained by means of the elements of his account of conscious experiences. Tye and Jackson, two other philosophers of mind that appeal to cognitive illusions in their attempts to resolve the puzzle of consciousness, also base their account of cognitive illusions on their theories of conscious experiences. Their theories, however, conflict significantly with Papineau’s. I will not address Papineau’s account of the nature of consciousness and of the cognitive illusion related to the reflective examination of the nature of conscious experiences in detail. Nor will I discuss Tye’s or Jackson’s alternative accounts in detail. The contention that the puzzle of consciousness is based on a cognitive illusion is merely one more defeasible contention involved in the debate on the nature of consciousness. The plausibility of this contention depends on the soundness of the theory that supports it. 28. See Papineau 2002: 3.

chapter 6

A priori knowledge and the explanatory gap 1. Introduction In 1983, a year after the publication of Jackson’s “Epiphenomenal qualia,” Joseph Levine published “Materialism and qualia: The explanatory gap,” in which the concept of the ‘explanatory gap’ was introduced into the debate on consciousness. This paper marks the beginning of another chain of controversies that deal with the question of whether the scientific explanation of the qualitative features of conscious experiences involves a unique problem. These controversies are not limited to the issue of the nature of consciousness. They deal with questions regarding the meaning of ‘conceivability,’ ‘apriority,’ ‘necessity,’ ‘identity,’ ‘scientific reduction,’ ‘scientific explanation,’ and related topics. Nevertheless, the important role of the problem of consciousness in these debates justifies viewing them as part of the third stage of controversies that were motivated by Kripke’s, Nagel’s, and Jackson’s respective arguments. My intention in the present chapter will be to examine one thread in this complex fabric of controversies. I will focus primarily on the controversy between Block & Stalnaker (1999) and Chalmers & Jackson (2001).1 This controversy exemplifies the extent to which a conflict between metaphysical intuitions is rooted in a wide range of topics. This is most notable with regard to the role of two-dimensional semantics (developed in the mid-1990s by Jackson and Chalmers) in this controversy. Chalmers (1996) and Jackson (1998) also introduced the twodimensional framework in order to explicate the semantic and modal basis of their novel arguments for dualism and for the explanatory gap. Drawing the attention of many philosophers, their use of two-dimensional semantics in this

1. Block & Stalnaker’s attack of Chalmers and Jackson was somewhat surprising. Jackson (1998) and Chalmers (1996) considered Stalnaker to be one of the main developers of the twodimensional framework, and therefore a potential ally in their project. Similarly, already in the 1970s, Block was one of the pioneers who emphasized the difficulties that a functionalist account of the mind faces in dealing with consciousness. (see Block 1978). Hence, this controversy can be viewed as indicating the parting of ways in contemporary philosophy of mind.

 Controversies and the Metaphysics of Mind

context established their positions as relevant alternatives to existing positions not only in the philosophy of mind but also in semantics.2 As we shall see, Block & Stalnaker do not accept Chalmers & Jackson’s position as a relevant alternative. Rather, their arguments aim to undermine the status of Chalmers & Jackson’s position as relevant alternatives. The fact that Block & Stalnaker attempted to undermine the main thread of ideas that underlies Jackson & Chalmers’s position had a significant effect on the relevant philosophical discourse.3 I will begin by presenting the origins of this controversy: Levine’s classic argument for an explanatory gap and Jackson’s and Chalmers’s accounts that involve the two-dimensional framework of the nature of serious metaphysics, conceptual analysis, and scientific reduction. I will then proceed to examine the controversy itself and its results. 2. Levine’s argument Levine’s argument for an explanatory gap can be viewed as a response to Kripke’s modal argument and to Jackson’s knowledge argument. Although Levine rejects the anti-physicalist conclusions of both arguments, he nevertheless considers them to be relevant alternatives in the sense of the term defined here. In what follows, I will primarily consider Levine’s argument in “On leaving out what it is like,” (1993) though my main claims here also apply to his position in The Purple Haze. In the first part of his paper, Levine criticizes the presumed anti-materialistic implications of Kripke’s modal argument and of Jackson’s knowledge argument. In Levine’s view, these arguments do not entail their conclusions. They seem to overlook features that could be used in a physicalist account of conscious states. Nevertheless, Levine believes that although the anti-physicalist conclusions are 2. Consider, for example, the following quotation from Soames’s book Reference and Description, which is devoted to a thorough critique of two-dimensionalism: “The subject under investigation, ambitious two-dimensionalism, has in my view, been one of the most important philosophical developments of the past twenty-five years. It is the most concerted and systematic attempt among many to reinstate descriptivism in the philosophy of language, internalism in the philosophy of mind, and some version of conceptualism in our understanding of modality in the face of the challenges that rocked these positions more than thirty years or so.” (2005: 329) 3. After the publication of the exchange between Block & Stalnaker and Chalmers & Jackson, several books and papers addressing the same issues were published. See in particular Szabó Gendler & Hawthorne (2002), Soams (2005), Kim (2005), García-Capitorino & Macià (2006), Thorin & Walter (2007).

Chapter 6. A priori knowledge and the explanatory gap 

not entailed by the argument, the intuitions involved in both arguments imply an epistemological problem rather than a metaphysical one. As Levine presents it, the problem is revealed by the differences between the available accounts of the appearance of contingency related to necessary a posteriori judgments such as ‘water = H2O’ and of identity statements between phenomenal states and physical states. If, as Levine believes, in all scientific explanations, the explanans necessarily entails the explanandum, then it is possible to point out the explanatory gap related to the qualitative features of conscious phenomena. Levine presents the task of explaining the qualitative features of conscious phenomena and the relevance of the difficulties involved in such explanations to the feasibility of physicalism as follows: I have argued that on a metaphysical reading of “leave something out,” Cartesian conceivability arguments cannot establish that physicalist theories of mind leave something out. However, there is also an epistemological sense of “leave something out,” and it is this sense that conceivability arguments, being epistemological in nature, can reveal a deep inadequacy in physicalist theories of mind. For a physicalist theory to be successful, it is not only necessary that it provides a physical description of mental states and properties, but also that it provides an explanation of these states and properties...what is at issue is the ability to explain qualitative character itself; why it is like what it is like to see red or feel pain. Conceivability arguments serve to demonstrate the inability of physicalist theories to provide just this sort of explanation of qualitative character. (1997: 548)

A functional description of a mental state is a reasonable candidate for a physical description of a mental state. In “Troubles with functionalism” (1978) Block claimed that functionalism involves serious problems when it attempts to deal with consciousness. But Levine goes beyond the argument presented by Block. In his view, functionalism must involve explanations of the qualitative features of conscious mental states. The “deep inadequacy” of functionalism or of other physicalist theories is revealed when one faces the task of explaining qualia. This “deep inadequacy” involved in physicalist theories of mind results from the apparent disanalogy between the available explanation of qualitative conscious phenomena and standard examples of explanations in the physical sciences, say, the explanation of the macro-features of water, and the explanation of heat by the physical and chemical theories. In Levine’s view, although the qualitative features of conscious phenomena are macro features of mental states, they do not allow explanations similar to the explanation by which, say, the macro features of water are explained. Levine’s first step is to note that the appearance of contingency of necessary a posteriori statements such as (1) can be explained by means of the contingency of (1’):

 Controversies and the Metaphysics of Mind

(1) Water is H2O. (1’) The substance that manifests [such and such macro properties of water] is H2O. (1’) is contingent since a substance could exhibit the macro properties of water without being H2O. Nevertheless, given that the macro properties of H2O are explainable by means of the relevant physical theory, it is inconceivable that the stuff which is H2O will fail to manifest the macro features of water. The account of the contingency of (1) is not analogous to the contingency of (2):

(2) Pain = C fiber stimulation.

Statement (2’) below is conceivable in both directions: (2’) The state that manifests the macro (phenomenological) features of pain is C-fiber stimulation. It is not merely conceivable that something may exhibit the macro (phenomenological) features of pain without being C-fiber stimulation (as required, for example, by the multiple realizability of functional states). It is also conceivable that a state that is C-fiber stimulation does not exhibit the phenomenological features of pain. This was the central point in Kripke’s modal argument for the real distinction between brain states and qualitative mental states. Yet, as I noted above, in contrast to Kripke, Levine does not think that the problem revealed by the disanalogy between (2’) and (1’) is a metaphysical problem but, rather, an epistemological one. Although the concept of scientific explanation involves the necessary connection between the explained phenomena and the explaining phenomena, the necessity at stake is epistemological and not metaphysical.4 Clearly, given that we have an explanation E of (1’), it is not merely inconceivable that a particular kind of stuff will be H2O and that it will not manifest the macro features of water. It is metaphysically impossible for a substance to be H2O and not to manifest the macro features of water. But the contingency of (2’) in both directions does not entail, in Levine’s view, the metaphysical possibility of a state being C-fiber stimulation and not manifesting the phenomenological features of pain. Levine’s account of the necessary connection between the explanans and the explanandum is founded on a link between the notion of scientific explanation and the notion of scientific reduction. As he notes, ... we justify the claim that water is H2O by tracing the causal responsibility for, and the explicability of, the various superficial properties by which we identify water... 4. As Levine notes (1997: 549), the idea that “explanations involve showing that the explanandum follows from the explanans” seem to commit him to the controversial deductivenomological model of explanation. But his account is also consistent with different models of explanations, for instance that of Cummins’s (1983) property theories.

Chapter 6. A priori knowledge and the explanatory gap 

But suppose someone pressed further, asking why being causally responsible for this particular syndrome of superficial properties should be so crucial. Well, we would say, what else could it take to count as water? But the source of this “what else” is obscure. In fact, I think we have to recognize an a priori element in our justification. That is, what justifies us in basing the identification of water with H2O on the causal responsibility of H2O for the typical behavior of water is the fact that our very concept of water is of a substance that plays such-and-such a causal role. (1997: 550)

Levine’s account of the problem of the explanatory gap can be summed up as follows: In scientific explanations, the explanans a priori entails the explanandum. In Levine’s view, the concept that does the explaining is a functional concept, which is defined in terms of the causal powers to produce the macro features of water. If one cannot produce an account of why it is inconceivable that a state is C-fiber stimulation and that it does not manifest the phenomenological features of pain, then one does not possess an explanation of pain. In other words, the feasibility of such an explanation requires a functional reduction of qualia. To the extent that qualia resist functionalization, they are unexplainable. It is clear that Levine believes that we currently lack explanations of the qualitative features of conscious experiences. What is not clear, however, is whether he believes that our current ignorance is merely temporary, something that could be remedied by further scientific inquiries, or whether the ignorance is of a deeper kind. In fact, Levine’s argument seems to oscillate between these two possibilities. On the one hand, his argument merely reveals the conditions that must be satisfied by explanations of the qualitative features of experiences, which we apparently do not yet possess. On the other hand, what he says seems to imply that the problem of the explanatory gap is a deeper problem. I will attempt to consider this question by examining the way Levine addresses Jackson’s knowledge argument. Let us remember that Mary’s case is an ideal case. Mary has all the available physical information regarding the perception of colored objects. We may therefore assume that she possesses all the elements required for an explanation. But as Jackson’s thought experiment assumes, Mary does not know what it feels like to see red objects. She apparently does not have an explanation of the qualitative features instantiated by experiencing red things. Given her ideal situation, it cannot be maintained that her ignorance can be removed if she is allowed to undertake further investigations. Since she already knows all the physical information, she cannot discover any additional physical fact that will assist her in overcoming her ignorance. Might her ignorance be merely epistemic? That is, could it indicate an explanatory gap without conveying anti-physicalist implications? When Levine discusses Jackson’s knowledge argument (when he examines conceivability arguments for mind-body dualism), he

 Controversies and the Metaphysics of Mind

refers to Horgan’s (1984) response to Jackson as a possible type of response that avoids the metaphysical implications of Mary’s conceivable ignorance. Let me remind the reader, according to this line of response, the meaning of ‘physical’ is not equivalent to ‘entities grasped by means of physical theories.’ Given Levine’s account, we may assume that physical theories and the objects and properties they involve provide causal explanations of observed phenomena. Apparently, not all physical properties of objects are causal properties that serve a role in causal explanations. The conceivability of such properties entails a gap between the scope of what can be grasped and explained within the physical sciences, and the nature of physical objects. This seems to be consistent with Levine’s intended claim that the existence of an explanatory gap does not have ontological implications of the anti-physicalist type. But Levine’s materialism is primarily based on the causal argument discussed in Chapter 5.5 Hence, given Mary’s idealized situation, it is questionable whether it is indeed conceivable that the following statements are possibly true together: a. [the causal intuition] The qualitative properties of experience are states that are caused by other states and processes and that serve causal roles in producing behavior. b. Relevant information regarding the occurrence of physical events and processes is information that is causally relevant to their occurrence. c. Mary knows all the causally relevant physical information there is to know about the perception of red objects; and d. [the knowledge intuition] Mary does not know what it feels like to see red objects. It seems that if (a)-(d) are true, then, at least in the ideal case, an explanatory gap entails an ontological gap.6 For if (a) is true and Mary does not know what it feels like to see red things, then she lacks some information that is causally relevant to the occurrence of the 5. See in particular his account of materialism in his book The Purple Haze (2001). 6. In “Conceivability and the metaphysics of mind,” (1998) Levine presents some arguments that aim to substantiate the conceivability/possibility distinction. In Levine’s view, a subject may know how to use two terms A and B without knowing that they refer to the same property or to properties that are necessarily instantiated together, if the mode of presentation that is linked, say, to A, is a non-ascritptive mode of presentation. Non-ascriptive modes of presentations “... establish relations... “behind the scenes,” not by being cognitively grasped by the subject. The subject’s competence with the term, her ‘knowledge’ of the meaning, consists entirely in her instantiating the requisite relation to something in the world. A priori ignorance is thus attributable to ignorance of what constitutes the meaning in this case”. (Levine: 1998: 458) It is clear however, that this explanation of the differences between conceivability and possibility does not work in Mary’s case.

Chapter 6. A priori knowledge and the explanatory gap 

conscious events and processes, and hence information that is causally relevant to the states and processes involved in the perception of red objects. Alternatively, the alleged explanatory gap entails some conceptual barriers that do not allow the knowledge of some facts about physical properties and objects (hence, Mary will not be able to have all the relevant physical information).7 As we shall see in what follows, the intuition that the latter possibility should be ruled out a priori is the main intuition that underlies Jackson’s and Chalmers’s positions. 3. Jackson: Serious metaphysics and conceptual analysis In From Metaphysics to Ethics (1998), Jackson provides an account of what he dubs ‘serious metaphysics,’ which deals with the connections between metaphysics, supervenience, conceptual analysis, and a priori knowledge.8 The issues dealt with in this book and in related papers are not only related to the mind-body problem – they address metaphysics and conceptual analysis as such. The mindbody problem influences the theory defended by Jackson in various ways. Jackson does not elaborate on the concept of explanation.9 Yet the concluding judgment of his account is that our modal concepts do not allow for the existence of a gap between metaphysical possibility and epistemic possibility, or conceivability. 3.1

Entry by entailment

In order to analyze the role of the concept of supervenience in serious metaphysics, Jackson uses physicalism as an example. The main idea (1998: 9) is that a complete physical story about the nature of our world entails a complete story about the psychological nature of our world. As Jackson notes (1998: 9–10), the notion of supervenience required by physicalism is that of global supervenience 7. According to Perry (2001: Chapter 4), epiphenomenalism need not be construed as a position that must involve property dualism. Epiphenomenal properties may, rather, be physical properties. See also Churchland (2004). Yet it is quite clear, I suppose, that if Mary has all the physical information that is causally relevant to the perception of red objects, she must know these epiphenomenal physical states that are produced by some physical causal mechanisms. Alternatively, it could be maintained that, given that qualia are epiphenomenal physical states, it is not possible to know all the physical facts that are relevant to the perception of red objects. But this contradicts the main assumption of the knowledge argument. 8. Jackson’s book is based on lectures given in 1995. 9. As we shall see in the next section, Chalmers (1996) provides a detailed account of the nature of explanation by employing a conceptual scheme similar to Jackson’s scheme that has the same result as the above.

 Controversies and the Metaphysics of Mind

and not of intra-world supervenience. The main thesis of physicalism must be expressed “in terms of quantification over worlds, rather than in terms of quantifications over individual in worlds.” (ibid.) Statement (A) below seems to be a candidate for the central thesis of physicalism: (A) Any two possible worlds that are physical duplicates (instantiated physical property, law and relation for instantiated physical property, law and relation identical) are duplicates simpliciter. (1998: 11)10

Yet according to Jackson (ibid.), physicalism need not be construed as a thesis about every possible world. It may be compatible with Cartesian dualism. According to Physicalism, our world cannot contain properties and objects that are not physical. The main thesis of physicalism could be construed as follows: (B) Any world which is a minimal physical duplicate of our world is a duplicate simpliciter of our world. (1998: 12)

As Jackson notes, physicalism entails (B), and (B) entails physicalism. For if (B) is true, it follows that physical nature “exhausts all contingent nature and in particular all psychological nature.”(1998: 13) In Jackson’s view, this seems to be linked to another idea. If (B) indeed follows from physicalism, then “the psychological account of our world is entailed by the physical account of our world.” (1998: 24) As we shall see below, in Jackson’s account of conceptual analysis, modality, and the related two-dimensional semantics, if physicalism is true, the psychological nature of our world is a priori entailed by the physical nature of our world. However, ‘entails’ as such does not mean ‘conceptually entails’ or ‘a priori entails.’ (1998: 25) Rather, it means ‘necessary truth preservation,’ ‘necessary determination’ or ‘fixing.’ ‘Entry by entailment’ means locating the features of possible worlds expressed in one vocabulary in terms of the features of a possible world expressed in another vocabulary. This is carried out by showing how the huge complex sentence that expresses the first features entails the complex sentence that expresses the second features. Here is how Jackson clarifies this notion: Let Φ be a story as told in purely physical terms, which is true at the actual world and all the minimal physical duplicates of the actual world, and false elsewhere; Φ is a hugely complex, purely physical account of our world. Let ψ be any true sentence which is about the psychological nature of our world in the sense that it can only come false by things being different psychologically from the way they actually are...Now if (B) is true, every world at which Φ is true is a duplicate simpliciter of our world, and so a fortiori a psychological duplicate of our world. But then every world at which Φ is true is a world at which ψ is true – that is, Φ entails ψ. (1998: 25) 10. For a similar position see Lewis 1983.

Chapter 6. A priori knowledge and the explanatory gap 

Let me note in passing that the connection between Jackson’s earlier knowledge argument discussed in Chapter 4 and his account of ‘entry by entailment’ is quite clear.11 Given that Mary knows all the relevant physical information, she knows Φ and, therefore, she must know ψ. If ψ exhausts the psychological character of our world, then Mary must have known what it feels like to see red objects before she was released. ‘Entailment’ as Jackson construes this term, does not involve ‘knowledge’ or any other human capacity. It applies directly to the world. Yet even if one endorses this realistic view, the connection between knowledge and entailment in Mary’s case naturally involves the idea of an ideal case.12 I will discuss the significance of this point later. 3.2

Conceptual analysis

Jackson’s account of ‘conceptual analysis’ is based primarily on his account of twodimensional possible-world semantics. I have already discussed some features of his theory in Chapter 5. In the present section, I will primarily present those elements in Jackson’s theory that are relevant to the debate with Block & Stalnaker. We may first note that for Jackson, conceptual analysis does not involve the presentation of definitions but, rather, the examination of the ways in which conceived possibilities are linked to one’s intuitions. Intuitions, as I noted in Chapter 3, are epistemic sources that, in Jackson’s view, reveal ordinary conception, that is, they reveal the folk theory assumed to be shared by all of us (Jackson 1998: 38). Philosophical or metaphysical thought experiments have precisely this role, that is, to present conceivable possibilities and to reveal the features of our shared folk theory by examining our intuitions. Yet as Jackson observes (1998: 42), one has to distinguish between two kinds of conceptual analysis: the modest and the immodest kind. In practicing the modest kind of conceptual analysis, one merely gives voice to one’s (supposedly shared) intuitions and “not to making any claim, one way or the other, about what the world is like.” (ibid.) By contrast, the

11. Compare Tye 2009: 32. 12. As Balog notes: “For someone to be able to tell, given a full fundamental description of the world considered as actual, whether a thought would be true there, one would have to understand the fundamental language. But it is implausible that anyone currently – even leading physicists – possesses the requisite fundamental concepts. AA, [the A Priori Availability Thesis] as it is stated, is an idealization; what Jackson has in mind is what an ideal logician, once she has learned the fundamental language of physics, and not being bound by time constraints, powers of physical endurance, etc. could figure out under optimal circumstances.” (2001: 648)

 Controversies and the Metaphysics of Mind

non-modest kind attributes a key role to conceptual analysis “in an argument concerning what the world is like.”(1998: 43)13 The distinction between the modest and immodest kinds of conceptual analysis in fact reflects the complex relations between the empirical and the a priori levels of conceptual analysis. In Jackson’s view, our language (and therefore conceptual analysis) allows a priori (though fallible) knowledge. But knowledge of the content assigned to a term is also an empirical matter and depends on the features of the actual world. The limits of conceptual analysis required by the distinction between the modest and immodest kinds of conceptual analysis results primarily from the two dimensions of meaning, one that allows for a priori knowledge, and one that requires empirical knowledge. According to Jackson (1998: 47), the roots of two dimensional semantics are found in Kripke’s Naming and Necessity (1980) and in works by Stalnaker (1978), Kaplan (1979, 1989), Davies & Humberstone (1980) and others. Jackson’s main idea is that possible worlds account of meaning has to distinguish between A- extensions and C- extensions of terms. In order to determine the A- extension of a term one has to consider a world as actual. In this case, “we are considering what the term applies to in w, given or under the supposition that w is the actual world, our world.” (1998: 48) By contrast, the C- extension is the following: In the second case, we are considering, for each world w, what T applies to in w given whatever world is in fact the actual world, and so we are, for all worlds except the actual world, considering the extension of T in a counterfactual world. We can call this the C extension of T in w. (1998: 48)

For example, consider the meaning of ‘water.’ As noted several times before, water has an underlying physical nature that can only be revealed a posteriori. In our case, ‘H2O’ refers to its underlying nature. But ‘water’ also connotes features such as ‘being the drinkable, tasteless transparent substance that fills the oceans and lakes... etc.’ When one considers a world as actual, its A-extension is the stuff that in the actual world is the extension of ‘water.’ Its A-intension is a function that assigns to each world the A-extension of ‘water.’ (ibid.) In considering a world 13. As Stalnaker notes (2001: 633), Jackson “thinks that a lot of implicit metaphysics can be extracted from conceptual analyses – from descriptions of what we mean by our words... Jackson’s modest metaphysician recognizes that we cannot move directly from the conceptual analyses that yield such results to metaphysical conclusions – that determinism or physicalism is false, that we are Cartesian egos.” Stalnaker adds (ibid.) that the only alternative to conceptual analysis that Jackson seems to allow is eliminativism. In contrast to Jackson, who believes that conceptual analysis is necessary for the possibility of metaphysics, Stalnaker maintains (2001: 636) that “[...] it is because of the unavailability of conceptual analyses that metaphysics is necessary.”

Chapter 6. A priori knowledge and the explanatory gap 

as counterfactual, one determines which stuff in the counterfactual world is the extension of ‘water’ given that the actual world is a certain world w. The function that assigns to each world w the C-extension in w is the C-intension of water (ibid.). For example, if in w considered as actual (not our world) the stuff that fills the oceans and lakes is XYZ, ‘water’ refers to XYZ in w. Given that in w considered as actual, ‘water’ refers to XYZ, ‘water’ refers in any counterfactual world w’ to XYZ. The distinction between the two types of intensions and extensions associated with the words of a given language seems to allow a novel account of the distinction between a priori knowledge and empirical knowledge.14 As Jackson notes, When a term’s A-extension and C-extension differ at some worlds...there is a crucial difference between the epistemic status of a term’s A-extension and it’s C-extension. To know a term’s C-extension we need to know something about the actual world. Although we understood the word ‘water’ before 1750, we did not know its C-extension at a world for any world other than the actual world...By contrast we did know the A-extension of ‘water’ at every world, for its A-extension does not depend on the nature of the actual world. (1998: 50)

The epistemic differences between A-extensions and C-extensions seem to directly imply an account of a priori knowledge that bypasses the traditional difficulties involved in such knowledge. Knowledge of A-extensions does not require one to know which world is the actual world: What we can know independently of knowing what the actual world is like can properly be called a priori. The sense in which conceptual analysis involves the a priori is that it concerns A-extensions at worlds, and so A-intensions, and accordingly concerns something that does, or does not, obtain independently of how things actually are. When we do conceptual analysis of K-hood, we address the question of what it takes to be a K in the sense of when it is right, and when it is wrong, to describe some situation in terms of ‘K’, and so we make explicit what our subject is when we discuss Ks. The part of the enterprise that addresses questions of what things are K at a world, under the supposition that that world is the actual world, is the a priori part of conceptual analysis... (1998: 51)

The question that Jackson’s account of the first two issues leaves open is “the relationship between what we are calling an entailment or necessary determination or fixing thesis, and issue of a priori deducibility.” (1998: 68) Again, the main example that Jackson employs is the metaphysical doctrine of physicalism. As he already established, if physicalism is a true doctrine, the above statement (B) must be true. If (B) is true, then the psychological nature of our world is entailed by the 14. It should be stressed, however, that in Jackson’s view (2001: 662) the distinction between A intensions and C intensions applies to words and not to thoughts.

 Controversies and the Metaphysics of Mind

physical nature of our world. But is this entailment an a priori entailment? As Jackson notes, his two-dimensional semantics and the related account of a priori knowledge seem to allow the a posteriori nature of the entailment of ψ by Φ. For it could be maintained that “although every possible world with a certain H2O character has a certain water character, we cannot a priori deduce water character from H2O character, because the necessary identity of water with H2O is a posteriori.” (1998: 68) Similarly, the main thesis of physicalism, according to which every world that is a minimal physical duplication of our world is a world identical to our world with respect to all the psychological features, also seems to be a case of a posteriori necessity. In other words, this thesis seems to grant that physicalism is not committed to the a priori entailment of the psychological character of the world from its physical character. Nevertheless, as Jackson aims to demonstrate in chapter (3) of From Metaphysics to Ethics, physicalism qua metaphysical doctrine is committed to the a priori entailment of the psychological character of the world from its physical character. It is not difficult to see the connection between Jackson’s goals in chapter (3) of his book and Levine’s contention according to which the explanatory gap is merely an epistemological problem. By claiming that physicalism qua metaphysical doctrine is committed to the a priori entailment of the psychological features of our world from the physical features of our world, Jackson, in fact, implicitly denies Levine’s contention that the explanatory gap has no bearing on physicalism conceived as a metaphysical doctrine. Jackson’s main intention in this chapter is to attack the alleged distinction between two types of necessity, metaphysical necessity on the one hand, and conceptual or epistemic necessity, the necessity linked to the notion of conceivability, on the other hand. In Chapter 5 of this book, I already examined what role the distinction between these two types of necessity plays in the controversy between Chalmers and Loar. Here a similar conflict seems to arise with respect to the connection between the notions of explanation, a priori entailment, and physicalism conceived as a metaphysical doctrine. At least if one considers Levine’s arguments with respect to ideal cases such as Mary’s case, the supposition that the explanatory gap is compatible with physicalism apparently implies commitment to two types of necessity. One type of necessity is required for the a priori entailment of the explanandum by the explanans. The other type is required for the necessity that binds qualia to their underlying physical nature. Given that the explanatory gap also exists in the ideal case, the knowable type of necessity must be distinct from the unknowable type. If one denies the distinction between these two types of necessity, and if one, in addition, holds the view that physicalism is committed to the a priori entailment of the psychological character of the world from the physical character of the world, as Jackson maintains, an explanatory gap entails a metaphysical gap.

Chapter 6. A priori knowledge and the explanatory gap 

Jackson employs two kinds of arguments against the existence of two types of necessity. First, by employing his distinction between A-extensions and A-intensions, on the one hand, and C-extensions and C-intensions, on the other hand, he rules out the assumption that the account of the differences between a posteriori necessity and a priori knowledge requires two notions of necessity. As the standard examples clarify, a person may understand the sentence below without knowing the underlying physical nature of gold in the actual world:

(1) Wedding rings are made of gold.

In Jackson’s view, the distinction between a priori knowledge and a posteriori necessity could be explained by noting that understanding a sentence does not entail knowledge of the truth conditions of the sentence used in a certain context, that is, of the proposition that it expresses. It does, however, involve “knowing how the proposition expressed depends on the context of utterance.”(1998: 73) According to another interpretation of the distinction between a priori knowledge and a posteriori necessity, the differences between knowing the truth conditions of sentences and knowing how the proposition expressed depends on the context in which it is uttered indicates that there are in fact two propositions involved in understanding a sentence.15 In Jackson’s view, the difference between these two accounts is superficial (1998: 77). At any event, both interpretations of the connection between a priori knowledge and a posteriori necessity do not require any distinction between two types of necessity. By using Ockham’s razor, one can therefore eliminate the distinction between the two types of necessity.16 One might ask what all this has to do with the claim that physicalism is committed to the a priori deducibility of ψ from Φ. The answer is straightforward. As Jackson notes, (3) below is not a priori entailed by (2):

(2) H2O covers most of the earth. (3) Therefore, water covers most of the earth.

Nevertheless, (3) is a priori entailed from (2) and (2a): (2a) H2O is the water of our acquaintance. (2a) expresses a proposition that identifies the actual world. Although (2a) is empirical, the entailment itself is a priori. 15. Jackson (1998: 76) ascribes this view to Tichy, Chalmers, Lewis, and Stalnaker. 16. Moreover, as Jackson notes, and this is his second contention against the theory that maintains that there are two senses of ‘necessity,’ the way in which Putnam and Kripke introduced the necessary a posteriori implies the denial of a distinction between two senses of ‘necessity.’ (1998: 77–78)

 Controversies and the Metaphysics of Mind

It is not difficult to see how this argument is related to Jackson’s knowledge argument. Although Mary is isolated from the world, given that she knows all the physical information there is to know about our world, she knows the truth of a sentence that is analogous to (2a); that is, a sentence that describes her actual world. Therefore, if the explanatory gap that physicalism faces in trying to explain the qualitative features of experience exists in Mary’s case, one must conclude that physicalism is a false metaphysical doctrine. 4. Chalmers: Supervenience and scientific explanations As noted above, Jackson is no longer a metaphysical dualist, and Chalmers continues to hold views close enough to mind-body dualism. Yet, they hold similar views regarding conceptual analysis, two-dimensional semantics, and the nature of physicalism. In the present section I will briefly present some of Chalmers’s claims regarding supervenience and reductive explanations that are relevant to the debate with Block & Stalnaker. According to Chalmers, supervenience is “a relation between two sets of properties: B-properties – intuitively, the high-level properties – and A-properties, which are the more basic low-level properties.” (1996: 33) Like Jackson, Chalmers distinguishes between several concepts of supervenience: local supervenience, global supervenience, natural supervenience, and logical supervenience. B properties locally supervene on A properties if any two individuals that instantiate the same A properties instantiate the same B properties (1996: 33–34). For instance, geometrical shape locally supervenes on physical properties. But values, say the value of the Mona Lisa, do not locally supervene on the physical properties of individuals. The value of the Mona Lisa globally supervenes on the physical properties of the actual world. In general, B-properties globally supervene on A-properties if no two possible worlds that are identical with respect to their A-properties differ in their B-properties (ibid.). B-properties logically supervene on A-properties if “no two logically possible situations are identical with respect to their A-properties but distinct with respect to their B-properties.” (1996: 35) As Chalmers notes (ibid.), logical supervenience is not defined in terms of deducibility in a formal logical system but rather in terms of possible worlds that are taken as primitives here. Bproperties naturally supervene on A-properties, if any two natural possible situations that are identical with respect to their A-properties have the same B-Properties (1996: 36). Like Jackson, Chalmers denies that there is a distinction to be drawn between logically possible worlds and metaphysical possible worlds. The notion of supervenience required for physicalism is that of logical global supervenience qualified by the limiting condition that it is our world with which we

Chapter 6. A priori knowledge and the explanatory gap 

are concerned (1996: 39). Logical global supervenience is conceptually linked to reductive explanations. As in Levine’s case, reductive explanations require a functional analysis of the phenomena to be explained, that is, by specifying the phenomena that require explanations in causal terms. Explaining how those functions are preformed is equivalent to explaining the phenomena in question (1996: 44). Reductive explanations are linked to the concept of logical supervenience in the following way: “a natural phenomenon is reductively explainable in terms of some low-level properties precisely when it is logically supervenient on those properties,” (1996: 47–48) that is, when no two possible worlds that are identical with respect to the low-level properties differ with respect to the (high-level) phenomena in question. The concepts of logical supervenience and of reductive explanations rely on those of logical possibility and logical necessity. Again, as in Jackson’s case, Chalmers’s two-dimensional semantics is designed with the intention of providing the foundations of a theory that bypasses the traditional concerns about logical necessity, conceptual analysis, and reduction. According to Chalmers, every referring term such as ‘water’ involves two levels of meanings or intensions: primary intension and secondary intension. Primary intensions are functions from possible worlds considered as actual to extensions, while secondary intensions are functions from centered worlds to extensions in worlds considered as counterfactual. As Chalmers notes, the distinction between these two sorts of intensions is close to Kaplan’s (1979, 1989) distinction between character and content with respect to indexical expressions. Yet here the distinction applies to all sorts of referring terms, including names and natural kind terms, and not just indexical terms. Chalmers’s distinction between the two dimensions of meanings seems to be a notational variant of Jackson’s corresponding distinction, but with one important difference. Although Chalmers’s primary intensions indeed correspond to Jackson’s A- intensions and his secondary intensions correspond to Jackson’s C-intensions, in Chalmers’s case, A-intensions are primary intensions and C-intensions are secondary intensions. According to Chalmers, primary intensions have a primary role to serve in explanations: It is the primary intension of a concept that is most central for my purposes: for a concept of a natural phenomenon, it is the primary intension that captures what needs explaining...The primary intension of a concept, unlike the secondary intension, is independent of empirical factors: the intension specifies how reference depends on the way the external world turns out, so it does not itself depend on the way the external world turns out. (1996: 57)

The same idea is also expressed in the following passage: ... when considering questions about explanation, primary intensions are more important than secondary intensions. As noted before, we have only the primary

 Controversies and the Metaphysics of Mind

intension to work with at the start of inquiry, and it is this intension that determines whether or not an explanation is satisfactory. (1996: 69)

Primary intensions allow for a priori knowledge: If a statement is logically possible or necessary according to its primary intension, the possibility or necessity is knowable a priori, at least in principle. Modality is not epistemically inaccessible: the possibility of a statement is a function of the intensions involved and the space of possible worlds, both of which are epistemically accessible in principle, and neither of which is dependent on a posteriori facts in this case. (1996: 68)

Yet, as the following passage verifies, epistemic accessibility and the apriority at stake mean ‘being epistemically accessible in principle’ and ‘being knowable a priori in principle:’ ... if B-properties are logically supervenient on A-properties according to primary intensions, then the implication from A-facts to B-facts will be a priori. So in principle, someone who knows all the A-facts about an actual situation will be able to ascertain the B-facts about the situation from those facts alone, given that they possess the B-concepts in question. This sort of inference may be difficult or impossible in practice, due to the complexity of the situations involved, but it is at least possible in principle. (1996: 70)

A priori knowability (in principle) and epistemic accessibility (in principle) involves conceiving an ideal possibility, a situation that is possible in principle, though not in practice. 5. Block & Stalnaker’s attack Recognizing a position P1 as a controversial relevant alternative to P2 at least implies that arguments introduced by the supporters of P1 are recognized by the supporters of P2 as involving convincing reasons that reveal an intrinsic problem in P2. The reasons introduced by the supporters of P1 are normally associated with what I described in Chapter 2 as the core theses of P1. They support these theses. Recognizing that Jackson and Chalmers’s apriorism and two-dimensionalism generate an internal problem in physicalism may lead to two types of responses. One type of response may endorse apriorism and/or two-dimensionalism while aiming to demonstrate why accepting apriorism and/or two-dimensionalism does not entail the supposed (controversial) epistemic or ontological implications. Some philosophers responded to Levine’s, Jackson’s, and Chalmers’s respective positions qua relevant alternatives in this sense. Block & Stalnaker’s

Chapter 6. A priori knowledge and the explanatory gap 

(1999) arguments do not belong to this type of response. Their arguments aim to undermine Levine’s, Jackson’s, and Chalmers’s respective positions as relevant alternatives to their position. As we shall see in what follows, Chalmers & Jackson’s official response seems to belong to the same type of response. In their official response to Block & Stalnaker’s arguments, they maintain that Block & Stalnaker’s arguments are basically irrelevant to their positions. Their response adds no novel element that significantly changes their earlier position. Nevertheless, I will argue that in “The foundation of two-dimensional semantics”(2006) (FTS), Chalmers also responded to Block & Stalnaker’s attack. His response in FTS differs from his (and Jackson’s) earlier response. In FTS, Chalmers’s response involves recognition of the internal problems in his earlier position revealed by Block & Stalnaker’s arguments.17 In FTS, Chalmers significantly changed his earlier position, a change that was at least partly motivated by the controversy he was involved in. One of Block & Stalnaker’s goals in “Conceptual analysis, dualism, and the explanatory gap” (1999) was to weaken the support provided by the alleged apriority of scientific explanations and by two-dimensionalism to the controversial epistemic mysterianism or ontological dualism. Another goal was to challenge the connection between the apriority of scientific explanations and two-dimensionalism. Block & Stalnaker’s argument can therefore be divided into two parts: The first part aims to weaken the claim that successful explanations in science are a priori in nature, while the second part aims to show that, in contrast to Chalmers ‘s and Jackson’s contentions, the two-dimensional logical apparatus is neutral with respect to apriority. 5.1

The arguments against the epistemic version

Let us begin with a short account of Block & Stalnaker’s argument against Levine’s argument for the explanatory gap. As I noted, Levine claims that our intuitions do not give us access to metaphysical reality. According to Block & Stalnaker, one could also refute the claim that there is an asymmetry between paradigms of successful scientific explanations, such as the water/H2O case and explanations of qualitative conscious states. More specifically, Levine’s contention is that one can reason a priori from, say, microphysical truths about H2O to truths about the macrophysical superficial features of water since it is inconceivable that the former hold while the later do not hold. But, in Block & Stalnaker’s view, this claim is based on fragile intuitions about conceivability. 17. To be sure, in FTS Chalmers responds to criticism that was stated against his theory also by many other philosophers.

 Controversies and the Metaphysics of Mind

If Levine’s account of the explanatory gap as an epistemic gap is correct, it entails that some statements are conceivable without being possible. But what could ‘conceivability’ mean if it has the role that Levine attributes to it? In Block & Stalnaker’s view, the attempt to explicate what ‘conceivability’ means for Levine reveals that his theory faces a dilemma. On the one hand, if we hold the meanings of our terms fixed, (1) below should be explicated as a misdescribed possibility in conformity with Kripke’s corresponding account in Naming and Necessity: 1. It is conceivable that water is not H2O. Sentence (1) could indeed be described as expressing a genuine possibility, say, the possibility that: 2. In Twin Earth, the term ‘water’ refers to XYZ and not to H2O. But (2) does not explicate the content of (1) as it is used by us. That sentence (2) expresses a possibility does not mean that, in our language, it is conceivable that water (as we use this term) is not H2O. Similarly, given that in Twin Earth the microphysical process of boiling differs from the process of H2O boiling, it cannot be maintained as Levine does that it is a priori inconceivable that microphysical truths about the boiling of water hold, while truths about the macro features of water do not hold. For given the meaning of ‘boiling’ in Twin Earth (in Twin Earth, boiling is a different physical process involving XYZ and not H2O), H2O cannot boil. On the other hand, it is indeed true that if the physics and chemistry of our world is held fixed, one could indeed maintain that it is impossible that in a possible world w, which has the same physics and chemistry as our world, the microphysical process that underlies the macrophysical features of boiling in our world would be instantiated in w, while the macrophysical features of the boiling of water would not be instantiated. Yet, it may be conceivable for someone in w who is ignorant of the microphysics of water that H2O cannot boil, though this would be a misdescribed possibility. In particular, as Block & Stalnaker note, if the physics and chemistry of our world is fixed, there seems to be no asymmetry between the case of H2O/boiling and, say, the case of pain/pyramidal cell activity pca. Assume that pain without pca is conceivable (since one cannot deduce the existence of pain from the existence of pca merely by using logic and conceptual truths). And assume further that pain is in fact pca. Any possible world with the same physics and chemistry as our world would be a world in which the existence of pca necessarily entails the existence of pain. Yet the fact that the existence of the former entails the existence of the latter cannot be conceived as an a priori conceptual truth, but rather as an a posteriori necessary truth in the very same sense in which water = H2O is an a posteriori necessity. It is therefore not clear in which sense the conceivability of pain without pca leaves any gap that is unique to the study of consciousness. As Block & Stalnaker conclude:

Chapter 6. A priori knowledge and the explanatory gap 

‘Pain’ as we use the term, obviously applies to pca in this possible world, but what about ‘pain’, as used by the people in this counterfactual world? Can we consistently say about such a possible world that the people in it, who are physically just like us, refer with ‘pain’ to something other than what we call ‘pain’? It is not clear that we can, and if we cannot, then this way of thinking about conceivability without possibility does not show that pca without pain is conceivable, and so does not show any asymmetry. (1999: 7)

Levine’s account of the explanatory gap invokes conceivability claims that do not express genuine possibilities. His account of conceivability without possibility appears to be similar Kripke’s account. But, as Block & Stalnaker note, the main idea that underlies Kripkean cases of conceivability without possibility is that the meaning and reference of the terms of our language “depend on empirical facts, facts that we might be ignorant or mistaken about.” (1999: 6) Given this sort of dependence, there seems to be no room for cases of conceivable impossibilities that allow the asymmetry between successful explanations in science and the explanation of consciousness – the asymmetry that underlies the idea of an explanatory gap. It can be easily seen that the type of argument that Block & Stalnaker use is that of type A discussed in Chapter 1. It points out difficulties in the position attacked without directly involving the core theses of the attacker’s position. 5.2

The apriority of reductive explanations and uniqueness

The type of criticism used by Block & Stalnaker against Levine obviously does not apply to Chalmers’s and Jackson’s respective accounts. In their case, conceivability entails genuine possibility. As we saw above, Jackson’s and Chalmers’s twodimensional semantics does not allow for the existence of sui-generis a posteriori impossibilities. But their two-dimensional semantics also does not allow for the existence of a posteriori necessary truths that are not knowable in principle. This issue seems to be central to Block & Stalnaker’s arguments. They attempt to undermine the supposition that if the conjunction of all the physical facts does not a priori entail the conjunction of all the psychological facts, then physicalism is a false metaphysical doctrine. Must reductive explanations involve the a priori conceptual analysis defended by Chalmers and Jackson? In Block & Stalnaker’s view, Jackson’s and Chalmers’s defense of a priori conceptual analysis is vulnerable to two types of criticism, the first of which is the following: Jackson and Chalmers seem to base their arguments on some cases of successful explanations, but the features of these examples are not exhibited by all cases of successful scientific explanations. For example, functional analysis is in Levine’s and Chalmers’s view an essential step in a reductive

 Controversies and the Metaphysics of Mind

explanation. Yet, not all cases of successful scientific explanations involve functional analysis. The explanatory gap regarding life was not closed by first providing a functional analysis of life.18 The second and more significant argument against the apriority of scientific reduction addresses what in Block & Stalnaker’s view is an irremovable gap in Jackson’s and Chalmers’s accounts of conceptual analysis. The arguments that establish the a priori entailment of facts expressed in one vocabulary (folk vocabulary) facts that are expressed in another vocabulary (scientific vocabulary) involve implicit claims to uniqueness that will be specified below. Although these uniqueness claims are required by Chalmers’s and Jackson’s a priori entailments, they nevertheless, resist a priori analysis. Block & Stalnaker’s example is in fact Jackson’s main example in From Metaphysics to Ethics. Block & Stalnaker present Jackson’s argument as follows: a. The earth is covered 60 percent by H2O. b. H2O = the waterish stuff. c. Water = the waterish stuff. d. The earth is covered 60 percent by water. (1999: 12) The term ‘waterish stuff’ is an abbreviation of the description ‘stuff which actually falls from the sky, fills the oceans, is odorless and colorless, is essential for life, is called by expert water etc...’ The term ‘actual’ used in the above description rigidifies this description. The first two premises of the argument express microphysical a posteriori facts. The third premise supposedly expresses a conceptual truth. Given that (c) is a conceptual truth, (a) and (b) a priori entail (d) (1999: 17). Yet, as Block & Stalnaker explain, (b) in the above argument is in fact the following: b. H2O = the waterish stuff around here. Premise (c) in the above argument is in fact the following: c. Water = the (unique) waterish stuff around here. (1999: 17) The question that Block & Stalnaker raise is whether use of ‘around here’ in (b) allows one to view (b) as referring to a microphysical fact, and whether the use of ‘around here’ and ‘unique’ in (c) allows one to view (c) as expressing a conceptual truth. Block & Stalnaker’s argue that (b) does not express microphysical truth and (c) a conceptual truth. As we shall see below, their argument involves the controversial intuition that Chalmers and Jackson share. That is, it involves the supposition that is similar to the supposition according to which the epiphenomenal non-physical character of conscious experiences is a genuine possibility. 18. As Block & Stalnaker note, life cannot be analyzed functionally. It cannot be identified with the fulfillment of a set of functions, even if functional analysis in terms of the causal roles was part of the explanation of life.

Chapter 6. A priori knowledge and the explanatory gap 

In other words, Block & Stalnaker’s argument is an argument of type (B) discussed in Chapter 2. The conclusion of the argument provides support for their position. But their argument makes use of a contention that is similar to the contention that belongs to the core set of Chalmers’s and Jackson’s respective positions. The same types of reasons that are used in the argument for epiphenomenalism are invoked in order to undermine the purely a priori character of scientific reductive explanations. Block & Stalnaker present the problem as follows: We doubt that there is a way to spell out ‘waterish’ so that (c) is a conceptual truth, and we doubt further that any remotely plausible analysis would get to first base when substituted in (b) as a truth of microphysics; but even if we set this issue aside, there is an additional problem with the argument: Even if ‘water’ can be given such a definition, the uniqueness assumption in the description throws doubt on the claim that premise (b) of the argument is a microphysical fact. (1999: 17)

If it is a conceptual truth that only physical stuff is ‘waterish stuff,’ then the presence of ‘waterish stuff around here’ is a microphysical fact. But ruling out the possibility that waterish stuff could be non-physical merely on conceptual grounds is an arbitrary stipulation. Moreover, this is precisely the type of an arbitrary stipulation ruled out by Jackson’s and Chalmers’s respective accounts. Let us remember that they both assume that non-physical properties are a genuine possibility. Hence that ‘the waterish stuff around here’ in (c) stands for physical stuff (as required by the analysis of water in microphysical terms) involves a claim to uniqueness, that is, that ‘water’ refers only to physical waterish stuff. Block & Stalnaker maintain that this uniqueness assumption cannot be a priori. Block & Stalnaker’s objection (1999: 19) is directly related to Jackson’s ‘nothing else’ condition that in his view physicalism requires.19 Their argument invokes this condition in the present context. As noted above, the physicalist cannot be merely satisfied with the claim that every possible world that is physically identical to our world is psychologically identical to our world. She also requires the claim that our world contains nothing but the physical facts. Physicalism is clearly false if, in addition to all the physical facts, our world also contains nonphysical facts. But is this restriction explicable as a microphysical fact? On the face of it, given the features of the uniqueness assumption noted above, it cannot belong to this type of facts. It should be stressed at this point that Block & Stalnaker do not maintain that explanations of facts described by means of folk vocabulary in terms of facts described by means of scientific vocabularies are not tenable. Their claim is rather 19. See Jackson 1998: 12–13.

 Controversies and the Metaphysics of Mind

that such explanations are not based merely on a priori reasoning and need not include a conceptual analysis of the explained phenomena. The gap involved in the a priori type of explanations could be filled by a posteriori identity statements.20 Block & Stalnaker’s contention is that identity requires no explanation. Nevertheless, identity statements could be used in explanations. Assume for a moment that a perfect correlation between pain and pyramidal cell activity is detected. According to Block & Stalnaker, one could explain the correlation by supposing that pain is identical to pca.21 Is this type of account plausible? Clearly, an explanation of the correlation between pain and pca that invokes the identity of pain and pca is defeasible. This, however, should not be conceived as a difficulty where ordinary human knowledge is concerned. Defeasibility is a feature of scientific empirical knowledge. One is apt to change one’s beliefs on the basis of new evidence. Yet, the implications of epistemic defeasibility of such identity statements are puzzling with regard to ideal epistemic situations, the type of cases used in Jackson’s and Chalmers’s respective arguments. Mary, the omniscient scientist, who already has all the relevant physical information about the perception of red objects, cannot change her beliefs regarding physical reality in light of new physical information. If she must also use a posteriori claims about the identity of two types of phenomena in order to explain the correlation between them (identity claims that, for all she knows, might be mistaken) then science and human knowledge may be threatened by an incurable type of skepticism. Since Mary already knows all the physical facts about the perception of red objects, the possibility that these identity statements could be mistaken cannot be eliminated by further inquiry. There therefore seems to be an unbridgeable gap between metaphysical and empirical reality, and our epistemic capacities. This seems to be precisely the concern that underlies Chalmers’s and Jackson’s intuitions. I will return to this point later. 5.3

The explanatory gap and two-dimensional semantics

As Block & Stalnaker note, Chalmers’s and Jackson’s two-dimensional accounts of meaning seem to provide them with an escape route from the above line of criticism. Here is how the claim that water covers 60% of the globe could be analyzed a priori as a microphysical truth: a. 60 percent of the globe is covered by H2O. b. H2O = the satisfier of X (the primary intension of water).

20. See Block & Stalnaker 1999: 43–44. 21. For a similar position see Papineau 1998.

Chapter 6. A priori knowledge and the explanatory gap 

c. Water = the satisfier of X. d. 60 percent of the globe is covered with water.22 It should be recalled that primary intensions are functions from centered worlds considered as actual to extensions. As Block & Stalnaker note, given that primary intensions are functions, the question of uniqueness does not arise. Yet, the main question raised by Block & Stalnaker is whether the two-dimensional framework could be properly understood as providing an analysis of meaning rather than a representation of meaning, whereas meaning itself is presupposed by (and not analyzed by) two-dimensionalism. Consider (b) above. Is (b) a microphysical truth? According to Block & Stalnaker, the claim that (b) is a microphysical truth begs the question: What (b) says is that the primary intension of ‘water’ maps the actual world onto exactly one item, H2O. But we can conclude that this is microphysical fact only if it is assumed that microphysical facts determine or include all the facts. The primary intension of ‘water’ is a function that takes WEarth to water, WTwin Earth to twin water etc. But we can’t, without begging the question, take for granted that the microphysical description of WEarth describes only WEarth. Suppose there are two microphysically indiscernible possible worlds, WEarth and WSuper Twin Earth. Suppose further that there are primary intensions that take WEarth to H2O, but WSuper Twin Earth to something else. (Primary intensions are just functions, and given any difference in inputs, there will be some functions that yield a difference in outputs.) If this is true for the primary intension of ‘water’, then (b) will be a fact, but not a microphysical fact, or even supervenient on the microphysical facts. Unless we assume that microphysical facts determine all the facts, or at least that the value of the primary intension for ‘water’ depends only on microphysical facts in the worlds that are the arguments to the function, the argument won’t work. (1999: 37–38)

The abstract two-dimensional account of meaning in terms of a particular kind of function renders the assumption that (b) is a microphysical fact implausible. If one assumes that microphysical facts determine all the facts, including facts about meaning, it seems that one can respond to this objection. But this response begs the question. For the abstract two-dimensional account of meaning in terms of a certain sort of function does not entail the statement that facts about meaning are determined by means of microphysical facts. Another type of response could now be presented. Chalmers, for example, could argue that primary intensions should be construed on the basis of our conceptual capacities. But what are these concepts over and above the account allowed by two-dimensionalism? One could have responded, for example, that the two-dimensional intensions are fixed by “the dispositions that the speaker has in 22. See Block & Stalnaker 1999: 37.

 Controversies and the Metaphysics of Mind

virtue of understanding, or being a competent user of, the term in question.” (1999: 39–40) Yet, as Block & Stalnaker point out, this response requires an account of how intensions are determined which is different from the one provided by Jackson’s and Chalmers’s two-dimensionalism. This response smuggles in a functionalist theory of concept possession. In Block & Stalnaker’s view, this move is unjustifiable: If the functionalist theory of concepts presupposes that there are analytic inferences that provide a purely conceptual component of the knowledge of competent users, that presupposition would have to be justified. If it does not suppose this, we will have to deal with the familiar holism problem of nonanalytic functionalism. How, and in what ways, can people differ in roles for a word while still expressing the same concept by that word?. (1999: 40–41)

This seems to be a pressing question that Chalmers’s and Jackson’s theories have to face. Chalmers, for example, indeed assumes that primary intensions should be explicated in terms of the concepts that we ourselves possess. Yet what precisely is the nature of the concept that is affected by the differences between possible worlds? As Block & Stalnaker note, Chalmers seem to be caught in a dilemma here. If what is transferred from world to world... ... is the meaning in the ordinary sense (the ‘wide’ meaning, or secondary intension), then it does not pick out twin water on Twin Earth...So it must be the narrow meaning – the purely conceptual content – that is retained. But the primary intension (or perhaps the two-dimensional intension), is supposed to be the explication of narrow, or purely conceptual content. So this answer is of no help in explaining what the primary intension of a concept is. (1999: 42)

Primary intensions have different extensions in distinct worlds considered as actual. Therefore, they cannot be identified with meaning in the (wide) ordinary sense. If ‘the purely conceptual content’ can be explicated by means of primary intensions, one must explicate the core of meaning that is retained in different possible worlds without invoking narrow content. For narrow content is supposed to be explicated by means of primary intensions. But this appears to be what Chalmers’s account fails to do. Block & Stalnaker’s criticism seems to revolve around the concept of meaning. Functional analysis of natural kind terms is not equivalent to an analysis of their meanings. The main intuition that underlies their objections is that the meaning of a natural kind term contextually depends on empirical facts that are part of the world of the people that use the term. Two-dimensionalism fails as an analysis of meaning. It presupposes meaning rather than analyzes it. Even if two-dimensionalism can in principle provide an account of apriority, it cannot serve as the foundation for the claim that physicalism requires a priory reductive explanations.

Chapter 6. A priori knowledge and the explanatory gap 

6. Chalmers & Jackson’s response Chalmers & Jackson’s main official response to Block & Stalnaker is presented in their paper “Conceptual analysis and reductive explanation.” (CARE) (2001)23 The intention that underlies their detailed response in this paper is to reveal the irrelevance of Block & Stalnaker’s arguments to their position. Chalmers & Jackson’s response is presented in three steps. First, they reiterate their previous (separate) accounts of a priori entailment. Then they provide reasons that seem to refute the claim that a priori knowledge must involve explicit conceptual analysis. Finally, they provide a critical examination of the alternative account of scientific explanation defended by Block & Stalnaker, the account that essentially involves a posteriori identity statements. One interesting feature that characterizes CARE is the separation between questions that concern apriority and questions that concern two-dimensionalism. In CARE, the reasons that validate a priori entailments do not directly involve the two-dimensional framework qua the rationalizing basis of a priori knowledge. This seems to conflict with the unified account of these two topics in their earlier work. Presumably, this move is related to Block & Stalnaker’s corresponding criticism. As we shall see below, the division between a priority and the two-dimensional framework is partly justifiable. Nevertheless, it destabilized the rational grounds of Chalmers’s and Jackson’s respective theories. Severing apriority from the two-dimensional framework undermined the rational basis of apriority. Moreover, disconnecting apriority from the two-dimensionalism left Chalmers and Jackson without a satisfactory response to Block & Stalnaker’s criticism with regard to the relevance of questions that concern apriority and scientific reduction to problems related to consciousness. As we shall see, an attempt to remedy these flaws is found in Chalmers’s FTS. Chalmers & Jackson first spell out the sense in which a macrophysical truth M is entailed by microphysical truths. According to their usage, p entails q if the material conditional ‘p → q’ is true. In their usage, ‘implication’ means a priori entailment, that is, an entailment the truth of which could be known independently of experience. Let M be... ... a typical macroscopic truth concerning natural phenomena, such as water or life, outside the psychological, social, and evaluative domains. Some such truths include: ‘water boils at 100 degrees Centigrade’, ‘water is H2O’, ‘water covers much of this planet’, ‘life propagates through the replication of DNA’, ‘there are many living beings’, and so on. (2001: 316-317) 23. There are other places where they each respond to Block & Stalnaker. See in particular Jackson 2003. I will not address Jackson’s paper directly in the present context.

 Controversies and the Metaphysics of Mind

Let P be the conjunction of microphysical truths about the world. Let T be the statement asserting that our world is a minimal world satisfying P (ibid.). Let I be the relevant indexical information, and let Q be the conjunction of all phenomenal truths. Then the claim that a macrophysical truth is implied by a typical microphysical truth is represented by the schema PTIQ → M. Implication does not require explicit analysis of concepts. This appears to be central to Chalmers & Jackson’s response. In their view, Block & Stalnaker’s account assumes that a priority involves explicit analysis. Their main counterexample for the supposition that apriority involves explicit analysis is the Gettier examples. We do not possess an analysis of the concept ‘knowledge.’ Yet, we seem to be able to reason a priori from the Gettier’s examples to the conclusion that these types of cases are not cases of knowledge. The Gettier examples are philosophical examples that involve merely conceptual truths, not empirical truths. One could therefore argue that it is not clear if the Gettier examples support the notion of apriority required for the a priori reductive explanations of empirical phenomena. Chalmers & Jackson do not provide any empirical example that verifies their claim. Instead, they attempt to convince their readers that the apriority of the entailment PTIQ→M is compatible with the fact that PTIQ involves empirical information. According to Chalmers & Jackson: It might be objected that if empirical considerations can play a role in affecting what is a priori involving a concept (as in the first alternative above regarding simplicity, and also on the version of the ‘rheumatism’ case involving terminological stipulation), then the notion of apriority is being watered down. But this seems wrong: apriority is a matter of non-empirical justification. Concept acquisition is usually empirically driven, and conceptual drift can occur in response to empirical factors, but neither of these is any bar to the apriority of resulting claims involving the concepts. (2001: 349)

Nevertheless: It is no part of our position that a posteriori methods are never used in determining our concepts’ extensions. They are used all the time: in particular, it is by using a posteriori methods that we form empirical hypotheses about how the world is. The information in PQTI itself is knowable only a posteriori; and simplicity considerations play a central role in our coming to know aspects of this information. But nothing here bears on the claim that given the information about PQTI ... the application of our concepts is determined a priori. (2001: 343)

And also: There is no question that empirical information can play a causal role in acquiring this knowledge. Empirical knowledge often plays a causal role in the

Chapter 6. A priori knowledge and the explanatory gap 

acquisition of concepts with certain a priori connections, and it sometimes plays a role in triggering changes in the a priori connections associated with a term, as in the cases above. There is also no question that E could play a mediating role in our knowledge of an application conditional. That is, E might itself be implied by PQTI, and one could then straightforwardly use E in combination with PQTI to deduce M. But neither of these possibilities entails that E plays an essential role in justifying knowledge of the relevant conditionals. We suspect that the cases in which Block and Stalnaker think that empirical information has a justificatory role are in fact cases where the information has a causal or mediating role. (2001: 346)

The debated issue is, therefore, whether the empirical information expressed in PTIQ has a justificatory role, qua empirical, with respect to the entailment PTIQ → M, or merely a causal role. Chalmers & Jackson’s answer is that empirical information merely serves a causal role in PTIQ → M, not a justificatory role. Yet, it should be noted that by involving all the microphysical truths, all the relevant indexical truths and all phenomenal truths, knowledge of PTIQ is an idealized epistemic situation that abstracts from cognitive human limitations. It might be objected here that the above type of idealized knowledge should not be used as the basis for examining the role of empirical information in scientific explanations. But this objection is not available to Block & Stalnaker. In their view, the meaning of a natural kind term contextually depends on empirical facts that belong to the same world as the people using the term without entailing any restriction on the ‘all the relevant information’ involved in Chalmers & Jackson’s PTIQ → M. Stated differently, if the differences regarding the status of PTIQ → M result from the fact that one party bases its position on ideal cases of knowledge while the other bases it on real cases of knowledge, the differences are eliminable. But conceiving the conflict along these lines involves conceiving identity statements such as ‘pain = pca’ as having merely a heuristic explanatory role that is eliminable in the idealized epistemic situations. And it is clear that the explanatory role assigned by Block & Stalnaker to a posteriori identity statements is not meant to be heuristic in the above sense. Indeed, this seems to be significant to Chalmers & Jackson’s response to Block & Stalnaker. As Chalmers & Jackson note, the explanatory role attributed by Block & Stalnaker to a posteriori identity statements may mean that they have the epistemic status of primitive (unexplained) psychophysical ‘bridge laws.’ They, therefore, involve similar difficulties to those raised by primitive (unexplained) bridge laws: Ontologically, these identities may differ from laws. But epistemically, they are just like laws. They are epistemically primitive psychophysical “bridging” principles that are not themselves explained, but that combine with physical truths

 Controversies and the Metaphysics of Mind

to explain phenomenal truths. An explanation of the phenomenal will have two epistemically irreducible components: a physical component and a psychophysical component. By calling the bridging principles identities rather than laws, this view may preserve the ontological structure of materialism. But the explanatory structure of this materialist view is just like the explanatory structure of property dualism. (2001: 353–4)

This similarity between a posteriori identity statements and primitive (unexplained) psychophysical ‘bridge laws’ raises the question of whether identity statements do not require explanations, as Block & Stalnaker suggest. Indeed, according to Chalmers & Jackson, Block & Stalnaker seem to conflate two distinct issues concerning the role of identity statements: It is sometimes held that “identities do not need to be explained” (for example, Papineau 1993). Block and Stalnaker say something similar (“Identities don’t have explanations”). But this seems to conflate ontological and epistemological matters. Identities are ontologically primitive, but they are not epistemically primitive. Identities are typically implied by underlying truths that do not involve identities. The identity between genes and DNA, or between water and H2O, is implied by the underlying truths in PQTI, for example. Once a subject knows all the truths about DNA and its role in reproduction and development, for example, the subject will be in a position to deduce that genes are DNA. So this identity is not epistemically primitive. (2001: 354)

Say that identity statements are not epistemically primitive. What, however, is the alternative type of explanation suggested by Chalmers & Jackson to viewing identity statements as epistemically primitive? The surprising answer is that instead of identifying say pain and pca for the purpose of explaining the correlation between pain and pca as Block & Stalnaker suggest, the alternative offered by Chalmers & Jackson is to accept implications such as (1) below as primitive a priori entailment: 1. ‘PTQI therefore water boils. But the term ‘a priori’ is misleading in this context. It seems to suggest that one has a unique type of justification at one’s disposal, when in fact one has no justification for accepting the conditional as true over and above one’s intuition that its falsity is inconceivable. The technical conceptual tools of two-dimensional semantics could perhaps have been used to respond to the above objection. But Chalmers & Jackson do not develop such a response in this context. Rather, they seem to accept Block & Stalnaker’s objections regarding the role of the two-dimensional framework regarding a priori knowledge: We can also agree with Block and Stalnaker that the mere existence of the twodimensional semantic framework does not imply that the a priori entailments

Chapter 6. A priori knowledge and the explanatory gap 

in question exist. As Block and Stalnaker say, the two-dimensional framework provides a good way to capture a priori connections when they exist, but the framework alone does not demonstrate specific a priori connections. But the two-dimensional framework plays no essential role in our arguments for a priori entailments. It played no role in the arguments given here. In the arguments given in Chalmers (1996) and Jackson (1998), it plays only a clarifying role, in removing certain confusions that may arise from the presence of a posteriori necessary connections, and in providing a convenient shorthand for discussing the patterns by which a concept applies to the world. The core of the argument for a priori entailments proceeds independently of these semantic details, just as it does in the Gettier case. (2001: 337–8)

In CARE, Chalmers & Jackson’s position is that the apriority of entailments could be established independently of considerations that involve the two-dimensional framework. I hope I have by now succeeded in convincing the reader why this supposition is problematic. The controversy between these two eminent teams in fact reveals another antinomy. On the one hand, there are convincing reasons for the view that an explanation in science involves a posteriori identity statements and, therefore, cannot be based on a priori entailments. But identities are not epistemically primitive and therefore require explanations. This was Chalmers & Jackson’s main point. However, a priory entailments that do not involve explicit analysis are no less epistemically arbitrary than the brute a posteriori identity statements they suppose to replace. They require explanations no less than the brute a posteriori identity statements. Therefore, the fact that Chalmers & Jackson disassociated apriority from the two-dimensional framework had damaging results for their philosophical endeavor. 7. Chalmers’s two-dimensionalism refined An implicit attempt to remedy the above shortcoming can be found in Chalmers’s FTS. One of Chalmers’s intentions in FTS is to reinstate the connection between apriority and two-dimensionalism by revising his version of two-dimensionalism. I will not address the question of whether Chalmers’s revisions are open to further criticism. Rather, my concern here will be to clarify the sense in which his reexamination of his earlier views in light his of opponents’ objections exemplifies the way in which progress in metaphysics is facilitated by controversies. Chalmers begins his paper as follows: Why is two-dimensional semantics important? One can think of it as the most recent act in a drama involving the three central concepts of philosophy: meaning, reason and modality. First, Kant linked reason and modality, by suggesting

 Controversies and the Metaphysics of Mind

that what is necessary is knowable a priori, and vice versa. Second, Frege linked reason and meaning, by proposing an aspect of meaning (sense) that is constitutively tied to cognitive significance. Third, Carnap linked meaning and modality, by proposing an aspect of meaning (intension) that is constitutively tied to possibility and necessity. (2006: 55)

But Several years later, Kripke severed the Kantian link between apriority and necessity, thus severing the link between reason and modality. Carnap’s link between meaning and modality was left intact, but it no longer grounded a Fregean link between meaning and reason. (2006: 55)

The task of two-dimensional semantics is “to restore the golden triangle” that consists of meaning, reason, and modality. In FTS, Chalmers reexamines the whole project of two-dimensional semantics with the intention of distinguishing between two types of two-dimensionalism, namely, the contextual type and the epistemic type. The question that motivates this distinction is whether two-dimensional semantics could restore the golden triangle. As he clarifies, not all types of two-dimensionalism allow this. Although the contextual type does not, the epistemic type does seem to be able to restore the golden triangle. As Chalmers notes, this novel account of two-dimensionalism constitutes a departure from his earlier position. Autobiographically: I think that primary intensions as I conceive them (both in Chalmers (1995) and in Chalmers (1996)) were much more like epistemic intensions than like contextual intensions. But the distinction between contextual and epistemic understandings was not sufficiently clear in my mind at the time of writing, and is certainly not clear on the page. (The current paper is in part mea culpa). Certainly, one must interpret primary intensions as epistemic intensions to make sense of the applications of the two-dimensional framework in these works. (For example, the main conceivability-possibility argument in Chalmers (1996) turns on a version of the thesis of Metaphysical Plenitude outlined earlier.) If one does so, I think the resulting arguments are sound. (2006: 129)

As Chalmers acknowledges, the absence of a distinction between the two types of two-dimensional semantics damaged the main arguments of his earlier position. The earlier position in fact confused the epistemic version and the contextual version.24 In Chalmers’s view, the new distinction (and its consequences) remedies the shortcomings of the earlier position. The question that needs to be addressed concerns the motives that underlie this interesting philosophical transformation. 24. Compare Chalmers 2006: 127.

Chapter 6. A priori knowledge and the explanatory gap 

I will not examine Chalmers’s rich and insightful paper and his related works in detail here. Rather, my concern will be with the sense in which FTS could be grasped as containing a response to the debate with Block & Stalnaker. In particular, I will attempt to examine whether the novel ideas included in FTS could be interpreted as attempts to resolve the antinomy revealed by the exchange between Chalmers & Jackson and Block & Stalnaker. Let us begin by noting that in contrast to Chalmers & Jackson’s response to Block & Stalnaker, in FTS, Chalmers does not separate the questions concerning apriority from the questions that concern two-dimensionalism but rather unifies the two issues as he did in The Conscious Mind. In the search for a restoration of the lost golden triangle, Chalmers formulates his question as follows: “Is there an interpretation of the two-dimensional framework that yields constitutive connections between meaning, reason, and modality?.” (2006: 64) This question is equivalent to the question of whether a position satisfying the ‘core thesis’ is feasible: Core thesis: For any sentence S, S is a priori iff S has a necessary 1-intension.” (2006: 64) According to Chalmers, the central neo-Fregean thesis is: Neo-Fregean thesis: Two expressions ‘A’ and ‘B’ have the same intension iff ‘A≡B’ is a priori. (2006: 58) The core thesis immediately entails a version of the neo-Fregean thesis: Neo-Fregean thesis (2d Version): Two expressions ‘A’ and ‘B’ have the same 1-intension iff ‘A≡B’ is a priori. (2006: 64) The contextual understanding of two-dimensional semantics is based on the idea of a contextual intension. The varied versions of the contextual understanding of two-dimensionalism are defined in terms of a variety of notions of contextual intensions. A contextual intension of an expression type is a function from centered worlds to extensions (2006: 66). The contexts of contextual intensions are centered worlds that include the utterances of tokens of sentences (relative to a type of which it is a token) (2006: 66). As Chalmers notes, first-dimensional intensions are often understood as some sort of contextual intensions. The contextual intension that is of particular interest to the present topic is the semantic contextual intension. A semantic contextual intension of an expression token T “...is defined at centered worlds with a token of T’s linguistic type at the center. It maps such a world to the extension of the relevant token in that world.” (2006: 68) The behavior of a semantic contextual intension depends on the choice of semantic values. As Chalmers notes, one could “stipulate that the relevant semantic value is its standing meaning: roughly, the aspect of meaning that is common to all tokes of the expression linguistic type.” (2006: 69) Alternatively, one may stipulate “that the relevant semantic value of an expression token is its

 Controversies and the Metaphysics of Mind

Fregean of descriptive content, corresponding roughly to the expression’s cognitive significance for the subject.” (ibid.) Both versions seem to be close enough to the way Block & Stalnaker interpreted Chalmers’s earlier version of two-dimensional semantics. As Chalmers himself suggested, this interpretation is supported by his earlier texts. Indeed, as Chalmers maintains in the present context: One can argue that when a statement is a priori, any possible statement with the same descriptive content will be a priori and so will be true, so that the expression’s semantic contextual intension will be necessary, as the Core Thesis requires. Correspondingly, one might suggest that when a statement is not a priori, then there will be possible statements with the same descriptive content that are false, so that the statement’s semantic contextual intension will not be necessary, as the Core Thesis requires. (2006: 69)

Yet, although one could interpret the earlier Chalmers as a supporter of the semantic contextual intension version of two-dimensionalism, in FTS, Chalmers views this version as inadequate for fulfilling his theoretical intentions: The Fregean two-dimensionalist, as sketched previously, intends to use the twodimensional framework to ground an aspect of meaning that is constitutively tied to meaning. But semantic contextual intensions as defined here presuppose such a Fregean semantic value, and so cannot independently ground such an account. If this is the best a two-dimensionalist can do, then if someone is independently doubtful about a Fregean aspect of meaning, two-dimensionalism cannot help. At best, two-dimensionalism will be a helpful tool in analyzing such a notion of meaning, given an independent grounding for the notion. (2006: 69–70)

As Chalmers notes in a footnote added to this passage (ibid.), this was precisely one of Block & Stalnaker’s objections regarding the relevance of Chalmers’s and Jackson’s earlier accounts of two-dimensionalism to the problem of the explanatory gap. In FTS, Chalmers replaces the contextual understanding of two-dimensional semantics with an epistemic understanding, which he believes can restore the golden triangle of reason, modality, and meaning. The epistemic version is formulated by Chalmers as a response to the above objection. Chalmers’s account of the epistemic understanding contains many novel elements by means of which the difficulties noted above are resolved. The three key ideas of the epistemic version are that of an epistemic space, the corresponding epistemic possibilities, and the idea of scrutability: “once we know how the world has turned out, or once we know which epistemic possibility is actual, we are in a position to determine the extensions of our expressions.” (2006: 75) Together these ideas suggest an epistemic intension. Here is a brief account of these ideas: epistemic space is depicted by scenarios. A scenario corresponds to “a maximally specific epistemically possible hypothesis,

Chapter 6. A priori knowledge and the explanatory gap 

or (for short) a maximal hypothesis: a hypothesis such that if one knew that it were true, one would be in a position to know any truth by reasoning alone.” (2006: 81) Correspondingly, the link between scenarios and epistemic possibilities is defined as follows: Plentitude Principle: For all S, S is epistemically possible if and only if there is a scenario that verifies S. (2006: 81)

One of Chalmers’s main intentions is to clarify why the epistemic version does not merely satisfy the Plentitude Principle, but also the thesis of Metaphysical Plentitude. The latter is defined as follows: Metaphysical Plentitude: For all S, if S is epistemically possible, there is a centered metaphysically possible world that verifies S. (2006: 82)

To these two plentitude theses, he adds a third thesis: Epistemic Plentitude: For all S, if S is epistemically possible, then some epistemically complete sentence of L implies S. (2006: 84)

D is an epistemically complete sentence of L when “(i) D is epistemically possible, and (ii) there is no sentence S of L such that both D&S and D&~S are epistemically possible.”(ibid.) It can easily be seen that the above PTIQ is an epistemically complete sentence. Also, if epistemic plentitude entails metaphysical plentitude, this means that all epistemic possibilities have metaphysical counterparts. A position that can establish the link between these principles will therefore be able to restore not only the connection between meaning, modality, and a priori knowledge but also the rational grounds for the knowledge argument and the zombie argument and, therefore, also the metaphysical interpretation of the explanatory gap. How does Chalmers reinstate these connections without being vulnerable to Block & Stalnaker’s objections? A key notion in this context is the idea of a canonical description of a scenario. A canonical description contains semantically neutral expressions. Such expressions are expressions that do not have “twinearthable” tokens.25 Stated differently, a neutral expression is an expression the extension of which in counterfactual worlds does not depend on how the actual world turns out (epistemically). A statement is in a canonical form “when it has the form D& ‘I am D1 & now is D2, where D, D1 and D2 are semantically neutral and D1 and D2 are identifying predicates relative to the information D.” (2006: 87) Clearly, a theory of meaning that is based on the notion of epistemic possibility and on canonical descriptions that involve only semantically neutral expressions 25. A twin-earthable expression token is an expression token that has a twin in Twin Earth with a different secondary intension.

 Controversies and the Metaphysics of Mind

seems to meet Block & Stalnaker’s objection. According to this view, meaning is not contextually dependent on the world that includes the uttered token. For any uttered token T, there is a complete sentence D that has a canonical form that implies T. Meanings, and therefore, contents of thought, are not contextually dependent on the world they belong to. In principle, we have epistemic access to all possibilities “ from nowhere,” so to speak. The account of meaning, reason, modality, and apriority provided by Chalmers abstracts from all limitations related to human cognitive and epistemic capacities. This could well serve as point of criticism, but as I suggested above, it is irrelevant to the present debate. For both sides seem to share the view that human rational epistemic capacities should be accessed on the basis of what is in principle knowable, and not on the basis of the actual capacities that we possess. 8. Conclusion The controversy on the nature of consciousness and the nature of scientific explanations does not end with Chalmers’s response. As I noted above, the matrix of controversies to which this response belongs has other ramifications. I will not examine them in the present context. Although Chalmers’s contribution is probably not the final stage in the chain of controversies examined in Part Two of this book, it at least clarifies the sources of the conflict that it deals with more sharply than before. Is there a gap that separates epistemic possibility and metaphysical possibility? Do we have rational access to reality that does not depend on the contexts from which we speak and think? Chalmers’s arguments clarify the hazard of irrationalism faced by those that emphasize the gap between epistemic possibility and metaphysical possibility and by those that endorse the contextual version of two-dimensionalism. Being part of the rationalist tradition, he denies the gap and the context-dependence of the foundations of thought and meaning. With Chalmers’s response, the controversy on the nature of consciousness has, however, reached its peak in at least one main respect. The connection between the factual issues dealt with by metaphysicians and their connection to logical and semantical issues is probably most evident in Chalmers’s contribution to the debate. The same type of connection also characterizes the metaphysical discourse on the nature of personal identity, which will be examined in Part Three of this book.

part three

Personal identity and revisionary metaphysics

chapter 7

Personal identity, self-consciousness, and bodily identity 1. Introduction Since early antiquity, philosophers have attempted to decipher the enigma of personhood and personal identity. In Western tradition, debates concerning the nature of personal identity and self-consciousness evolved with the progress of early modern philosophy. Locke’s theory played a key role in these debates. His account of personhood emphasized the significance of self-consciousness in the concept of a human person and disconnected the identity of persons from that of all types of substances – mental substances and physical substances. Locke’s position, which conflicted with current metaphysical theories, was extensively discussed and disputed by 17th and 18th century philosophers. Kant’s attempt to end this chain of debates is probably the culmination of the early modern discourse on this topic. But debates on the nature of human persons and their identities did not end with Kant’s critical philosophy. The same perplexing relations between personhood and self-consciousness continued to be a part of Western philosophy in the 19th and 20th centuries. They were particularly significant to the development of German idealism, and to the phenomenological tradition that emerged in the early years of the 20th century upon the publication of Husserl’s Logical Investigations. The first analytic philosophers paid little attention to the subject of personal identity and to the debates regarding the nature of personal identity. This approach has gradually changed since the early 1950s. Scholarly attention was increasingly devoted to the question of personal identity in the analytic tradition, which was the subject of new controversies. As I hope to show in Part Three of this book, these notable examples illustrate how fruitful controversies have been in the evolution of metaphysics. At the beginning of “Personal identity and individuation” (1956–7)1 (PII), which constitutes a landmark in the new debates on this topic, Bernard Williams claimed that:

1.

Reprinted in Williams 1973.

 Controversies and the Metaphysics of Mind

There is a special problem about personal identity for two reasons. The first is self-consciousness – the fact that there seems to be a peculiar sense in which a man is conscious of his own identity...The second reason is that a question of personal identity is evidently not answered merely by deciding the identity of a certain physical body. (1973: 1)

In this paper, Williams defends a version of a theory which views bodily identity as necessary for personal identity. Although he does note the force and the reasonableness of the intuitions on which the opposite positions are based, his defense of his version of bodily identity theory aimed to reveal the illusions to which these intuitions lead due to their extensive use of dialectical means. Interestingly, however, the only alternative which Williams addresses as a relevant threat to his version of materialism is Locke’s position. Williams’s return to the debates of the 17th and 18th century is both surprising and significant. The Lockean position was merely the first position in the Western tradition to characterize persons as constituting a unique type of entities, which evades the objective descriptions by which other objects are characterized, by virtue of the role of self-consciousness in personal identity. But Locke’s main themes, the themes that in Williams’s view underlie the philosophical problem of personal identity, evolved into a variety of conflicting positions within the Kantian, Hegelian, and phenomenological traditions. Yet, these traditions are overlooked by Williams and by most analytic philosophers that joined the new debate. It seems as if what went wrong in the long and painstaking historical process of philosophical deliberations in the years between the publication of Locke’s Essay Concerning Human Understanding and the emergence of the current debate – which involved some of the finest philosophers in the history of Western tradition – could only be fixed by retreating to the initial point of the debate where Locke and his contemporaries had left it. Attempts at new beginnings often raise some obvious objections. Yet, they need not be conceived as bound to bear unfruitful consequences. In fact, the debates on personal identity exemplify the advantages of such attempts. The fresh beginning allowed Williams, Strawson, Shoemaker, and those that followed them to effectively use their methods of analysis in their attempts to resolve the conflicting relations between the features of human persons. It was apparently more effective than carrying out these analytic endeavors by linking them to intricate interpretations of their predecessors and contemporaries with whom they disagreed. Williams, Strawson, Shoemaker, and those that followed them did not entirely severe the connection to earlier phases of the debate. The return to the early modern debates was at least partly motivated by the Lockean type of thought experiments such as Locke’s imagined story of the prince and the cobbler, which supported his position. Locke’s thought experiment revealed conflicting, but

Chapter 7. Personal identity, self-consciousness, and bodily identity 

equally sound intuitions. Presumably it was easier to realize why these intuitions, as revealed by the Lockean type of thought experiments, were relevant to the characteristic accounts of the human mind in the analytic tradition. As we shall see below, they were particularly relevant to the novel functionalist accounts of mental representations, mental states, and mental processes. As I will show below, the Lockean type of thought experiments largely influenced the theories of personal identity in the last 50 years, as well as the chain of debates motivated by them. These debates did not merely reveal the conflicting metaphysical intuitions and the problems related to them. They also delineated the relevant focal point of the inquiry by allowing only some of the (conflicting) intuitions related to the concept of a human person into the debate. The metaphysical debates regarding the nature of persons and personal identity involve metaphysical progress no less than the debates on the nature of consciousness. Yet, to the extent that the debates on the nature of consciousness could be characterized as exemplifying ‘descriptive metaphysics’ in Strawson’s terms, some of the significant changes related to the debates on personal identity exemplify a revisionary approach to metaphysics. In the previous chapters, I examined attempts to resolve the problems of consciousness that are consistent with common sense. By contrast, the new debates on the nature of personal identity motivated changes that in many cases depart from common sense. The new debates on personal identity began with three philosophers: Strawson, Williams, and Shoemaker. In their accounts, all three emphasize the conflicting relations between what can be called ‘the self-consciousness intuition’ and ‘the bodily-continuity intuition.’ There are, however, significant differences between these three positions. These differences affected the later developments of this chain of debates. The rest of this chapter is devoted to an examination of these positions and the dialectical relations between them. I will first examine Strawson’s position in Individuals.2 I will then proceed to examine Williams’s position and Shoemaker’s position. 2. Strawson’s Individuals In “Persons,”, the influential Chapter 3 of Individuals, Strawson’s main concern is to provide solutions for traditional epistemic and semantic problems: the problems of ‘solipsism,’ ‘other minds,’ ‘the meaning of psychological predicates,’ and similar issues. He does not attempt to reveal the conditions of personal identity. 2. Strawson’s Individuals was published in 1959. However, the book was based on lectures given at Oxford University in 1954–55.

 Controversies and the Metaphysics of Mind

Yet, as we shall see, his account is relevant to the way in which the philosophers that followed him addressed the question of personal identity. Strawson presents the problem he wishes to resolve by means of two questions: “(1) Why are one’s states of consciousness ascribed to anything at all? And (2) Why are they ascribed to the very same thing as certain corporal characteristics, a certain physical situation &c.?.” (1959: 93) In Strawson’s view, these questions are raised by virtue of the connection between two facts. The first is that the contents of experiences depend on the positions and location of the human body and its sense organs. The second is that the relevant type of dependence is a contingent type of dependence. Visual experiences, for example, could be analyzed as being dependent on the subject’s body in three different ways: (a) with respect to the question of whether the subject’s eyes are open or closed; (b) with respect to the orientation of the subject’s eyeballs; and (c) with respect to the location of the subject’s body – that is, where she sees from. The fact that each visual experience depends on one human body in all three of these ways does not entail that they must depend on one human body. Strawson clarified this by using the following thought experiment: Let A, B, and C be three different bodies. According to Strawson (1959: 90–92), it is conceivable that the visual experiences of a given subject S depend on whether or not her eyes are open on whether or not A’s eyes are open (and not whether B and C’s eyes are open or closed). With respect to the orientation of the eyeballs, they depend only on the orientation of the eyeballs of B (and not on A or C). And with respect to the location of the body, they depend only on the location of C (and not that of A or B). The (single) visual experience of, say, seeing a porcelain vase on the shelf in the drawing room, contains all the above three features. It has the content that it has by virtue of the fact that one’s eyes are open, one’s eyeballs are oriented in a certain way in space, and one’s body is in a particular location in space. But it seems that the fact that the content of this visual experience has all three features does not entail that it causally depends on a single body. The distinct features of the content of the experience could causally depend on different bodies. According to Strawson: Such points illustrate some of the ways in which each person’s body occupies a special position in relation to that person’s perceptual experiences. We may summarize such facts by saying that for each person there is one body which occupies a certain causal position in relation to that person’s perceptual experience, a causal position which in various ways is unique in relation to each various kinds of perceptual experiences he has...We also noted that this complex uniqueness of the single body appeared to be a contingent matter, or rather a cluster of contingent matters; for it seems that we can imagine many peculiar combinations of dependence and independence of aspects of our perceptual experience on facts about different bodies. (1959: 92)

Chapter 7. Personal identity, self-consciousness, and bodily identity 

It is well-known that Strawson’s intention in Chapter 3 of Individuals was to establish that persons are basic particulars (particulars that are irreducible to other types of particulars, Cartesian souls or material bodies) that must have (non-derivatively) both mental properties and physical properties. That is, persons must instantiate both psychological (P) predicates and material (M) predicates. It is usually not noticed, however, that this ontological contention is established by Strawson as a solution to an antinomy that consists of the conflict between the characteristic contentions of two opposed positions, each of which emphasizes one feature of conscious experience: the Cartesian position and the ‘no-ownership’ position. The ‘no-ownership’ position emphasizes the contingent dependence of experiences on material bodies. The pure ego (Cartesian) theorists claim that conscious experiences are necessarily the experiences of the subject that has them. Each of these positions can only give a satisfactory answer to one of the two questions above mentioned. Each of these two positions is able to provide convincing arguments that clarify why the rival position is unable to account for the feature of experiences it is able to account for. The solution – viewing persons as basic particular – is presented by Strawson after he points out the inherent inadequacy of each of the rival positions. The ‘no-ownership’ theorist’s argument begins by noting that the unique causal roles of material bodies in the account of experiences “is sufficient to give rise to the idea that one’s experiences can be ascribed to some particular individual thing.” (1959: 95) Yet, as the ‘no-ownership’ theorist proceeds: So long as we thought in this way, then to ascribe a particular state of consciousness to this body, this individual thing, would at least be to say something that might have been false; for the experiences in question might have been casually dependent on the state of some other body...it might have ‘belonged’ to some other individual thing. (1959: 96)

The contingency noted above seems to suit the Cartesian ego theorist’s goals, whose theory posits a pure ego qua the possessor of experiences and mental states. She maintains that experiences are necessarily had by the subject that has them. Therefore, recognizing the contingency of the relation between a body and a mental state apparently supports the positing of a pure ego. Yet according to the ‘noownership’ theorist, the ego theorist unjustifiably slides from the contingent mode of possession to the necessary mode of possession. If the source of the idea of the ownership of experiences is the causal dependence of a person’s experiences on an individual body, introducing a pure ego as the possessor of experiences in order to give content to the idea of necessary possession cannot be justified. Nevertheless, in Strawson’s view, the ‘no-ownership’ theory is incoherent. Its incoherence is revealed by means of a transcendental argument:

 Controversies and the Metaphysics of Mind

It is incoherent, in that one who holds it is forced to make use of that sense of possession of which he denies the existence, in presenting the case for the denial... The theorist means to speak of all the experiences had by a certain person being contingently so dependent. And the theorist cannot consistently argue that ‘all the experiences of a person P’ means the same thing as ‘all the experiences contingently dependent on a certain body B’; for then his proposition would not be contingent, as his theory requires, but analytic. He must mean to speak of some class of experiences of the members of which it is in fact contingently true that they are all dependent on body B. The defining characteristic of this class is in fact that they are ‘my experiences’ or ‘the experiences of some person’, where the idea of possession expressed by ‘my’ and ‘of’ is the one he calls into question. (1959: 96–97)

Although experiences are contingently (causally) related to particular bodies, the idea that they are contingently had by a given subject involves a confusion. The identity of an experience involves the identity of the subject that has it: “states, or experiences, one might say, owe their identity as particulars to the identity of persons whose states or experiences they are.”(ibid.) The Cartesian theorist seems to be able to account for the necessary ownership of experiences. But her position is equally incoherent. The argument that reveals why her position is also incoherent begins by pointing out a feature that is missed by both theories: ... a necessary condition of one’s ascribing states of consciousness, experiences, to oneself, in the way that one does, that one should also ascribe them, or be prepared to ascribe them, to others who are not oneself. (1959: 99)

As Strawson observes (ibid.), phrases that ascribe states of consciousness are not ambiguous. They do not express different meanings when used in the first-person or the third-person modes. It must be stressed that Strawson’s claim is about our language, our conceptual scheme and our self-conception, that is, how we use and understand our language. And indeed, to the extent that self-knowledge requires one to distinguish oneself from another self, one’s own mental states from the mental states of other people, the Cartesian theorist must explain how, within her position, it is possible to ascribe mental states to others and to recognize persons different from herself. But if knowledge of others is based on sense perception, if what is perceived are material bodies, and if, as the above argument demonstrates, persons qua subjects of experiences are required for the account of the necessary ownership of experiences (that is part of the essence of experiences), persons cannot be identical to material bodies. Nevertheless, one cannot indentify other persons merely as subjects of experience, that is, without being able to perceive them. A necessary condition of self-ascription of mental states is the ascription of mental states to others. The necessary ownership of experiences by the subject that

Chapter 7. Personal identity, self-consciousness, and bodily identity 

has them is a necessary condition that all experiences must satisfy. It might be assumed that though the existence of other people cannot merely be based on perception (and, therefore, the possibility to ascribe mental states to them cannot merely be based on perception), one can, nevertheless, infer the existence of other people from their perceived bodies, if one supposes that the same type of relation between one’s own experiences and one’s body also holds between another person’s experiences and her body. Yet, as Strawson notes: “...this suggestion is useless. It requires me to have noted that my experiences stand in a special relation to body M, when it is just the right to speak of my experiences at all that is in question.” (1959: 101) Indeed, the mistake involved in this way of conceiving other persons is, in fact, the same mistake that the ego theorist commits, according to the ‘no ownership’ theorist: the ego theorist unjustifiably slides from the admissible meaning of ‘ownership’ of experiences (that they contingently depend on a given material body) to the supposition that they are necessarily owned by a given person. The two conflicting positions, therefore, represent two sides of an antinomy. Each of the positions cannot establish itself without involving reasons that lead to the main contention of the rival position. And yet, the main contention of each position motivates an argument the conclusion of which is the denial of the rival position’s main contention. In other words, the type of arguments used by each position against the rival position is that of type (B) discussed in Chapter 2. The ‘no-ownership’ theorist reveals the groundlessness of the shift from the contingent sense of ownership to the necessary sense of ownership. But, according to Strawson, the ‘no-ownership’ theorist cannot establish the claim that no subject necessarily owns the experiences that are contingently related to a given body, without presupposing the necessary sense of ‘ownership.’ The Cartesian theorist, on the other hand, for whom the necessity of ownership of experiences raises no difficulty, slides from the contingency of the dependence of one’s experiences on one’s body to the necessity of the ownership of experiences. The single body, the source of the idea that experiences should be ascribed to a particular thing (according to the ‘no-ownership’ theories), is left out of the Cartesian’s account. But since, in Strawson’s view, self-ascription requires the possibility to ascribe mental states to others, the Cartesian theorist is unable to account for self-ascription, and therefore, for the necessary sense of ownership. Strawson’s solution for this conflict reveals a possibility that both theorists seem to miss: Persons are basic particulars that must instantiate (non-derivatively) both psychological predicates and material predicates. The fact that they are basic particulars implies that they are distinct from the material body that they possess. Although persons necessarily have bodily features, they cannot be identified with bodies. The distinction between the two types of basic particulars, material bodies

 Controversies and the Metaphysics of Mind

and persons, is based primarily on the role of P predicates in the characterization of persons: ... this characterization of the type is still very opaque and does not at all clearly bring out what is involved. To bring this out, I must make a rough division, into two, of the kind of predicates properly applied to individuals of this type. The first kind of predicate consists of those which are also properly applied to material bodies to which we would not dream of applying predicates ascribing states of consciousness. I will call the first kind M-predicates...The second kind consists of all the other predicates we apply to persons. These I shall call P-predicates. (1959: 104, my italics)

In Strawson’s view, P-predicates cannot be ascribed to mere material bodies. Indeed, as Strawson notes in the same context, “though not all P-predicates are what we should call ‘predicates ascribing states of consciousness’ (e.g. ‘going for a walk’ is not), they may be said to have this in common, that they imply the possession of consciousness on the part of that to which they are ascribed.” (1959: 105) As I noted in the introduction to this book, Strawson is a descriptive metaphysician. Nevertheless, his dialectical arguments seem to have far-reaching ontological consequences. It is clear that in Strawson’s view the identity of a human person cannot merely consist in the identity of a human body. Yet, properly speaking, the identity of a human person cannot be separated from the identity of her body. As noted above, one of the two questions with which Strawson begins his investigations is: “Why are they (mental states and experiences) ascribed to the very same thing as certain corporal characteristics, a certain physical situation &c.?.” (1959: 93) Even if one accepts Strawson’s view that the ascription of mental states is necessarily correlated with the ascription of material states, Strawson’s account seems to leave open the question regarding the relation of identity between a person and her material body. The reasons that supported Strawson’s contention that persons must satisfy both P-predicates and M-predicates were basically epistemic reasons that concerned the possibility of ascribing mental states to others. But the conditions required for ascribing mental states to others do not involve the supposition that one necessarily has the body that one has. The question with regard to the relation of identity that arises with respect to persons and their material bodies is naturally invited by Strawson’s philosophical reflections. On the one hand, the necessary connection between the ascription of P-predicates and M-predicates, and the necessary ownership of mental states by the subject that has them (qua features that belongs to the very conception of a mental state or an experience) seem to imply that the identity of the subject of experience must involve or be depended upon the identity of her body. Nevertheless, the contingency of the

Chapter 7. Personal identity, self-consciousness, and bodily identity 

relation between the particular body and the person’s mental states (one of the reasons for distinguishing persons and bodies), seems to imply that the identity of the body is not a necessary condition for the identity of a person. 3. Williams: Personal identity and individuation It is rare to find articles that set the stage for a whole discourse on a given subject by containing the main questions concerning this subject, the main types of answers presented as a responses to these questions, as well as an expression of the gaps hidden in the conflicting intuitions regarding the given subject matter. Bernard Williams’s “Personal identity and individuation” (1956–7) (PII) is close to being an example of such a paper. On first approximation, Williams’s position regarding persons and their conditions of identity seems to be close to Strawson’s position. They both appear to be motivated by the same, apparently conflicting, features of personhood, and they both appear to arrive at similar conclusions. Yet, while one of Strawson’s goals was to explain why one’s own states of consciousness (presumably including states in which one is conscious of oneself) must be ascribed to the very same thing as certain corporal characteristics, Williams aspired to clarify why bodily identity is always a necessary condition for personal identity. Indeed, as will be shown in the present section, both positions differ. Williams does not explicitly address Strawson’s arguments in PII. But explicit criticism of Strawson’s position regarding his account of persons is found in Williams’s review of Strawson’s Individuals.3 The critical arguments presented in Williams’s review in fact use the same types of reasons that he used in PII in order to criticize views that maintain that the identity of a person can be separated from the identity of her body. As noted above, in Williams’s view, the question of personal identity poses a unique philosophical problem for two reasons: the relation between self-consciousness and personhood, and the fact that questions of personal identity are not answered “merely by deciding the identity of a certain physical body.” (1973: 1) Williams’s objective in PII is primarily polemical: to undermine the supposition that bodily identity is not a necessary condition for personal identity. For that purpose, he distinguishes between two versions of such views. The weak version maintains that in some conceivable cases, personal identity does not entail bodily identity. In these cases, the psychological conditions suffice to identify a person. The stronger thesis maintains that “there is no conceivable situation in which 3. Williams’s review of Strawson’s book was originally published in 1961. It was reprinted in Williams 1973.

 Controversies and the Metaphysics of Mind

bodily identity would be necessary.”(ibid.) The strong version is identified with Locke’s position, whereas the weak version is not identified with any position. Nevertheless, it is not hard to see that Strawson’s position could in fact be interpreted as a version of the weak position. This was indeed how Williams interpreted Strawson’s views in other places. Although in Williams’s view the problem of personal identity is created by the role of self-consciousness of identity in personhood, Williams does not look in PII for an account of self-consciousness that clarifies why bodily identity is a necessary condition for personal identity. Rather, his arguments are devised to reveal the illusions related to self-consciousness when investigating the grounds of personal identity: It is not part of my aim to discuss in general consciousness of self. I have tried in this section to show in a limited way that although we may have the feeling that, by consideration of it alone, we may be given the clue to personal identity, this is in fact an illusion. That it is an illusion is disguised by those theories of personal identity which, by assuming no particular point of view, try to get the best of both worlds, the inner and the outer. If we abandon this for a more realistic approach, the facts of self-consciousness prove incapable of yielding the secrets of personal identity, and we are forced back into the world of public criteria. (1973: 14–15)

Williams’s argument in PII has three steps. The argument begins by examining the conditions that determine the identities of others. It continues with the examination of the possibility of body exchange, and the final step clarifies why memory, qua non-bodily criterion of identity (as Locke conceives it) could serve as a criterion for self-identity. Williams explains first the differences between identity and exact similarity. Two persons may share the same personal character, even the same cognitive capacities and, nevertheless, be distinct persons. Hence, if the reasons that support the statement that ‘X = Y’ suffice merely to establish exact similarity, this statement may be false. Next he examines the strong version of the ‘not-necessary’ thesis, that is, the Lockean type of memory theory according to which the identity of the body is always not a necessary condition for personal identity. As Williams notes: ... important features of memory have to be borne in mind. I. To say ‘A remembers x’, without irony or inverted commas, is to imply that x really happened; in this respect ‘remember’ is parallel to ‘know’. II. It does not follow from this, nor is it true, that all claims to remember, any more than all claims to know, are veridical; or, not everything one seems to remember is something one really remembers. (1973: 3)

Memory theories of personal identity maintain that if X remembers that Y observed E or that Y did A, X is identical to Y. According to one interpretation, this

Chapter 7. Personal identity, self-consciousness, and bodily identity 

is clearly false. If Bill remembers that John did A, Bill is not identical with John. On the other hand, if ‘X remembers doing A’ means ‘X remembers that he himself did A,’ the identity of X and Y follows, if this proposition is true. But in this case, if the evidence of memory is supposed to function as a criterion of personal identity, memory theories must face the problem of circularity invoked by Butler against Locke’s theory (1973: 4). It might be assumed that circularity may well not be much of a problem for the memory theories. If the identity of the X that remembers doing A with the Y that did A is entailed by the truth of the relevant kind of memory statements, memory theories succeed in basing personal identity merely on memory. But here the distinction between what memory statements mean and what is entailed by their being true is called for. True memory statements (of the relevant kind) indeed entail the identity of the remembering person and the person that did the remembered action or that observed the remembered event. But memory statements could be false. There must be a way to distinguish between false claims to memory and true ones. As Williams observes, no such distinction could merely be based on firstpersonal capacities. Remembering doing A or observing E does not involve the application of criteria of identity. But, first-person memory statements could be false. They could be assessed as such on the basis of third-personal criteria of identity. In other words, first-person memory assertions the making of which does not involve criteria of identity do not suffice to determine the identity of the person who makes the first-person statement and the person that did the remembered action or that experienced the remembered experience. Williams establishes the last contention by examining what is involved in the Lockean type of thought experiments that supposedly support the claim that personal identity differs from bodily identity. Say that when Charles wakes up one morning, he manifests significant changes in his personal character, memories and cognitive capacities. Further investigations reveal that his new personal character and his new memories and cognitive capacities are exactly similar to those of the late Guy Fawkes. Should we say that Charles’s body is now connected to Guy Fawkes, that is, that Guy Fawkes has been resurrected in Charles’s body? According to Locke, the answer must be “yes!” However, as Williams notes, if such a change could have occurred to Charles, it is not possible to deny that the same change could also have happened to his brother Robert. (1973: 8) In other words, both of them may manifest the same personality as well as the memory and cognitive capacities as those of the late Guy Fawkes. Personality and cognitive capacities allow reduplication. But reduplication suffices merely for exact similarity. It does not suffice for identity.

 Controversies and the Metaphysics of Mind

It might be argued that the above argument leaves out what seems to be so important for the memory theorist, that is, the fact that one’s personal identity is somehow connected to one’s first-person capacities to remember past experiences and actions. Yet, as Williams clarifies, it is useless to introduce these capacities at this point of the argument. Clearly, Charles, the person suffering a radical memory change cannot compare the memories he had in the past to his present ones. Supposedly, he simply seems to remember what Guy Fawkes used to remember. In remembering, he does not rely on any criteria of personal identity. Yet, as Williams observes, this does not mean that the individuation of persons need not involve such criteria. Given the possibility of reduplication “a criterion must be used by someone.” (1973: 13) Hence to the extent that memory theorists aim to disconnect the conditions of the identities of persons from bodily identity, they fail both in the weaker version of this thesis and the strong version.4 As noted above, Williams’s intentions are polemical. He undermines the rival position without introducing an alternative theory that clarifies how the following feature of his own position is possible: “that some reference to him [a person] cannot be understood as references to his body or to parts of it.” (1973: 2) The same discrepancy between reference to persons and reference to bodies also motivated Strawson’s account of persons. Strawson, however, aimed to provide one answer to two questions: “(1) Why are one’s states of consciousness ascribed to anything at all? And (2) Why are they ascribed to the very same thing as certain corporal characteristics...?.” (1959: 93) Williams seems to address the second of these questions and does not provides an answer to Strawson’s first question. As we shall see below, this is significant for subsequent developments of the debates on personal identity.

4. As noted above, Williams does not refer explicitly to Strawson in PII. Yet, the consequences of his argument are applied to Strawson’s theory in other places. His main objection concerns Strawson’s distinction between P-predicates and M-predicates. As Williams notes, Strawson’s distinction is based on the notion of ‘state of consciousness’ that is “slipped without introduction” into the discussion. (1973: 122) Not all P-predicates are, according to Strawson, predicates that involve consciousness. But what distinguishes the different sorts of P-predicates? And how, given this distinction, is it possible to distinguish P-predicates such as ‘x is hit by a stone’ from the corresponding M-predicate? In Williams’s view, Strawson’s appeal to consciousness and, in particular, to perceptual experiences, indicates traces of Cartesianism hidden in Strawson’s theory. These traces are particularly noticeable in the thought experiment that aims to reveal the contingency of the relation between experiences of persons and their bodies. If persons, who are the subjects of experiences, are contingently related to their bodies, bodily identity cannot be a necessary condition for personal identity. Hence, Williams’s argument in PII is applicable to Strawson’s theory.

Chapter 7. Personal identity, self-consciousness, and bodily identity 

4. Shoemaker: Self-Knowledge and Self-Identity The beginning of Shoemaker’s Self-Knowledge and Self-Identity parallels the beginning of Williams’s PII. At the beginning of this book, Shoemaker refers to self-knowledge as the feature of persons by virtue of which “there is a special philosophical problem, or set of problems, about the nature of persons (selves) and personal (self) identity.” (1963: 2) Even though Williams is not explicitly mentioned in this book, Shoemaker’s overall argument there could be interpreted as a critical response to Williams’s contentions. His argument aspires to overcome Williams’s dismissive appraisal of first-person memory states and memory statements in the account of persons’ identities without abandoning the necessary role attributed by Williams to the body in constituting personal identity.5 Shoemaker’s body-change thought experiment had a considerable impact on a whole generation of philosophers of mind. His thought experiment invites us to imagine a brain transplant involving two persons, Mr. Brown and Mr. Robinson. Assuming that future medical technology will make it possible to transplant brains, let us consider the following possibility: Two men, a Mr. Brown and a Mr. Robinson, had been operated on for brain tumors, and brain extractions had been performed on both of them. At the end of the operations, however, the assistant inadvertently put Brown’s brain in Robinson’s head, and Robinson’s Brain in Brown’s head. One of these men immediately dies, but the other, the one with Robinson’s body and Brown’s brain, eventually regains consciousness. Let us call the latter ‘Brownson.’ Upon regaining consciousness Brownson exhibits great shock and surprise at the appearance of his body. Then, upon seeing Brown’s body, he exclaims incredulously “That’s me lying there!” Pointing to himself he says “This isn’t my body; the one over there is! When asked his name he automatically replies “Brown.” He recognizes Brown’s wife and family (whom Robinson had never met), and is able to describe in detail events in Brown’s life, always describing them as events in his own life. Of Robinson’s past life he evidences no knowledge at all. (1963: 23–24)

The Brown-Brownson case is similar to Locke’s case of the cobbler and the prince. Yet, in contrast to Locke’s case, Shoemaker’s case is a case of partial bodily continuity. Nevertheless, as Shoemaker clarifies, if Brownson is the same person as Brown, their identity could hardly be based on bodily criteria. It is absurd to suppose that Brown is identical to a proper part of his body, that is, to his brain. It should be noted that the Brown-Brownson case may in fact serve as a counterexample to Williams’s main contention in PII, according to which bodily continuity 5. Shoemaker’s book grew out of his Ph.D. dissertation that was submitted to Cornel University in 1958, and a paper published in 1959.

 Controversies and the Metaphysics of Mind

is a necessary condition for personal identity. Regarding the identity of Brown and Brownson, it is psychological rather than bodily criteria – that is, the fact that Brownson has the same personality and the same memories – that appear to be most significant. But in contrast to Locke’s case, Shoemaker’s case does not seem to be open to the charge of reduplication.6 The additional feature added to Locke’s original case by the Brown-Brownson case concerns the type of causal connections between Brown and Brownson’s respective psychological states. The causal connections between Brownson’s memories and Brown’s past experiences appear to be the same type of causal connections exhibited in the lives of normal persons. Assuming some form of mind-brain supervenience, Brownson’s personality traits and cognitive capacities supervene on Brownson’s brain as Brown’s characteristics and cognitive capacities supervene on his brain. As we shall see, this is significant for subsequent developments of the debate. As Shoemaker notes, “it is primarily the way in which first-person statements are made and known, not the way third-person statements are made and known, that gives rise to a special problem about the nature of persons and personal identity.” (1963: 9) His position regarding the meaning of ‘I’ is close to Strawson’s position regarding the meaning of P-predicates: The idea that the word ‘I’ stands for or refers to something – an idea that is presupposed by the question ‘What am I?’ – has its basis, not primarily in the fact that the word ‘I’ functions grammatically as a subject, but in the fact that corresponding to any first-person statement there are third-person statements that are in a certain sense equivalent to it and are certainly ‘about something.’ (1963: 12–13)

The term ‘I’ has the same meaning when it is used in first-person statements and when it is used in third-person statements. Although Strawson is not explicitly referred to in Self-Knowledge and Self-Identity, Shoemaker apparently also shares Strawson’s idea that the word ‘person’ has a special status that distinguishes persons from mere bodies: While there is a sense in which what I am, what the word ‘I’ when I use it refers to, is anything that I can truly be said to be, it now appears that the sense in which I am a person, or in which ‘I’ stands for a person, is different from the sense in which I am, or in which ‘I’ stands for, a mammal, an American, or a male. It is only contingently true that I am mammal or an American. But since the word ‘I’ refers to the being that is using it to refer, it necessarily refers to a person. (1963: 15)

6. As we shall see in the next chapter, this is not exactly so.

Chapter 7. Personal identity, self-consciousness, and bodily identity 

As Shoemaker notes (ibid.), the above difference might encourage one to suppose that one is essentially a person and only per accident a human body. It may seem to follow “from the view that a person (or self) is something logically distinct from a human body...that the identity of a person is not the same as and does not necessarily involve the identity of a person’s body.”(1963: 18–9) Yet, that some facts about a human person are not facts about her body implies that some properties of persons are not physical properties and not that the person is not a body: Only if we take the word ‘body’ to mean ‘something all of whose properties are physical properties’ can we conclude from this argument that a person is not a body. And that something is not a body in this sense does not imply that it is completely nonphysical entity, that it does not have physical properties as well as non-physical ones, or that its identity does not involve the identity of a body. (1963: 19)

The above passage contains at least two contentions. First, it seems that Shoemaker here endorses property dualism. Property dualism seems to have direct implications regarding the connection between personal and bodily (physical) identity. The fact that persons are subjects of predicates expressing non-physical properties means that they are not identical to their (merely) physical bodies.7 Secondly, in Shoemaker’s view, the fact that persons satisfy predicates that attribute non-physical properties to them does not entail that the identity of a person could be detached from the identity of her body. It might seem that the latter contention entails that cases of body exchange, cases in which the person continues to exist though she is not connected to the particular body with which she was identified, are impossible, in conformity with Williams’s position. However, the first contention seems to allow body exchange. Even if persons must necessarily have some physical properties, it is not entailed that their identity must be equated with the identity of a particular body. The Brown-Brownson case apparently exemplifies the middle ground that Shoemaker wishes to defend according to which personal identity may be preserved in cases of body exchange without abandoning the necessary connection between personal identity and bodily identity. The strong intuition to the effect that Brownson is Brown apparently involves the supposition that the relation between Brown-Brownson and the distinct bodies to which he is connected in the different periods of his existence (before the operation to Brown’s body and after the operation to Robinson’s body) is a contingent relation. The implicit question addressed by Shoemaker is whether one can devise a coherent account that would satisfy Williams’s weak version of the ‘not-necessary’ contention regarding the 7. Shoemaker does, however, allow for a position that distinguishes between mere physical bodies and live bodies.

 Controversies and the Metaphysics of Mind

connection between the identities of persons and their bodies, cases “in which bodily identity fails, but in which the other conditions will be sufficient for an assertion of personal identity,” (1973: 1) without abandoning the necessary connection between personal identity and bodily identity. At first approximation, such a position appears to be a contradictory position. Yet, Shoemaker’s arguments aim to clarify that this is a false impression. I will not present the arguments with which Shoemaker defends his position in detail. I will briefly examine only a thread of arguments that is relevant to the subsequent development of the debate on personal identity. His arguments establish four theses: 1. Knowledge of identity expressed in first-person memories states (states in which one remembers one’s own past experiences or actions) does not involve criteria of identity. 2. Bodily criteria are criteria for personal identity. 3. It is a conceptual truth that memory statements must be generally true. 4. In some contexts, first-person memory statements can, from the third-person viewpoint, serve as criteria for personal identity. In Self-Knowledge and Self-Identity, Shoemaker accepts the concept of a criterion that became current through the work of the later Wittgenstein. Criteria for the truth of a judgment are “those states of affair that are (whose existence would be) direct and noninductive evidence in favor of the truth of a judgment.” (1963: 3) Determining that this table is the same table as the table I used yesterday involves criteria of identity. But first-person memory statements do not appeal to criteria of identity. Remembering watching the football match between France and Brazil a few years ago or eating breakfast a few hours ago does not involve any act of identification. I do not identify myself with the person that experienced the past events or that performed the action in the past. Memory statements entail the identity of the remembered person and the person that remembers. Their truth conditions involve the respective personal identity. But memory statements are not immune to error. The fact that they are open to the possibility of mistakes implies that it must be generally true that the body of the remembering person is the same body as the body of the person that experienced the past events or that performed the actions in the past. Sameness of body is, in other words, a third-person criterion of identity. The identity of the person that misremembers must necessarily be linked to the identity of that person’s body. That bodily identity functions as a criterion for personal identity is not merely required in order to establish that some first-person states of memory might be mistaken. It is also required by our most basic conceptual and perceptual capacities. The content of perceptual experiences, and the conceptual capacities that involve perception entails that it is generally impossible for a person not to be

Chapter 7. Personal identity, self-consciousness, and bodily identity 

connected to a particular body. It is impossible, for example, that the place from where she observes physical objects and states of affair will steadily shift in such a way that it will be impossible to identify her, qua observing person, with a given spatiotemporal (bodily) entity. Yet, as the Brown-Brownson case attests, the fact that it cannot be generally true that persons are disconnected from particular bodies, does not entail that cases of body exchange are not possible. In other words, there are reasons to claim that the identities of persons must (qua conceptual truth) in general be linked to the identities of particular bodies. But we also have reasons to claim that body exchange is conceivable. In other words, in contrast to Williams’s position, the necessary connection between the identity of a person and the identity of her body allows exceptions. Shoemaker’s treatment of the connection between first-person memory statements and personal identity makes use of similar modal ideas. Not all first-person memory statements are true. Yet, this does not entail that they could all be false or even that the majority of such statements could be false. According to Shoemaker, it is conceptually impossible that all or the majority of them are false. A person all or most of whose memories are mistaken would be unable to have knowledge of language and of how language applies to reality. In contrast to Williams’s contention, memory states and first-person memory statements are relevant to the identities of persons, if they are true, not merely by entailing that the remembering person is identical to the person who performed the remembered action, or who experienced the remembered event. Also, they are not merely sources of illusions and fallacious philosophical theories of personal identity. Rather, ... it seems to me that the fact that persons can make identity judgments about themselves, or statements implying the persistence of themselves, without using criteria of personal identity, and that such statements are generally true and can be said to express knowledge, itself constitutes an important difference between the identity of persons and the identity of other things. (1963: 244)

Self-knowledge of identity is not only close to being an essential feature of persons that distinguishes them from other things in the world. If memory statements are generally true, memory could serve as a criterion of personal identity: I have maintained that it is a necessary truth, not a contingent one, that confident and sincere memory statements are generally true. If this is so, inferences of the form ‘He claims to remember doing X, so he probably did do X’ cannot be merely inductive, for they are warranted by a generalization that is necessarily rather than contingently true. (1963: 245)

Based on this, one could explain the conceivability of the Brown-Brownson case. Indeed, the Brownson case is more complicated than expressed by (a) below: a. He claims to remember doing X, so he probably did X.

 Controversies and the Metaphysics of Mind

If (a) can establish identity in normal cases, there is reason to suppose, according to Shoemaker, that (b) below is also true: b. He claims to remember doing X, Y, and Z under such and such circumstances and at such and such times and places, and it is true that X,Y, and Z were done by someone under precisely those circumstances and at such and such times and places so there is reason to think that he is the person who did those actions. (1963: 245) As Shoemaker notes, the fact that cases such as the Brown-Brownson case “incline us to admit the possibility of body exchange, or leave us in doubt as to what to say, is itself prima facie evidence that memory is a criterion of personal identity.” (1963: 246) Endorsing this type of criterion for the identity of persons is indeed not common linguistic practice. In a sense, it is a new criterion. But accepting the new criterion does not mean that one has completely abandoned the old criteria (ibid.). The judgments of those who employ it cannot be said to be “not in accord with our criteria.” (1963: 247) Moreover, It cannot be said that they would be abandoning bodily identity as a criterion of identity; all that can be said is that they would be refusing to regard this criterion as decisive in all cases, and would be allowing it to be overweighed by other criteria in some circumstances. (1963: 247)

Shoemaker’s conclusion in the last quotation clearly refutes one of Williams’s main contentions. The dichotomous confrontation between bodily criteria of personal identity and Lockean criteria is replaced by a middle-ground position that seems to be able to espouse the necessary role of bodily identity in constituting the identities of persons as well as the first-person type of self-knowledge of identity expressed in first-person states of consciousness. Yet, the necessary role of bodily identity requires a revision of Locke’s original position. In the same vein, the conceivability of cases in which the bodily criterion is not the criterion that determines the identity of A and B (but rather the revised memory criterion does) requires a significant revision of Williams’s account. By employing a thought experiment in which a (so far) technically impossible case is imagined, Shoemaker supports a middle-ground position regarding the connection between personal identity, bodily continuity, and self-consciousness, which attempts to overcomes the previous dichotomy. By requiring revisions in the relevant positions, Shoemaker’s position became a relevant alternative to the positions he aspired to overcome. It should be noted, however, that what seems to be left unanswered in Shoemaker’s account, no less than in Strawson’s and Williams’s respective accounts, is the question regarding what constitutes the identity of persons. Given the BrownBrownson case, bodily identity does not suffice for personal identity, even if

Chapter 7. Personal identity, self-consciousness, and bodily identity 

bodily identity has a necessary role in determining the identity of a person. Identity is a one-one relation. The sense in which in Shoemaker’s theory bodily features function as necessary conditions for personhood allows for the contingency of the relation between persons and their bodies. Given that identity is a one-one relation, even if B is the actual body of S, and even if S must instantiate physical properties at each point of her existence, B is not identical with S. Neither the bodily criterion nor the psychological criterion suffice to determine the identity of persons. Each of these criteria may not be satisfied without entailing that personal identity itself is lost. Shoemaker’s account was not meant to provide a full-fledged account of the grounds that constitute the identity of a person. Rather, he strove to characterize the relations between the conditions or criteria of identity. By attempting to explicate the difficulties related to the connection between the apparent conflicting criteria and by suggesting the most plausible solution to these difficulties, Shoemaker’s account did not remove the mystery inherent in the nature of the identities of persons, but rather enhanced the mystery and rendered it even more perplexing than before.

chapter 8

From transcendental arguments to revisionary metaphysics 1. Wiggins: Relative identity and personhood In 1967, the British logician and philosopher David Wiggins published Identity and Spatio-temporal Continuity (ISC), a significant contribution to the debate on identity and personal-identity. Wiggins’s main purpose in this book was to defend an account of identity that establishes what he calls ‘principle D’ without entailing or involving what he calls ‘principle R.’ D is the principle according to which “if someone tells you that a = b, then you should always ask them ‘the same what as b?’.” (1967: 1) R is the principle of relative identity according to which a could be the same f as b but not the same g as b, even when a or b itself is a g.1 According to Wiggins, though D is true, R is false.2 Wiggins’s arguments in the first part of ISC refute R. His main idea, which will not be examined in detail here, is that R could be shown to conflict with Leibniz’s law. The second part of the book clarifies the reasons for accepting D, and presents a formal theory of substances. In the final part of his book, by using the logical reflections included in the first parts, Wiggins endeavors to resolve the apparent conflict between the two supposedly incompatible criteria of personalidentity: the memory criterion and the bodily continuity criterion. Wiggins’s arguments do not always explicitly address the reasons that support the conflicting contentions of the opposed positions. Yet, it is easy to detect the dialectical features of his argument aiming to resolve the debate on personal identity. To begin with, according to D, spatiotemporal continuity cannot be a sufficient basis for the identity of a and b. Even given spatiotemporal continuity, one could always ask: ‘the same what as b?’ The statement ‘a is the same person as b’ could be interpreted as: 1. a is the same psychological (Lockean) object as b. 1.

Throughout this chapter I will use Wiggins’s notations.

2. Wiggins also claims (1967: 1) that D does not entail the principle C, according to which “to specify the something or other under which a and b coincide is necessarily to specify a concept f which qualifies as adequate for this purpose, and hence as a sortal, only if it yields a principle of counting for fs”. In Wiggins’s view, C is also false.

 Controversies and the Metaphysics of Mind

2. a is the same body as b. 3. a is the same person as b. Given the difficulties of (1) as revealed by Williams, we are left with either (2), or (3) (Strawson’s suggestion). However, given Wiggins’s account of the logical features of the identity of substances, both possibilities appear to be inadequate. The problem with (2) is that the predicates ‘...is a human body’ and ‘...is a person’ appear to express two distinct sortal concepts (assuming that ‘person’ and ‘body’ are sortal concepts). Shoemaker’s Brown-Brownson’s case supports the intuition that one could conceive of cases in which a will be the same person as b without being the same live human body as b. As I showed in the previous chapter, although Shoemaker’s theory distinguishes between personal identity and bodily identity, it rejects the Lockean psychological account of the identities of persons that disconnects the bodily continuity criterion and the psychological criterion. But if one rejects relative identity (that is, Wiggins’s principle R), Shoemaker’s theory leaves one without an answer to the question regarding the identities of persons. On the one hand, assuming as does Shoemaker that the account of the identity of persons must be linked to bodily identity, it appears that Shoemaker is forced to accept (1) below as being generally true: 1. If a is the same person as b, it must be generally true that a is the same living body as b. On the other hand, given that with regard to each person there are contexts in which being the same person does not entail being same body, it also seems that Shoemaker must accept (2) below: 2. a could be the same person as b without having the same live human body as b. Clearly, in Shoemaker’s view, (1) and (2) must both be true of each human person. There are, indeed, contexts in which the bodily continuity criterion is the relevant criterion that determines “which of the bs is identical with a?” There are other cases, however, in which the psychological criterion is the relevant criterion. But if different criteria determine persons’ identities in different contexts, it appears that the method of identification implicitly employed by Shoemaker implies relative identity. As noted in the previous chapter, according to Shoemaker, a must in most contexts be the same living human body as b. But a (in the past) could in other contexts be the same person as b but not the same live human body as b. Although Shoemaker’s position, as presented in Self-Knowledge and Self-Identity, is open to the charge of being committed to relative identity, it is not clear that it must be open to this charge. Presumably, given some qualifications, it can be saved from this charge. Shoemaker might avoid the charge, if one distinguishes, as Wiggins does (1967: 10–11), between three senses of ‘is’: (α) the ‘is’ of predication;

Chapter 8. From transcendental arguments to revisionary metaphysics 

(β) the ‘is’ of constitution; and (γ) the ‘is’ of identity. Clearly, (γ) represents the sense of ‘is’ that creates a problem for Shoemaker’s theory. If ‘is’ in ‘a is the same body as b’ is used in the sense of strict identity, then given that both (1) and (2) are true of each person, the position is vulnerable to the charge of being committed to relative identity. Also, it is clear that the sense of ‘is’ involved in ‘a is the same body as b’ cannot be that of predication. If ‘is’ in ‘a is the same body as b’ is used in its constitutive sense, the relation between persons and their bodies is similar to that between a lump of clay and the statue it constitutes. If the lump of clay constitutes the statue, the lump of clay is not the same object as the statue, but, rather, a different object that spatiotemporally coincides with the statue at some stages of its existence. If ‘x constitutes y’ is the relation between persons and their bodies, persons are not identical to their bodies.3 They are constituted by their bodies. Yet, it is not clear whether ‘x constitutes y’ is the relation between Brown and Brownson and their bodies. The above distinction between personal identity and spatiotemporal bodily continuity might raise the concern that the criterion that determines the identity of persons is the psychological criterion. Since the two criteria lead to conflicting results, the bodily criterion cannot serve as a criterion for the identity of a person in the strict sense. Wiggins’s main intention in Part Four of his book is to relieve this worry. In Wiggins’s view, the bodily criterion of personal identity seems to involve three difficulties. Firstly, since my body lasts longer than I do, or perhaps I last longer than my body, I must be distinct from my body (1967: 35). Secondly, as Wiggins notes, the bodily criterion proposed by Williams “seems to ride intolerably roughshod over the memory criterion.” (ibid.) And, finally, there is a strong intuition, voiced by philosophers such as Reid, that bodily identity is never perfect identity and that, therefore, since the concept of identity does not allow ambiguities or degrees, bodily identity is never perfect identity. According to Reid, the concept of (perfect) identity is applicable to persons. In addition, the identity of a person a with a person b cannot be arbitrarily determined by arbitrarily fixing the conventions of our language. In contrast to that, “questions about the identity of a body are often questions about words.” (1967: 45) The main idea underlying Wiggins’s solution to the conflict between the two criteria is expressed in the following passage: To see the answer to this problem, and make the appropriate concessions to these three objections, is to see the essential unity of the concept of a person and the 3. The claims that constitution is not identity and that bodies constitute persons are central to the theory of personal identity defended By Rudder Baker (1997, 2000). Rudder Baker emphasizes first-person consciousness of ‘oneself as oneself’ as essential to the identities of persons. First-person consciousness of ‘oneself as oneself’ distinguishes persons from the bodies that constitute them.

 Controversies and the Metaphysics of Mind

equivalence of the revised spatiotemporal criterion to any admissible memory criterion of personal identity. This equivalence does not arise simply and trivially from the role of spatio-temporal criterion in determining the veridicality of single claims to remembering X-ing or Y-ing. The spatio-temporal criterion and the memory criterion, when it is properly founded in the notion of cause, inform and regulate one another reciprocally – indeed they are really aspects of a single criterion. For the requirement of spatio-temporal continuity is quite empty until we say continuity under what concept (see Part Two). And it will be argued that we cannot specify the right concept without mention of the behavior, characteristic functioning, and capacities of persons, including the capacity to remember some sufficient amount of his past. It is this characteristic functioning which gives the relevant kind of spatio-temporal continuity for kinds of parcel of matter we individuate when we individuate persons. (1967: 45–6)

Wiggins, in fact, invokes some version of a functionalist conception of the mind in order to specify the relevant type of “the same what...?” required for the relevant criterion of spatiotemporal continuity for persons. When the spatiotemporal criterion and the bodily criterion are conceived as founded on the notion of cause, they are aspects of one and the same criterion. Supporters of the bodily continuity criterion of personal identity maintain that the answer to the question “the same what...?” must include the behavior, characteristic functioning, and memory of the spatiotemporal continuous person. According to Wiggins, this Aristotelian idea is helpful in resolving the supposed conflict between the two criteria for personal-identity, and can even explain how Shoemaker’s body exchange thought experiment is conceivable: I think the Aristotelian view of what a life and a psūchē are, taken in conjunction with a causal theory of memory and our previous analysis of spatio-temporal continuity, put us in a position to improve a bit Shoemaker’s rather hasty conclusion. Abandoning Aristotle’s terminology, the point is that spatio-temporal continuity under the concept psūchē or person (living-body as opposed to living body) is not quite the same spatio-temporal continuity under the concept body. And if we take the conceptual analysis of vital functions as seriously as Aristotle and the memory theorist do, and as seriously as anyone who thought Brownson was the same man as Brown would have to take it, then surely what matters is not bodily continuity but the continuity of Brown’s life and vital functions as they are planted in one body and recognizably and traceably transposed in another body. (1967: 50–51)

In Shoemaker’s thought experiment, psychological continuity was not divorced from bodily continuity. Although Brown’s body is not identical to Brown’s brain, Brownson is psychologically continuous with Brown by virtue of the fact that he has Brown’s brain. Yet, as Wiggins notes: “The brain does not figure in the a priori

Chapter 8. From transcendental arguments to revisionary metaphysics 

account of person or same person except perhaps under the description ‘seat of memory and other functionally characteristic capacities.’” (1967: 51) In different terms, the brain is relevant to the continuation of the life of a given person by virtue of the fact that it realizes the functions linked to the concept of a person. The resolution of the conflict between the bodily criterion and the psychological criterion that invokes the functional characterization of the concept ‘person’ is not immune to difficulties. As Wiggins notes, the thorniest question in this context is whether this way of resolving the conflict conforms with condition (D. ix) of his formal theory of substances, which expresses the transitivity of ‘...is the same f as...’: If f is a substance concept for a, then coincidence under f must be a determinate notion, clear and decisive enough to exclude this situation: a is traced under f and counts as coinciding with b under f, and a is traced under f and counts as coinciding with c under f while nevertheless b does not coincide under f with c. (1967: 38)

(D. ix) is satisfied by the bodily criterion. But it is possible to present counterexamples to the functionalist account of ‘person’ in which a person a is viewed as the continuer of the ‘same life’ of a person b, though (D. ix) is not satisfied. In order to clarify this difficulty, Wiggins introduced a variation of Shoemaker’s Brown-Brownson case. Based on split brain cases, Wiggins suggests the following thought experiment: Suppose that we split Brown’s brain and house the two halves in two different bodies. The difficulty is that there is no inclination at all to suppose here that our ‘decision’ floats free of the interests which normally animate our application of the concept person. Since there is memory character and life in both brain transplants it precisely cannot... We are supposing that the transplanted persons, Brown I and Brown II, claim to remember exactly the same things, that they are equally intelligent, and that they are equally at home in their new bodies. In this case, where cerebral material is actually transplanted we cannot simply disregard their (claimed) memories. For we understand far too well why they have these memories. On the other hand if we say each is the same person as Brown, we shall have to say Brown I is the same person as Brown II. That is an inescapable part of what was meant by saying that each was the same person as Brown. But Brown I will have all sorts of experiences which Brown II will not. They will be in different places and have separate experiences from now on. And they will communicate interpersonally. (1967: 53)

Given Wiggins’s above functional characterization of the concept ‘person,’ if the brain is indeed the seat of the life of Brown, the transplanted persons Brown I and Brown II have equal rights to be Brown’s continuer. But in this case (D. ix) is not satisfied. In Wiggins’s view, this difficulty is not surprising. For, the predicate ‘...

 Controversies and the Metaphysics of Mind

remembers X-ing on such and such occasion’ is a multiple satisfiable predicate, and the predicate “...is causally essential to the performance of such and such functions” applies to stuff as it does to things (1967: 54). Therefore: The thought experiment with brain splitting forces us to conceive of circumstances in which we should have to think of consciousness or personhood as something which could belong derivatively in the category stuff, and it forces us to conceive that this stuff could be done up in separate parcels which could have additional separate histories of their own for such time as they were separate parcels, as well as having other bits of shared history. (1967: 54)

Wiggins suggests a way out of this difficulty. As he notes, What (D. ix) requires of any substantial sortal f is only that coincidence under f should exclude the possibility of splitting...It would be better, after a conceptual analysis of the essential and characteristic vital functions, to analyze person in such a way that coincidence under the concept person logically required the continuance in one organized parcel of all that was casually sufficient and causally necessary to the continuance of essential and characteristic functioning, no autonomously sufficient part achieving autonomous and functionally separate existence. This logically excludes splitting, and makes the right empirical questions arise at the right place. ‘Have we in this case transferred all that was causally sufficient and necessary?’ which involves both general questions and particular questions of causality. (1967: 55)

It is not clear, to say the least, how Wiggins’s analysis of the concept ‘person’ can overcome the difficulty raised by the demand to satisfy his condition (D. ix) in the face of his own thought experiment. The italicized condition in the above quote is meant to rule out the logical possibility of splitting as part of the concept ‘person,’ and thereby to distinguish ‘person’ qua sortal concept of individual substances from uses of this term in which reference to individual persons is reference to them as members of clones. If cases such as Wiggins’s thought experiment are conceivable, it is not clear whether the sortal concept ‘person’ is applicable to individual persons. One way of interpreting Wiggins’s argument is to view it as separating between the a priori logical analysis of the sortal concept ‘person’ and the empirical questions that concern the applicability of the concept ‘person’ to actual cases. In light of this interpretation, Wiggins conceives of his thought experiment as relevant to pursuing the right logical analysis of the sortal concept ‘person’ and not to the question of whether ‘person’ is applicable to Brown I and Brown II. Clearly, if his thought experiment is generally relevant to human persons, Wiggins seems to leave us with no answer to the question of whether there are human persons, if ‘person’ is a sortal concept.4 4. Wiggins’s position gradually changed in his later writings on this topic. In Sameness and Substance (1980), he abandoned the idea that “there was a kind of entity whose nucleus was a

Chapter 8. From transcendental arguments to revisionary metaphysics 

Wiggins’s contribution to the debate on identity and personal identity was significant in more than one respect. As in Shoemaker’s, Strawson’s, and Williams’s cases, Wiggins’s analysis of personal identity aims to resolve the problem that is created by the two supposedly incompatible criteria by means of which a person’s identity is determined – the psychological criterion and the bodily continuity criterion. Apparently, Wiggins’s attempt to resolve the conflict parallels those of Strawson and Shoemaker. Yet, by introducing new logical and functionalist considerations into the debate, Wiggins’s arguments revealed the gaps in Shoemaker’s argument. As we shall see below, the connection between the functional analysis of mental states and the concept of personal identity significantly affected the later developments of the debate. This idea could be easily integrated with the novel philosophical and scientific theories of the mind that gradually emerged in the 1960s through the works of Putnam, Armstrong, Lewis, Fodor, and other philosophers of mind.5 Yet, the optimistic expectation that this account would be able to resolve the conflict was bound to be dampened by the subsequent developments of this debate. 2. Shoemaker: Quasi-memory and revisionary metaphysics In October 1970, Shoemaker published two papers: “Wiggins on identity” (WI) and “Persons and their pasts.”(PP) The first paper was a detailed review of Wiggins’s ISC. In the second paper, Shoemaker abandoned his earlier position, and developed and defended a novel, neo-Lockean account of personal identity that based the concept of personal identity on that of memory – on the special access that persons have to themselves qua themselves in considering themselves as “the same thinking being, in different times and places.”6 7 Shoemaker’s revised

vital organ, the brain, which organ could be differently attired at different times in different bodily accouterments.” (2004: 604). In Sameness and Substance Renewed (2001), he completely abandoned this position. In this book and related writing, his position is that a person is a human being. In his contribution to the exchange with Shoemaker in the Monist (2004a: 605), he maintains “Brownson is, at best, a stump, a relic, a reminder, a residue, a remnant, a vestige, a scrag, or a layered offshoot.” He is not identical to Brown. 5. In his later writings, Wiggins sharply criticized this approach to personal identity. See in particular Wiggins 2004a, 2004b. 6. Locke 1975: 335. 7. In his later writings, Shoemaker bases his account of personal identity on the psychological continuity that is far richer than “mere” memory. See in particular his contribution to Shoemaker and Swinburne (1984), and Shoemaker 1997, 2004a, 2004b.

 Controversies and the Metaphysics of Mind

account of personal identity is, among other things, based on a novel concept – that of quasi-memory (hereafter q-memory). As I will show below, Shoemaker’s Brown-Brownson thought experiment, Wiggins’s thought experiment, and the logical gaps that Wiggins revealed in Shoemaker’s earlier account apparently motivated the introduction of q-memory into the debate. Shoemaker’s q-memory is consistent with the background of the relevant sciences that deal with the human organism. But q-memory, in the context it is presented, is a product of philosophical deliberation that is motivated by a philosophical debate. It was introduced in order to solve a philosophical problem, that is, to resolve the conflict between the bodily continuity criterion and the memory criterion that are implicit in our ordinary self-conception. Yet q-memory was a novel concept that ventured beyond common sense. The arguments that introduced it into metaphysics, therefore, exemplify revisionary metaphysics, similar to the introduction of transcendental idealism as a solution for the conflict of reason within itself (though clearly less radical than Kant’s move). The new concept introduced by these arguments and the novel account of personal identity that they motivated remodeled the relevant space of real conceptual possibilities. These theories invited new questions, and motivated new controversies. In the present section I will first examine some of Shoemaker’s arguments in WI that are relevant to the present concern. I will then examine his novel account and the role of q-memory in it. 2.1

Wiggins on identity

In WI, Shoemaker’s assessment of Wiggins’s accounts of identity and personal identity goes beyond an examination of Wiggins’s main themes. Although his assessment emphasized the importance of Wiggins’s theory, it also aimed to improve on it by pointing out what, in Shoemaker’s view, are the gaps that it contains. The attempt to avoid theses gaps turned into a full-blown novel theory of personal identity in PP. As Shoemaker notes, Wiggins’s account of putative cases of relative identity “overlooks what seems to me the most plausible sort of case.” (1970: 530) The case that Shoemaker refers to in this context is the relation between a statue and the piece of bronze that coincide at a given temporal interval. Say that one makes the following assertion: 1. The statue on my table and the piece of bronze on my table are one and the same thing. What meaning should be assigned to ‘are’ in (1)? It cannot be assigned the meaning of ‘...is identical to...’. The statue could cease to exist without the piece of

Chapter 8. From transcendental arguments to revisionary metaphysics 

bronze ceasing to exist. As Shoemaker notes (1970: 531), Wiggins distinguishes between three senses of ‘is:’ (a) the ‘is’ of predication; (b) the ‘is’ of identity; and (c) the ‘is’ of constitution. But it appears that none of these senses fits the relation between the statue and the piece of bronze asserted in (1). The sense that seems to best fit their relation in (1) is (c). Nevertheless, (c) means ‘composed of’ as in ‘the jug is a collection of pieces of china.’ But it seems improper to hold that the statue is composed of a piece of bronze: While the statue is composed of bronze, and of a particular portion or quantity of bronze, it does not seem right to say that it is composed of a piece of bronze – and still less is it right to say that the piece of bronze is composed of a statue. (1970: 531)

The relation asserted in (1) is symmetrical unlike that of ‘composed of’ or ‘constitutes.’ The correct description in this context seems to be that the relation that holds between the statue and the piece of bronze is that of ‘being composed of the same matter as.’ This relation is easily confused with that of identity. In Shoemaker’s view, although Wiggins does not discuss this relation in dealing with the principle of relative identity, he does imply that this is the relation between a person and her body. According to Shoemaker, this view ... has the great merit of reconciling two highly plausible doctrines that may seem on the face of them to be incompatible – namely, (a) the doctrine (recently associated with Strawson) that physical properties belongs to persons in nonderivative way (that is belong to the same subject as do states of consciousness) and (b) the view that it is logically possible for a person to have different bodies at different times. Doctrine (a) may seem to imply (what is contradicted by [b]) that a person is identical to his body, but Wiggins holds, quite rightly I think, that it does not imply this. (1970: 531–2)

Clearly, the two views are Strawson’s view and Shoemaker’s own view. In 1963, Shoemaker did indeed think that the relation between a person’s body and her capacity for self-knowledge was a necessary relation (in some sense of the term). Yet, his claim that first-person memory statements suffice to identify persons in contexts such as the Brown-Brownson case seems to be in conflict with Strawson’s position. The apparent incompatibility of the two positions acquired a new meaning in light of Wiggins’s criticism of the concept of relative identity. According to Shoemaker’s earlier position, the person’s identity can be determined by means of two criteria. The person may not have the same body in all the temporal intervals in which she may exist. As I noted above, Shoemaker’s earlier position seems to imply commitment to the notion of relative identity. But, this was not the only problem Wiggins’s critical argument revealed. As I noted earlier, Shoemaker’s resolution of the supposed conflict between the two criteria of personal

 Controversies and the Metaphysics of Mind

identity, namely, the memory criterion and the bodily criterion, was based on the assumption that most of one’s first-person memory statements must be true. Yet Wiggins’s thought experiment of the branching Brown is, in fact, a counterexample to Shoemaker’s contention. Assuming that true memory statements entail that the person who is remembering is identical to the person who had experienced the remembered experience or who had performed the remembered action, all of Brown I and Brown II first-person memory statements seem to be false.8 Both Brown I and Brown II are continuers of Brown. But Brown I is not identical to Brown II. Each of their conscious first-person assertions ‘I experienced E’ or ‘I did A,’ where the E was experienced by Brown and A was done by Brown, seem to be false. That most of one’s first-person memory statements must be true was in Shoemaker’s earlier position the link that connected the two criteria of personal identity. It was required in order to clarify the intelligibility of cases of body exchange. The loss of this link rendered unintelligible what for Shoemaker’s earlier position was apparently possible (on the basis of his Brown-Brownson thought experiment). 2.2

Persons and their pasts

In WI, Shoemaker does not address the difficulties noted above, and the need to resolve them. They are discussed in PP, which contains what is perhaps one of the most impressive confessions by a contemporary philosopher about the inadequacy of his or her earlier position.9 In the present section, I will examine the way Shoemaker’s revised account could be interpreted as a response to Wiggins’s arguments. In the last part of WI, Shoemaker espouses Wiggins’s functional account of personal identity with some revisions. Shoemaker and Wiggins both accept the conceivability of cases of body exchange. But the complication that Wiggins added to Shoemaker’s thought experiment led Wiggins to accept a principle that is false in Shoemaker’s view. Let us remember that Wiggins distinguishes between personal identity and bodily identity. Nevertheless, personal identity must, in his view, involve the continuation of the life of a person “in one organized parcel of matter.” The latter restriction is needed to rule out the possibility of branching. Yet, as Shoemaker notes, the restriction is apparently an arbitrary one. One could conceive of cases in which it is plausible to maintain that the continued life of a person does not involve “the same parcel of matter.”10 In Shoemaker’s view, 8. As we shall see in Chapter 9, Perry and Lewis suggest a plausible resolution to this difficulty. 9. See Shoemaker 2003: 19. 10. See in particular his variation of Wiggins’s thought experiment 1970a: 540–541.

Chapter 8. From transcendental arguments to revisionary metaphysics 

Wiggins’s view is supported by his unwarrantable denial of criteria of identity that involve a “no competing alternative” condition. These are criteria in which the condition that y is identical with x unless x has a competitor is part of the criterion of identity itself: ... we can offer the following as a logically sufficient condition of personal identity: If X’s memories and so forth at t stand in R to Y’s past life, and if there is no one other than X who at t or at some time prior to t has memories and so forth which stand in R to Y’s past life, then X is the same person as Y. (This presupposes that we already have a necessary condition of personal identity that can be used to establish that someone is “other than X;” but clearly we do have this, for clearly Z is other than X if X and Z occupy different places at the same time.) The criterion just suggested excludes splitting by making X’s having a certain relationship to Y sufficient for X’s being identical with Y unless X has a “competitor” (Wiggins’ term) for being identical with Y – that is, unless there is some Z (≠X) which also stands in that relationship to Y. (1970a: 542–543)

In Shoemaker’s view, Wiggins’s denial of this type of criterion derives from a mistake: Wiggins twice objects to this way of excluding splitting, but one of his objections rests on a mistake and the other is inapplicable to the present case. He twice invokes the principle that if “p” is grounds for “q,” then “p and r” (for any r) is grounds for “q,” and argues that if a certain fact is grounds for the truth of an identity statement then it must be grounds for it even in the case where there is known to be a “competitor” (see notes 38 and 47, pp. 69 and 72–73). But the principle invoked is false, unless “grounds” means “entailing grounds,” and in any case the objection is inapplicable if the statement of the grounds includes an assertion of the nonexistence of any competitors. (1970a: 543)

If the memory criterion p is grounds for the identity of X and Y, where there is no competitor Z, the memory criterion should also suffice for the identity of X and Y, in which case a competitor Z does exist, a context in which p is apparently insufficient for identity. Yet as Shoemaker notes, the principle invoked here is erroneous, unless ‘grounds’ means ‘entailing grounds.’ At any event, it is irrelevant as an objection to a criterion in which the requirement that there is no competitor to X is built into the criterion itself. As we shall see below, this view is central to Shoemaker’s revised account. Let us try to recapitulate the various steps in the debates examined so far. As Shoemaker observes (1970: 531–2), the two theories that emphasize the role of first-person knowledge of identity as a constitutive feature of personal identity – Strawson’s theory and his own theory – were equally plausible. The first theory underscores the presumption that physical predicates belong non-derivatively to

 Controversies and the Metaphysics of Mind

persons, while the second theory defends the plausibility of the presumption regarding body exchange. According to Shoemaker, in the earlier stages of the debate, these two theories appeared to be equally plausible and yet incompatible theories, a conflict that apparently has the characteristics of an antinomy. Wiggins’s theory aimed among other things to resolve this antinomy. His relevant contribution to the debate seems to possess the following features: On the one hand, his application of his logical account of identity to the question of personal identity could be interpreted as an implicit polemical argument that criticizes Shoemaker’s earlier account of body exchange cases. This argument has the features of a type B polemical argument discussed in Chapter 2. The argument grants the plausibility of the presumption regarding body exchange cases. Moreover, it introduces cases of body exchange that are in fact cases of splitting. Yet, his highly plausible argument against the notion of relative identity undermined the reasonability of Shoemaker’s earlier theory. Nevertheless, as noted above, Wiggins seems to leave his readers without a clear answer as to how the identities of the offshoots could be determined. Shoemaker’s response to Wiggins’s implicit argument did not consist in an attempt to defend his earlier theory, for example, by trying to defend the notion of relative identity as others did. His response was quite different. He abandoned his earlier account, although he did not endorse any of the other theories that dominated the debate since the early 1950s. Rather, as noted above, he introduced, or better yet, reintroduced a neo-Lockean account, that is, a psychological continuity account, into the debate. This was the same type of account that he had attacked in his earlier work. Two motives seem to underlie this development in Shoemaker’s views: the attempt to clarify the plausibility of the presumption regarding body exchange cases, and the attempt to provide an answer to the question regarding the conditions that constitute personal identity with regard to cases of splitting. Merely reiterating Locke’s earlier views did not suffice to attain these tasks. It required some significant conceptual innovations. In PP, Shoemaker presents his revised account of personal identity in four steps. The first step consists in the introduction of what, in Shoemaker view, is one of the main epistemic features of first-person conscious states of awareness, their being immune to error with respect to the term ‘I.’ A person who reports what she observes may be mistaken with regard to what she observes. But according to Shoemaker, she cannot be mistaken as to whether she herself is the person that at least seems to have perceptions of objects. First-person psychological reports are immune to the error of misidentification with respect to the term ‘I.’ In an earlier paper “Self-reference and self-awareness,”(1968) where this idea is first introduced, Shoemaker connects this type of immunity to error to the

Chapter 8. From transcendental arguments to revisionary metaphysics 

existence of psychological predicates that can be self-ascribed (in Strawson’s sense). In PP, the fact that first-person states of awareness are immune to error through misidentification with respect to the term ‘I’ is introduced for a different purpose. When one is conscious that one seems to observe an IPod on the table, one might be mistaken as to what one seems to see, or even regarding to whether one is seeing anything (one could be victim to an illusion). Yet, it is not possible to be mistaken with respect to the identity of the subject that at least seems to be seeing something, that is, with respect to what ‘I’ stands for. But how should one construe reference to oneself in first-person memory statements, say statements such as ‘I remember watching the demonstration in Rabin Square on November 4, 1995?’ Shoemaker’s earlier analysis of this kind of statements involves the following features: a. The remembering subject at least seems to remember an event E, or an action A. b. If the statement reporting the relevant kind of memory is true, we have excellent reasons to suppose that the remembering person is identical to the person that experienced E or that did A. c. Memory statements are not infallible. d. Knowledge of the identity of the remembering subject and of the subject that experienced E or that did A is not based on an act of self-identification that makes use of criteria of personal identity. In Shoemaker’s earlier position, the fact that most of our memory reports must be true formed the basis for reconciling two other facts: that memory statements are fallible, and that making them does not involve criteria of identity. According to Shoemaker’s earlier position first-person memory statements of the relevant kind serve as evidence for the identity of the remembering subject and the subject who had done or experienced what she remembers by virtue of the fact that most memory statements must be true. First-person memory statements were conceived as being generally immune to error due to misidentification with respect to the term ‘I.’ Wiggins’s complication of Shoemaker’s thought experiment renders this supposition false. Conceivable cases of fission are cases in which it seems plausible to maintain that most of the relevant first-person memory statements made by the two branches that result from the person split are false. Clearly, this massive mistake does not affect the immunity to error through misidentification of the ‘I’ that is conscious of herself at a given time, the person most of whose firstperson memory statements are false. But it does affect the general immunity to error of the identity of the ‘I’ that remembers and of the ‘I’ who had experienced the Es or had done the As. In PP, Shoemaker’s view is (2003: 28) that first-person veridical cognitive states that represent (from “within”) past experiences and past actions are not immune to error through misidentification with respect to the

 Controversies and the Metaphysics of Mind

term ‘I.’ In fact, they need not be memory states at all. Rather, as we shall see below, they could be q-memory states. When Shoemaker discussed the ordinary cases of immunity to error through misidentification with respect to the term ‘I’ in “Self-reference and self-awareness,” he distinguished between two types of errors related to first-person reports, namely, mistakes related to the referent of ‘I,’ and mistakes related to the objective content of one’s report. Consider (1)-(4) below: 1. I see a canary 2. I believe I see a canary 3. I seem to see a canary 4. I had the illusionary experience of seeing a canary. If I sincerely assert (1) and I don’t see a canary, (1) is false. But if ‘I’ is sincerely used in (the first-person mode) in asserting (1) (that is false), (1) may entail (depending on the context of the assertion) a true first-person statement such as (2), (3), and (4) in which the intended referent of ‘I’ is the same as the intended referent of ‘I’ in (1). The idea of the possibility of error regarding most of one’s memory statements has a different result in the contexts of Wiggins’s thought experiment. All of Brown I’s and Brown II’s memory reports of the form ‘I [remember that I] experienced E,’ ‘I remember [that I] did A,’ are false. The mistake is due to what the term ‘I’ stands for in these assertions, not due to what E or A stand for. Consider (5) below: 5. I remember eating breakfast yesterday If (5) is false by virtue of what ‘I’ refers to, it may entail in the context of branching case that (6) below is true: 6. Someone ate that breakfast (the breakfast to which ‘breakfast’ refers to in (5)) yesterday (the same time to which ‘yesterday’ refers to in (5)). The idea that underlies the introduction of q-memory is that in cases of branching persons, such as Wiggins’s case, (false) sincere memory reports such as (5) are sufficient grounds for true statements such as (6). In contrast to true memory statements, the truth conditions of true q-memory statements do not involve the identity of the person that q-remembers and the person that experienced or that did what is q-remembered. Clearly, this is a third-person characterization of cognitive states, not a first-person characterization. We may assume that when asserting sentence tokens of the same type as (5), Brown I’s and Brown II’s intentions are not to assert q-memory statements. Hence, their assertions are false. But this does not rule out the fact that belief states that are expressed by assertions similar to (5) can be false by virtue of what ‘I’ stands for and, nevertheless, provide reliable grounds for the truth of assertions similar to (6).

Chapter 8. From transcendental arguments to revisionary metaphysics 

In developing the concept of q-memory, Shoemaker distinguishes between two types of such states. All q-memory statements truly refer to past events, not merely to generic events, but to singular events and actions. All q-memory statements must satisfy the previous awareness condition. Yet without any further restrictions, this type of state is open to the possibility of confusing exact similarity with identity (Williams’s problem that is pointed out also by Strawson’s idea of massive reduplication). Moreover, assuming that knowledge of continuous temporal events, such as a stone rolling down the hill, involves memory states of the past phases of the event that are linked to the phase actually observed, an unrestricted cognitive capacity for q-memories may connect the earlier phase of an event e1 of type T with the later phase of an event e2 of the same type, even though e1 and e2 are two different particular events. The required restriction is a strict causal requirement that q-memories must satisfy: If, as is commonly supposed, casual chains must be spatiotemporally continuous, then if quasi-memory claims implied the satisfaction of this strengthened previous awareness condition they would, when true, provide some information concerning the location of the quasi-remembered events and actions. We would know at minimum that the spatiotemporal relationship between the quasi-remembered event and the making of the quasi-memory claim is such that it is possible for them to be linked by spatiotemporally continuous causal chain, and if we could trace the causal ancestry of the quasi-memory we could determine precisely when and where the quasi remembered event occurred. Thus if we construe the previous awareness condition of quasi-memory as including this causal requirement, it seems that a faculty of quasi-remembering could enable us to identify past events and to reidentify persons and things...without giving us special access to our own past histories. (2003: 34–5)

Quasic-memory (hereafter qc-memory) is the type of q-memory that satisfies the causal requirement. Shoemaker’s main point is that having qc-memories of past experiences and actions does not suffice for the identity of the person that qc-remembers the past experiences and actions, and of the person that experienced these past experiences, and did these actions in the past. Brown I and Brown II both qc-remember Brown’s past experiences and past actions. But the concept of qc-memory can serve as the basis for a non-circular definition of personal identity in terms of qc-memories. Shoemaker calls the relevant causal links between the past experiences and actions and the present states of qc-memories ‘M-type of causal chain.’ Yet as Wiggins’s thought experiment clarified, the M-type of causal chain can branch, and given that identity is a one-one relation no criterion that allows branching suffices for identity. Clearly, a definition of personal identity that uses qc-memory as a necessary but not sufficient condition is not circular. In contrast to his earlier position, Shoemaker’s additional condition is not the bodily

 Controversies and the Metaphysics of Mind

continuity criterion. Rather, a person X is identical with a person Y if X quasicremembers “from the inside” Y’s experiences and actions and if there is no Z (that is not identical with X) such that Z also quasic-remembers Y’s experiences and actions “from the inside,” that is, if there is no competing alternative to X for being identical to Y.11 Shoemaker’s concept of q-memory, and the account of personal identity that is based on it, are products of revisionary metaphysics, which was motivated by the polemical dialogues in which he was involved. By introducing a novel relevant alternative to this debate, it paved the way for further developments, as we shall see in the next chapter. The causal concept quasic-memory, the M-type of causal chain, and the definition of identity based on them, complied with basic intuitions regarding the objective world and mental reality. Yet, the impersonal nature of the q-memories upon which the definition of identity is based invited attempts to refute it.12 As we shall see in the following chapter, Shoemaker’s definition of personal identity raised at least three types of objections. One type of objection accepted Shoemaker’s quasic-memory, and extended the Q-strategy to all types of mental states, but denied that an account of the identities of persons could be based on it. This later became known as ‘reductionism,’ the position defended by Parfit and his followers. Another type of objection accepted quasic-memory and extended the Q-strategy to all types of mental states, but questioned Shoemaker’s ‘no-competing alternative’ condition of identity. The neo-Lockean positions defended by Lewis and Perry belong to this type. Finally, a third type of objection aimed to undermine the very idea of q-memory and with it the neo-Lockean definitions of identity, and recommended a refined version of animalism. This was basically the position defended by Wiggins (in his later writings) and his followers. These positions and variations of them more or less exhaust the possibilities discussed in the last three decades or so by analytic philosophers addressing the problem of self-identity. Most of them dispute Shoemaker’s contentions in PP. And yet, the changes that Shoemaker introduced to the debate in PP affected all three (conflicting) positions. In other words, Shoemaker’s arguments render his position a controversial relevant alternative to the optional accounts of personal identity.

11. See Shoemaker 1970: 42. 12. As Noonan notes (1989: 170), Shoemaker believes that one’s q-memories (“from the inside”) provide special access to one’s own past. Yet, as Noonan adds (ibid.), there are in fact two concepts of q-memories one of which differs from Shoemaker’s concept in this respect. The differences between the two concepts are important for an account of the conflicts between the neo-Lockeans.

chapter 9

Neo-Lockeanism, reductionism, and animalism The emergence of a new debate

1. Introduction In 1970, Bernard Williams published “The self and the future” (SF) that presented a puzzle regarding personal identity that could be called the antinomy of personal identity.1 The antinomy that Williams’s argument revealed consisted of two equally plausible conflicting presumptions. One presumption apparently supports the bodily continuity position, according to which bodily continuity is a necessary condition for personal identity; the other presumption supports the neo-Lockean position, according to which the memory criterion suffices to establish identity. The importance of Williams’s paper does not merely consist in uncovering the antinomy of personal identity. By emphasizing the significance of the ethical aspect of self-concern to the question of personal identity, his paper contributed another significant component to the new debate. This feature was ignored by the earlier accounts that primarily addressed the epistemological, the metaphysical, and the logical features of the identities of persons. Yet, the issue of self-concern and the question regarding the importance of identity soon became a vital feature of the chain of debates that addressed the problem of personal identity. Williams’s emphasis on self-concern in personal identity was a significant step in the process that led to the emergence of a new stage in the chain of debates on personal identity. The explicit topics of these debates differed from those of the earlier debates. The element of self-concern served as the basis for Parfit’s skeptical solution to the problem of personal identity that linked self-concern to the main features of Shoemaker’s neo-Lockean account of identity.2 The theory 1. Johnston (1987) is among those who explicitly responded to Williams’s puzzle. As Johnston notes (1987: 65), Williams’s puzzle is a “conundrum about personal identity which, on the face of it, seems to threaten both parts of the orthodox position.” A problem that has these features is similar to an antinomy. 2. Parfit’s novel theory was later called reductionism. Parfit developed his theory in detail in his book Reasons and Persons (1984). Since my intention here is to follow the historical process

 Controversies and the Metaphysics of Mind

developed in Parfit’s paper “Personal identity” (1971) used Shoemaker’s novel, neo-Lockean Q-strategy that impersonalized mental states. Shoemaker adopted the Q-strategy in order to avoid the classical charge of circularity regarding the Lockean definition of personal identity, and in order to clarify the possibility of imagined cases of body exchange and fission (Wiggins’s type of cases). Yet, Parfit’s account uncovered the implausibility of Shoemaker’s definition. The question it raised did not merely pertain to the logical adequacy of Shoemaker’s definition, but, also or even primarily, to whether the definition could capture what is significant and important to us from the perspective of self-concern. In Parfit’s view, what matters is survival, and identity is not what matters in survival. Responses to Shoemaker’s definition and to Parfit’s challenge are found in Perry’s and Lewis’s new neo-Lockean accounts of personal identity that could be regarded as an additional stage in this chain of debates. These accounts attempt to avoid Parfit’s aporia (regarding personal identity), though they differ significantly from Shoemaker’s theory. This stage in this chain of debates clarifies the nature of the symbiotic relationship between the metaphysical intuitions regarding the nature of persons and the logical intuitions regarding the nature of identity. In addition, it deepens the revisionary features of Shoemaker’s neo-Lockean account. I suggest that this circle of debates culminates in Chapter 6 of the first edition of Wiggins’s Sameness and Substance, where Wiggins attempts to incorporate the Lockean intuition within his animalist position. Wiggins’s animalism does not merely attempt to resolve the old controversies but, in addition, tries to steer the debates on the nature of personal identity back to the realm of common sense. This chapter is devoted to an interpretation of some of the debates from the 1970s which led to the present stage of the chain of debates on personal identity in the analytic tradition. A detailed examination of this historical process requires a far more extensive discussion than the one undertaken here. My hope is that my interpretation at least suffices to clarify and support the main claims defended in this book. In Section 2, I will present William’s argument. In Section 3, I will clarify the significance of Parfit’s theory as presented in “Personal identity” (1971) to the debate on identity. In Section 4, I will examine Perry’s neo-Lockean approach. Finally, in Section 5, I will examine Wiggins’s revised account of personal identity in the first edition of Sameness and Substance.3 in which new ideas emerge in the context of polemical exchanges, I will not discuss his mature position here. 3. The second edition of Wiggins’s book was published in 2001. In this edition, the chapter on personal identity contains additional revisions in Wiggins’s position. I will not discuss them in the present context.

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

2. Self-identity and self-concern Williams begins SF with an argument that seems to confirm the intuition regarding the possibility of body exchange. Williams uses Shoemaker’s Brown-Brownson case with one important revision. In Williams’s version (1970: 162), the fact that Brown’s brain is transplanted in Robinson’s body is not as significant as the fact that Brown’s brain contains information regarding Brown’s personal character and past life. Williams invites his reader to imagine a possibility in which the information encoded in the brain of A is recorded on some machine and then transferred to the brain of B and vice versa. Before the operation, A and B are asked to choose one of two possibilities: one possibility is that, after the operation, the A-body-person will receive painful treatment, whereas the B-body-person will be given $100,000. Alternatively, the B-body-person will receive painful treatment after the operation, whereas the A-body-person will be rewarded $100,000. A and B may choose whatever they wish as long as they obey the following condition: only one of the persons (the A-body-person or the B-body-person) will receive the pleasant treatment of a monetary reward, while the other will receive the painful treatment. Williams then examines several possibilities. Say that, before the operation, A chose for the B-body-person to receive the pleasant treatment (the reward of $100,000) after the operation. If, after the operation, the B-body person (whose brain now contains the information that was contained by the brain of the A before the operation) wakes up and receives the pleasant treatment, it seems reasonable to suppose that she would say that she had received the treatment of her choice. On the other hand, if, before the operation, A had decided that the A-body-person would receive the pleasant treatment after the operation, it seems equally natural to suppose that, when the B-body-person wakes up after the operation and receives the painful treatment, she would say that she had received what she originally wanted. Williams then examines different sorts of possibilities. Say that A is told that she would receive a painful kind of treatment at a given future time, though she is told that she will be in a state of complete amnesia when she actually receives this treatment. As Williams notes, given her egoistic attitudes, it seems plausible to suppose that A will fear the future treatment, even if she has reasons to suppose that she will have no memories of her (future) past self. But in this case, her beliefs regarding her identity are based primarily on the identity of her body. They seem to support the view that bodily identity is a necessary condition for personal identity, in contrast to the conclusion that apparently follows from the body exchange thought experiment. The idea of an amnesic future self could be used for the purpose of questioning the neo-Lockean contentions that are based on intuitions invoked by cases of body exchange by means of the following steps:

 Controversies and the Metaphysics of Mind

i. A is subjected to an operation which produces total amnesia; ii. amnesia is produced in A, and other interference leads to certain changes in his character; iii. changes in his character are produced, and at the same time certain illusory “memory” beliefs are induced in him; these are of a quite fictitious kind and do not fit the life of any actual person; iv. the same as (iii), except that both the character traits and the “memory” impressions are designed to be appropriate to another actual person, B; v. the same as (iv), except that the result is produced by putting the information into A from the brain of B, by a method which leaves B the same as he was before; vi. The same happens to A as in (v), but B is not left the same, since a similar operation is conducted in the reverse direction. (1970: 172)

It should be noted that the argument presupposes that a totally amnesic person does not cease to exist by virtue of her being amnesic. This seems to conflict with the main Lockean contention. Yet, one must observe that the Lockean intuition is invoked with respect to certain type of cases, that is, cases of body exchange. It is not at all clear whether the Lockean contention is equally plausible with regard to intuitions invoked with respect to cases of total amnesia. It seems rather plausible to maintain that identity is preserved in such cases, or for that matter, in cases in which the person’s character changes. A powerful intuition seems to support the supposition that a person does not cease to exist if she loses her memory, or if her personality undergoes dramatic changes. Williams’s first rhetorical goal is to convince his readers that this intuition is as powerful as the Lockean intuition. That the bodily criterion is necessary for personal identity is apparently warranted by the set of steps that represents the changes in A’s memories and her personality as changes that occur to A, that is, changes that leave A’s identity untouched. The fact that A has acquired memories and personal characteristics that are exactly similar to those of another person seems to be irrelevant to the question regarding the identity of A, that is, the question regarding the identity of the person that has acquired the new personality and the new memory-like states. If in cases of complete amnesia or in cases of radical changes of personality the person’s identity is preserved, there seems to be no reason to claim that it is not preserved in the above type of cases. Viewed from this perspective, the neo-Lockeans’ main contention appears to rest on the difference between what occurs to A in step (v) and what occurs to her in step (vi). Yet, as Williams clarifies, anyone who maintains that A of step (vi) is identical to (B) of step (v) seems to face a serious difficulty. In step (v), A and B are considered different persons. We may assume that the A of step (v) is identical to the A of steps (i-iv). The only difference between the A of step (v) and that of step (vi) concerns changes that occur to another person, that is, changes that occurred to B.

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

It is not difficult to see that up to this point Williams’s argument is apparently an argument against the neo-Lockeans. It is an instance of a type B argument discussed in Chapter 2. Claims that are assumed to be non-controversial (apparently commonsensical claims), which are accepted by both interlocutors, are used as premises in his argument. However, his argument entails a contention that is incompatible with the neo-Lockean position. Williams is clearly more attached to the bodily continuity view. Yet, it is interesting to note that in Williams’s view, his argument does not really warrant a version of the bodily continuity position. He believes that the differences between the two standpoints, between the bodily continuity view and the Lockean view, depend on the way one describes the situation at hand, that is, on how the thought experiment is contrived. In other words, in Williams’s view, the relevant intuitions seem to pull us in two different directions, one supporting the psychological criterion, and the other supporting the bodily criterion: Thus, to sum up, it looks as though there are two presentations of the imagined experiment and the choice associated with it, each of which carries conviction, and which lead to contrary conclusions. The idea, moreover, that the situation after the experiment is conceptually undecidable in the relevant respect seems not to assist, but rather to increase, the puzzlement; while the idea (so often appealed to in these matters) that it is conventionally decidable is even worse. Following from all that, I am not in the least clear which option it would be wise to take if one were presented with them before the experiment. I find that rather disturbing. (1970: 178-179)

I suggest that the puzzle related to personal identity mentioned by Williams in the passage quoted above is close enough to an antinomy in the modified Kantian sense. It is interesting to add that Williams’s argument weakens the assumption that the Lockean position emphasizes the first-person features of persons, whereas the animalist emphasizes the third-person features. This is indeed one of the significant results of his rhetorical move: Whatever the puzzlement, there is one feature of the arguments which have led to it which is worth picking out, since it runs counter to something which is, I think, often rather vaguely supposed. It is often recognized that there are “firstpersonal” and “third-personal” aspects of questions about persons, and that there are difficulties about the relations between them. It is also recognized that “mentalistic” considerations (as we may vaguely call them) and considerations of bodily continuity are involved in questions of personal identity (which is not to say that there are mentalistic and bodily criteria of personal identity). It is tempting to think that the two distinctions run in parallel: roughly, that a firstpersonal approach concentrates attention on mentalistic considerations, while a third-personal approach emphasizes considerations of bodily continuity. The

 Controversies and the Metaphysics of Mind

present discussion is an illustration of exactly the opposite. The first argument, which led to the “mentalistic” conclusion that A and B would change bodies and that each person should identify himself with the destination of his memories and character, was an argument entirely conducted in third-personal terms. The second argument, which suggested the bodily continuity identification, concerned itself with the first-personal issue of what A could expect. That this is so seems to me (though I will not discuss it further here) of some significance. (1970: 179)

This important observation is compatible with Shoemaker’s neo-Lockean transformation of his earlier position. Shoemaker seems to emphasize the third-person features of concepts such as memory and psychological continuity. According to Williams, when the question of personal identity is assessed from the viewpoint of one’s own self-concern, the difficulties related to the supposition that personal identity is not a decidable issue are different than those attached to the claim that the identity of other physical objects (clouds, ships, etc.,) is not a decidable issue. 3. Parfit and “what matters” Parfit’s paper “Personal identity” (PI) was published in 1971. In this influential paper, Parfit attempts, in fact, to resolve William’s puzzle by means of three steps.4 He first clarifies why some questions regarding the identity of a person cannot be answered in some cases. He then adopts Shoemaker’s Q-strategy and extends the impersonalizing account to all type of mental states (and not merely to memory). Finally, he undermines the importance of the question of personal identity. Following Williams, we may assume that personal identity is connected to our practical concerns, that is, to egoistic attitudes such as self-preservation, fear of death, fear and hope for future possibilities, and so forth. Yet, as Parfit aims to show, our interest in personal identity lies not in identity, but, rather, in survival. Although identity is a one-one relation, psychological continuity and psychological connectedness, the relations that matter in survival, are not one-one relations.5

4. Parfit does not explicitly refer to Williams’s puzzle. I suggest, however, that this is a reasonable interpretation of his views. 5. Parfit distinguishes between psychological continuity and psychological connectedness. Psychological connectedness is the direct relation between a q-memory and the q-remembered experience, or between the q-intention and the q-intended action. Although psychological continuity is a transitive relation, psychological connection is not transitive. See Parfit (1971: 20). However, both relations are important for survival. The connection between them supports the claim that there are degrees of survival.

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

The resolution that Parfit suggests to Williams’s antinomy is in a sense a skeptical resolution. In contrast to Shoemaker, his aim is not to remove the problem related to personal identity. At least in his 1971 paper, he seems to accept that the identity of a person is not a decidable matter.6 Nevertheless, he does not merely leave the reader with the puzzle related to this question. In Parfit’s view, resolving the puzzle of personal identity means weakening the belief that personal identity is what matters to us. This task is carried out by clarifying that what is significant to us is not identity but, rather, survival. Parfit’s argument in PI may mislead one to believe that he in fact endorses some version of the neo-Lockean approach. Yet, his intentions differ from those of the neo-Lockeans. In contrast to the neo-Lockean attempt to accommodate the idea of personal identity in the space of possibilities that includes Shoemaker’s and Wiggins’s thought experiments, Parfit in fact emphasizes what these neoLockeans attempts have missed. Let us grant, for a moment, that Shoemaker’s revised account of personal identity, which he presented in “Persons and their pasts,” is an adequate account that allows for cases of body exchange without conflicting with the logical features of identity. It should be noted, however, that Shoemaker’s account can hardly be viewed as a satisfactory response to Williams’s objections, that is, to the objections regarding the importance of identity. Wiggins’s case of fission, in which A’s brain is divided and transplanted in two distinct bodies B and C, are cases in which it seems plausible to maintain that B and C are both continuers of A. According to Shoemaker’s account, personal identity is not preserved in these cases. Yet, even if Shoemaker’s account is granted, the question of whether A is completely annihilated in such cases, or whether she somehow survives, raises a disturbing problem. If survival entails identity, the splitting of A’s brain into two equally potent halves seems to be equivalent to the death of A. But equating splitting with death must surely be wrong. For according to Shoemaker’s account, if only one half of A’s brain survives, say, in B’s body, A survives. Therefore, how can it be maintained that if the two parts of A’s brain survive (albeit in different bodies), A does not survive? It should be noted that the difficulty is not whether the logical features of identity are satisfied by these cases, but rather whether this way of solving the problem is faithful to the interest that we have in ourselves. If the question is asked by A (the person that is about to undergo the process of fission) from the perspective of her self-concern, then, surely, it seems absurd to suppose that in envisaging the possibility of splitting, she would regard it as equivalent to death. Each of her continuers has the same psychological features, the same memories, 6. According to the reductionist view advanced in Reasons and Persons (1984) eliminating persons from a canonical description of the world leaves all states of affairs untouched.

 Controversies and the Metaphysics of Mind

the same intentional attitudes about the future that she has. Each of them is connected to her past life with the normal sort of causal chains. There has to be a difference between complete annihilation and fission. Nothing can be self-identical and nevertheless exist as two individuals. But if persons qua psychological entities can be split into two distinct (not identical) persons, our interest in survival must differ from our interest in identity. It might be supposed that a position that emphasizes the significance of selfconcern, which attempts to free us from the presumed importance of personal identity on this very basis, would also emphasize the importance of self-consciousness in the account of psychological states and processes. Yet, this is not Parfit’s approach. As I noted above, he endorses Shoemaker’s Q-strategy, and extends the impersonal account to all the other types of mental states: intentions, desires, and emotions. Moreover, freeing the concept of survival from the concept of identity becomes feasible only if mental states are impersonalized (given some restrictions that concern the coherence of our metal lives).7 As noted above, Parfit does not resolve the puzzle of personal identity. He does not offer an account of the identities of persons, but rather weakens the belief regarding the importance of being identical, that is, of the ethical significance of the question of self-identity. Parfit’s arguments marked a new stage in the debate on personal identity. It is important to note that his arguments could easily be interpreted as arguments against the neo-Lockeans and as belonging to the type B arguments discussed in Chapter 2. Parfit’s argument presupposes the possibility of body exchange. This already seems to separate him from Williams’s earlier position. Parfit also adopts and even extends Shoemaker’s impersonalizing strategy in his account of mental states. But by connecting what is important in survival to the question of personal identity, Parfit’s arguments emphasize the difficulties related to Shoemaker’s neo-Lockean account of personal identity. This argument, which begins by accepting the basic contentions of Shoemaker’s neo-Lockean solution and which clarifies the unreasonableness of this solution by emphasizing what is important in survival, also suggests an alternative solution that is incompatible with the competing positions. By means of this polemical argument, Parfit’s position became one of the main relevant alternatives in the discourse on identity. As several philosophers have noted, Parfit’s position is apparently similar to that of Hume.8 But this should not prevent us from appreciating what was so novel in Parfit’s contribution to the debates on personal identity. I suggest that the novelty of his contribution does not consist only in the core themes of his position, 7. A’s survival on the basis of her being psychologically continuous with {B,C,D...} allows multifarious forms and structures. See Parfit 1971: 19–23. 8. See for example Shoemaker 1997: 138–139.

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

but also in the fact that he used a type B argument in the particular context in which he used it. This argument was persuasive in clarifying the relevance of his contentions, which are part of the core set of his position, to the core set of the competing positions. 4. Perry; identity and “what matters” Shoemaker’s (revisionary) neo-Lockean account of personal identity and Parfit’s (revisionary) approach to the question regarding the importance of personal identity soon became the topic of a new controversy. The notable difficulty related to Shoemaker’s account, the fact that it seemed to be unable to clarify the sense in which Brown I and Brown II were both Brown before the operation was, as I showed above, significant for Parfit’s account. In Parfit’s view, that identity is a one-one relation and that cases of fission and fusion are real possibilities undermine the feasibility of a philosophical account of personal identity. However, Shoemaker’s ‘no other alternative’ condition, which is essential to his definition of personal identity was also counterintuitive for reasons related to the logical concept of identity. The question faced by those supporting a neo-Lockean account of personal identity was whether it was possible to provide an account of identity that differed from Shoemaker’s account which could also serve as the basis for a satisfactory response to Parfit’s queries. In “Can the self divide?,” (1972) John Perry’s main intention is to defend a neo-Lockean account of personal identity that resolves the concerns related to Shoemaker’s theory and Parfit’s skeptical solution. According to Perry, the neoLockean approach can be freed from Shoemaker’s ‘no other alternative’ condition as far as the logical features of the concept of identity are concerned: Shoemaker would deny that either Brown-Johns or Smith-Johns was Johns. At least part of Shoemaker’s motivation for denying this is his belief that it involves “modifying the usual account of the logical features of identity.” That I deny; this paper is an argument for that denial. (1972: 465, footnote 6)

In this paper and related works published in the 1970s, Perry’s main intention was to support a theory of personal identity that could clarify why Brown I (who roughly corresponds to his Brown-Johns) and Brown II (that roughly corresponds to his Smith-Johns) had both been Brown in the past. Interestingly, in Perry’s view, this account does not refute Parfit’s contention that identity is not what matters in survival.9 In fact, Perry’s views are close to Parfit’s views regarding the 9. See Perry 1976.

 Controversies and the Metaphysics of Mind

importance of identity. What he wishes to deny is that the concept of identity is not applicable to persons that survive fission and fusion. Taken together, these claims reflect Perry’s belief that the question regarding what is important in survival cannot serve as the basis for an account of personal identity, in contrast to what Parfit seems to assume. As we shall see, by disconnecting the two issues, Perry is also able to resist Parfit’s recommendation to give up our ordinary concept of the human person. What seems to underlie this suggestion is the belief that what is important in being self-identical depends on the metaphysical concept of personal identity, a view that, according to Perry, begs the question. Perry’s account at least attempts to bridge the gap between common sense and philosophical theories of personal identity. Yet, as we shall see, his account disconnects the metaphysical and logical investigations from the commonsensical view even more than before.10 Splitting cases such as Brown → Brown I/Brown II naturally lead to the following contentions: (a) Before the operation Brown was a self-identical person; (b) Brown cannot exist as two persons, and (c) after the operation there are two distinct (self-identical) persons, each of whom is psychologically continuous with Brown, and therefore has the right to claim that he has done and experienced all the things that Brown has done and experienced. How can (a)-(c) be true together? The claim that Brown was a self-identical person before the operation seems to support the contention that fission is equivalent to the death of (the selfidentical person) Brown. One could cling to the connection between identity and survival, if either Brown I or Brown II dies. Yet, as shown above, this leaves the connection between identity and survival unexplained in cases such as that in which Brown survives as Brown I and Brown II. How can it be true that Brown does not die in fission, that before the operation Brown is a self-identical person, that he survives as two persons, and also, that the link between survival and identity is not broken? Perry’s main intention is to resolve this apparent conflict by introducing a novel account of the case at hand. His argument begins by drawing a distinction between the identity of an object and the unity relation connected to the concept of an object. Mastering the concept of ‘table’ involves knowledge of two types of unity relations: spatial unity and temporal unity. For example, the ability to apply the concept ‘table’ to a given object involves the knowledge that the left and right parts of the object, as well as its legs, are all parts of a single object. This type of 10. Lewis (1971, 1976) developed a theory similar to Perry’s theory of personal identity. However, there are some differences between Lewis’s and Perry’s theories of personal identity. Their accounts differ with regard to the importance of identity. As I noted, while Perry’s views are close to Parfit’s views, Lewis believes that his theory of personal identity can preserve the commonsensical link between identity and survival. I will not discuss Lewis’s views in this book.

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

relation, which connects the table’s spatial elements, is the spatial unity relation. Similarly, mastering the concept ‘table’ involves the knowledge that tables may change, that is, that the same table that used to be blue in the past is now (after it was painted) green. This second type of relation, which connects the different temporal stages of an object, is the temporal unity relation. As Perry notes, one must be careful not to confuse the unity relations with identity. The logical features of the unity relations are not the same as the logical features of identity. It might be supposed that given that the unity relations constitute an object, they must share the logical features of identity. Yet, as Perry notes, this is mistaken. Suppose there to be two Siamese twins joined at the thumb. Now consider three thumbs a, b, and c (b is the shared thumb). There is a single body of which both a and b are thumbs. That is the (spatial) unity relation for human bodies holds between a and b. And similarly there is a single body of which both a and c are parts. But there is no single human body of which both a and c are parts. So if Rb is the spatial-unity relation for human bodies, a has Rb to b, b has Rb to c, but not-(a has Rb to c). Thus the spatial unity relation for human bodies is not transitive. (1972: 468–9)

Here one might wonder whether the non-transitivity of the spatial unity relation for human bodies does not lead to the breakdown of identity. However, as Perry notes, it does not jeopardize identity, because there is no single body which has both a and c as parts. Although Perry does not explicitly say so, his move seems to be a response to the reductio argument that presumably motivated Shoemaker to introduce his definition of personal identity. This argument has the following form: (1) (2) (3) (4)

It is not true that Smith-Johns is the same person as Brown-Johns. Smith-Johns is the same person as Johns. Brown-Johns is the same person as Johns The negation of (1) follows from (2) and (3) and the transitivity of identity: Smith-Johns is the same person as Brown-Johns. (1972: 469)

In Perry’s view, the mistake involved in this argument consists in the supposition that the name ‘Johns’ used in (2) and (3) refers to a person. But although in (2) and (3) ‘Johns’ refers to a person stage, which belongs both to Brown-Johns and to Smith-Johns, it does not refer to a person. In the context in which it is presented, the distinction between a person and a person stage, which is crucial to Perry’s account (and also to Lewis’s account), naturally leads to the idea that different persons can share stages or parts. Although this contention seems to be implied by Wiggins’s thought experiments, it conflicts

 Controversies and the Metaphysics of Mind

with other intuitions, such as the intuition articulated by Strawson that mental states owe their identity to the identity of the individual subject that has them. Perry’s account of personal identity constructs persons from person-stages. The truth conditions of sentences about the past and future properties of persons are stated “in terms of statements about the properties of person stages and the temporal unity relation for persons”(1972: 470) that he designates by ‘R.’ He further distinguishes between basic and non-basic properties of persons. Basic properties are properties that a person has at any time by virtue of events that occur at that time. Non-basic properties are properties a person has by virtue of events that occur at other times. Non-basic properties that a person has at any given time are a function of the basic properties that she has at other times. This method allows three different accounts of personal identity that analyze English (or any other natural language) as being either a ‘branch language,’ a ‘person stage language,’ or a ‘life-time language,’ respectively. According to the first possibility, persons are branches consisting of person stages, say, the person stages thought to be Brown or Brown I. It should be noted that Brown, Brown I, and Brown II do not together form a branch. A person has a given property if the branch with which she is identified contains a person stage that has that property. A name is assigned to person stages. It names a person if and only if there is a branch that contains this person stage. Sentences having the form of ‘N is F at t’ are true if and only if the branch named by N contains a person stage P that is F at t. If there is no such branch, if for example, there is more than one branch to which P belongs, the above sentence is false. In Perry’s view, the branch analysis is not an adequate account. According to this conception, there is no sense in which Brown was a single person before the operation who did all the things that Brown I and Brown II remember doing (1972: 472). How can it be maintained that Brown was a single person before the operation and also that, nevertheless, ‘Brown’ does not name a single person after the operation? The two other options seem to be able to satisfy both conditions. However, for that purpose one is required to view the role of temporal adverbs in sentences such as (1) as qualifying what ‘Brown’ refers to: 1. After the operation, Brown will be in room 102. Although ‘Brown’ in (1) is adequate, ‘Brown’ that is asserted in (2) after the operation is improper: 2. Brown is now in room 102. In other words, the adequacy of a name depends on the time it is used to assert a sentence. Names do not timelessly name a person. Whether a name is proper or improper depends on the temporal adverbs involved in the asserted sentence.

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

According to the person-stage analysis, names are assigned to person stages. According to this theory, the ‘is’ of past tense statements such as (3) below does not express strict identity “but simply the temporal unity relation R.” (1972: 476) 3. The person you danced with last night is the person sitting on the sofa. In terms of strict identity, (3) is, in fact, false. According to this approach, if the name ‘Hilda’ is supposed to name the same person you danced with last night and the person who is now sitting on the sofa, ‘Hilda’ is ambiguous. Names of persons are systematically ambiguous. As Perry notes, though the stage language analysis seems to give the mentalist (the neo-Lockean) what she wants, according to this approach “being personally F at t is like being married to a janitor; it’s not being the janitor, but having an intimate relation to someone who is.” (1972: 479) Perry’s alternative is the lifetime language. A person stage determines a lifetime. A lifetime includes all and only the person stages that are R related to the person stage P that determines it.11 Names are assigned to person stages. Their primary references are the lifetimes determined by the person stages that they name. Hence, the lifetime named ‘Brown’ (before the operation) is the lifetime that has the Y-shape structure, which includes both Brown I and Brown II stages. But the lifetime named ‘Brown I’ is not identical to the lifetime named ‘Brown’, or to the lifetime named ‘Brown II.’ In other words, although person stages determine one lifetime to which they belong, they can be parts of more than one lifetime. Given the primacy of person stages in the construction of persons, lifetimes are determinable at t only if the person stage that determines them (say, Brownson I being in room 102 after the operation) exists at t. Hence, before the operation the lifetime named Brown I is not determinable. Perry believes that the lifetime analysis is better than its competing alternatives. It appears that Perry is also able to somehow preserve the connection between identity and survival. The lifetime named ‘Brown,’ the lifetime that has the Y-shape structure, is a self-identical entity that survives the operation. It is not identical to the lifetime named ‘Brown I’, or to the lifetime named ‘Brown II’ (each of which is self-identical). Yet, as I noted above, one of Perry’s motivations was to avoid the baffling results of Shoemaker’s definition. It appears, however, that Perry’s account trades Shoemaker’s bewildering results for a position that involves contentions that are no less perplexing. According to this position, 11. The relation R is the relation that constitutes the temporal unity relation for persons. R in Perry’s intended sense is a version of Locke’s “same consciousness” relation that is interpreted by some of the neo-Lockeans as the memory relation. The R relation is interpreted in two different ways by Grice (1941), and Quinton (1962). As Perry notes (1972: 468, footnote 7), both positions are compatible with his intentions.

 Controversies and the Metaphysics of Mind

distinct persons can share stages, a person can survive as a Y-shaped lifetime, and the adequacy of names depends on the time when they are used. Each of these contentions violates the pre-theoretical, commonsensical intuitions that we connect to the concept ‘person.’ They raise the question of whether this type of account can rightfully be viewed as a conceptual analysis of our concept of a human person, or whether a new concept of personhood is being constructed. As I noted above, Perry’s views are, in fact, close to Parfit’s view regarding what is important in survival. He thinks that “the importance of identity is derivative.” (1976: 84) The importance of identity depends on the interest in personal projects. These interests seems to be sustained even in cases in which the continuer is an exact copy of myself, that is, in cases I have reasons to believe that my continuer is not identical to me but is rather a duplicate produced by some sophisticated star-track like process on the basis of information extracted from me. According to Perry, the importance I attribute to properties and projects that apparently seem to be linked to me (qua self-identical entity), which is related to my reasons for action, is not affected in such bizarre cases. My reasons for acting in accordance with these motives are not affected by the fact that the results of these actions do not affect my own life, but rather the life of my duplicate!12 Nevertheless, Perry believes that the importance of our ordinary concept ‘person’ is not affected by these considerations. According to Perry, the stages of a single human being are H-related, and the stages of a single person are P-related. P-relations are the unity relation for persons. H-relations can be distinguished from P-relations. Although the entities that are related by means of P-relations “are stages of human-bodies ... there is nothing about the way we introduce the P-relation which makes it necessary that human-stages which stand in the P-relation invariably stand in the H-relation, or vice versa.” (1976: 71) Yet, “in normal circumstances, human-stages are P-related if and only if they are H-related.” (ibid.) This does not rule out the conceivability of abnormal cases. The normal cases explain our habit for caring about ourselves, the importance of which is not based on the metaphysical concept of personal identity, but rather on its evolutionary worth: “our having such private projects can be explained by what I take to be the correct theory of personal identity, as a result of habit, and the demands of evolution.” (1976: 80) Let us sum up the present section by noting that although Perry’s intention is to bridge the gap between what is, in his view, the best metaphysical account of personal identity and our ordinary concept ‘person,’ his account deepens the gap between the two. Perry’s account of topics such as self-interest and personal projects does not depend on his account of identity. This theory uses intuitions related to our concept of a 12. See Perry 1976: 83.

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

human person. But it does not merely analyze our concept. Rather, it seems to revise it. As Parfit observed, Perry’s account is in fact a variation of his own account.13 5. Wiggins’s animalism In 1980, David Wiggins published Sameness and Substance, one of the most influential books in contemporary metaphysics. In the final chapter of this book, he readdressed the topic of personal identity. It contained significant revisions of Wiggins’s earlier views that were nevertheless faithful to his earlier motivations. There is more than one respect in which Wiggins’s main arguments in Chapter 6 of Sameness and Substance could also be considered a response to the neo-Lockeans’ arguments, indeed, to the controversy on personal identity as a whole. It is quite clear, I suppose, from what has so far been established that the debates between the neo-Lockeans and their opponents are not debates between materialists and substance or property dualists. All parties are materialists of some kind. The question these debates addressed concerns the importance and significance of the notion of a stream of experiences and memories to the identity of a person. Theories of bodily continuity (such as Williams’s early views) apparently denied that the continuous stream of psychologically connected states could serve as the basis for a definition of the identities of persons, whereas the neo-Lockeans believed that some relation between q-memories (or similar mental entities) makes such a definition possible. In Sameness and Substance, one of Wiggins’s intentions was to separate between the importance for personhood of the stream of connected mental states and its relevance for the identity persons: ...what seems not merely suspect, however, but almost inconceivable is the suggestion that this sort of mental connectedness is simply no part of the concept of a person... The pretheoretical conception of mental connectedness is an intuitive idea that we bring to philosophy... (1980: 151)

However, as Wiggins adds: The hard question is not the question of its relevance to philosophy but of just what, if anything, it implies about the nature of personal identity. (1980: 151)

In Wiggins’s view, distinguishing between these two topics is vital to the resolution that could be offered for the controversy that separates the neo-Lockeans from their enemies: 13. See Parfit 1976: 101–2.

 Controversies and the Metaphysics of Mind

Indeed I shall urge in the last section of the chapter that the issue of mental connectedness, considered as a mark of what it is to be a person, has to be separated from the main issue of identity and survival, and that, even though experiential memory is one component in an inner nucleus of conceptual constituents of what it is for a person to continue to exist or perdure, there is no strictly necessary or sufficient condition of survival that we can formulate in its terms. I hope that it will be the result of this separation, not only to disrupt the even flow of controversy between friends and enemies of the Lockean conception, but also to refine our general understanding of the intimate but complex relations of the notions of sameness and substance. (1980: 151–2, my italics Y.S.)

I suggest that the “even flow” of the controversy between the neo-Lockean and their enemies is close enough to a conflict with the features of the revised notion of a Kantian antinomy. Wiggins, in other words, attempts to overcome the antinomic state of the relevant philosophical discourse on personal identity. As we shall see, the resolution offered is the typical positive resolution of an antinomy. It aims to resolve the conflict by clarifying why the grain of truth in each of the conflicting positions could be preserved without entailing the conflict itself. According to this solution, although the stream of the variety of mental events and processes that are mentally connected by means of the memory relation are essential features of personhood, no necessary or sufficient conditions of identity can be based on these features of persons. Wiggins’s aim in this chapter could therefore be interpreted as an attempt to resolve Williams’s antinomy, to which Parfit responded. Nevertheless, the resolution that he offers for it differs significantly from Parfit’s skeptical solution. Wiggins’s resolution begins with an examination of Butler’s classical charge against the Lockean views, that is, the problem of circularity. As he aims to demonstrate, the charge of circularity begs the question. It could be rephrased and considered a charge of non-effectiveness (1980: 160). Yet, in Wiggins’s view, the neo-Lockeans can also respond to this charge. As we shall see, Wiggins believes that the neo-Lockean view is mistaken, although this mistake differs from what both these charges maintain. To remind the reader, according to the circularity objection, memory presupposes personal identity and therefore cannot be part of the definition of personal identity. The simple version of this objection can easily be responded to. The objection maintains that ‘A remember A’s X-ing’ is the canonical and equivalent form of ‘A remembers X-ing’ (1980: 153). Yet, as Wiggins observes: ... so far from ‘A remembers X-ing’ having ‘A remembers A’s X-ing’ as its canonical and equivalent form...the two locutions are importantly different in both construction and sense. (1980: 153)

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

The more promising version of the circularity charge is expressed in the following passage: Suppose P claims to remember planting a fig tree. We cannot settle the truth of his claim that he remembers planting the fig tree until we can establish by some criterion or principle independent of memory whether P is identical with the person (if any) who planted it. But then memory is doomed always to bring too little too late to determine any identity-question. For we must already have settled the identity-question by other means of determination before memory is allowed on to the scene at all. (1980: 161)

Clearly this objection parallels one of Shoemaker’s main contentions in SelfKnowledge and Self-identity (1963) according to which memory requires bodily criteria. Wiggins’s formulation of this objection continues as follows: If the memory condition is an extra condition applied after it is already fixed by some criterion K what we are to trace through space and time, then either it contradicts K, or at best it restricts (determines some subkind of) the kind associated with K. Suppose K goes with animal, for instance. Then the memoryrestriction may define self-conscious animal. But this’, the objector will say, ‘is going to create a very nasty situation for you. It is going to be possible for the victim of the car crash who loses his memory to be the same animal as the patient who walks out of hospital to start a new life but not the same person – even though both are persons. (1980: 161)

The argument presented in the above objection apparently points out the difficulties related to the concept of identity for a position that wishes to connect the memory criterion to the bodily criterion. If memory or self-consciousness is an extra criterion that restricts something determined in advance by some criterion K, say, the bodily-continuity criterion, the two criteria may conflict in a way that is incompatible with the concept of identity. Yet, according to Wiggins this objection can be responded to as follows: What is this other and precedent principle of criterion K? It will be replied: bodily continuity. But on one reading of ‘body’ (interpreting it as standing for abstract determinable material substance) the suggestion is in no way at variance with the proposal that is being explored. And on the other reading (‘body of a human being’) the respondent will have to explain how he is to avoid a conceptual obligation to count among present persons Jeremy Bentham and even the Pharaohs...Bodily continuity (in the second sense of ‘body’) is not enough without life. And surely sufficient exercise of the capacity for recall is conceptually connected with a particular mode of activity, and a whole cluster of vital functions and faculties of which Locke is entitled to argue that memory is one member. (1980: 162)

 Controversies and the Metaphysics of Mind

The question that is addressed here concerns the nature of K, namely, the criterion of bodily continuity. According to one interpretation, K stands for the abstract determinable material substance, an interpretation that is at variance, in Wiggins’s view (for reasons that will be clarified only later), with the option that is being criticized. According to the second interpretation, K refers to the body of a human being. But in that case, it is not clear why mummies are not persons. As Wiggins observes, if the first interpretation is endorsed, memory could be conceived as a crucial part of the continuity principle that we choose for persons without entailing that “it determines identity questions autonomously.” (1980: 162) What the objection really had to confront was a thesis about memory as part and parcel with other vital functions and faculties characteristic of persons. But against that thesis it simply begs the question. (ibid.)

Wiggins’s argument against the charge of circularity accommodates the neoLockean intuition at least in one important sense. According to Wiggins, a position that views the criterion of bodily continuity as the spatiotemporal criterion that constitutes the identity of persons, that is, that views the psychological criterion as being merely further restrictions of an antecedently determined individual person (that is determined as such by means of the bodily continuity criterion), is misguided. Yet, Wiggins’s arguments do not appear to support an unrestricted neo-Lockean position. The neo-Lockeans assume that the relation involved in memory or psychological continuity can serve as the basis for a criterion of personal identity. But some remarks that are embedded in Wiggins’s arguments against the charge of circularity and the charge of non-effectiveness already clarify at this stage of his argument that the defects in the arguments against neo-Lockeanism (offered by those who endorse the bodily-criterion), do not suffice to establish neo-Lockeanism. What these remarks seem to hint at is that a third position is available. This position incorporates elements taken from both conflicting positions. It is freed from the supposition that the memory criterion is a further determination of an independently conceived animal, as well as from the supposition that the relation involved in memory or psychological continuity can alone determine personal identity. Needless to say, the response to the charge of circularity was essential for the neo-Lockean positions examined above. Shoemaker and Perry provided different versions of the neo-Lockean account of personal identity which meet the logical requirements of identity. Their accounts responded to the charge of circularity effectively. It seems that up to this point in his argument, Wiggins in fact joins the neo-Lockeans against the representatives of the bodily-continuity position. Yet, Wiggins’s criticism of the circularity charge does not merely reiterate the neoLockeans’ arguments. There are two important differences between his arguments

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

and theirs. The neo-Lockean arguments were shaped by a context that includes the conceivability of cases of body exchange, fission, and fusion. By virtue of the features of the context that was shaped by these imaginary and extraordinary types of cases, their non-circular definitions of personal identity required the impersonalizing Q-strategy that stripped particular memory-like states (and other kind of mental states) from the identities of the persons that have them. By contrast, Wiggins’s arguments do not presuppose a context that is already shaped by these types of conceivable cases, which apparently provide the strongest grounds for the Lockean and neo-Lockean positions. Consequently, Wiggins does not need to appeal to the impersonalizing Q-strategy in order to avoid the circularity charge. The fact that the impersonalizing Q-strategy, which was devised by Shoemaker and which was used by the other neo-Lockeans, is not required for the response to the circularity charge was significant, as we shall immediately see, for the argument that supported Wiggins’s solution. That q-memories and their kin were not required for a response to the circularity charge diminished the importance of the neo-Lockeans’ imagined cases. These were used primarily in order to emphasize the significance of the stream of experiences that are mentally connected by means of memory to the concept of personhood. This move apparently reflected Wiggins’s belief that one could maintain that the stream of experiences that are mentally connected by means of memory is essential for personhood without relying upon the bizarre neo-Lockean types of cases. Rather, these features could be known as essential for personhood on the basis of our commonsensical pretheoretical conception. Moreover, this move reflects Wiggins’s belief that if the attempts to reveal the significance of those aspects of personhood emphasized by the neo-Lockeans presuppose contexts that are infected by the bizarre neo-Lockean imagined cases, then these attempts are bound to distort the “real” significance of these features. In other words, this hidden layer of Wiggins’s argument operates on the contextual background of the neo-Lockeans’ arguments and theories. Wiggins’s response to the neo-Lockeans contentions culminates in Section 6 of Chapter 6. He compares two senses in which a person is said “to transcend his body, one correct, the other impossible.” (1980: 163) The correct sense is, in fact, a variation of Strawson’s idea: persons should be conceived as basic particulars distinct from their “mere” bodies. Indeed, this will turn out to also be essential for the animalistic alternative that Wiggins endorses. The absurd and impossible sense to which he refers in this context is the above discussed neo-Lockean accounts suggested by Perry and Lewis.14 In Wiggins’s view, the position that conceives persons 14. It is interesting to note that Shoemaker’s version is not addressed by Wiggins in this context.

 Controversies and the Metaphysics of Mind

as mereological sums of “aggregate of person-moments or person-stages” (1980: 168) is the best version of an unqualified neo-Lockean approach. But Perry’s and Lewis’s positions are seriously flawed. The point about such aggregates... ...is that distinct aggregates may have common constituents, and that such aggregates will coincide as one person if, and only if, they have all their constituents person-moments in common...The trouble is that they are also much too unlike people. How is a person who is going to divide to conceive of himself as two or three people? Nor is this the only or the worst problem. Anything that is part of a Lesniewskian sum is necessarily part of it... But no person or normal material object is necessarily in the total state that will correspond to the person – or object-moment postulated by the theory under discussion. Finally look at the contrast between the predicates of ordinary material objects and the predicates of a Lesniewskian sum. An aggregate of this kind has its relevant properties derivatively from the properties of its constituent moments. But at least half of the things we want to say about persons cannot be even tortuously explained in terms of the states at an instant of person-moments. Consider for example weak, clever, cowardly, a bad slip fielder, resolute, opportunistic, erratic, honest...It is true that (however unnaturally) we can see each of these as supervening on all other properties of those Lesniewskian sums... But this is not unproblematic. It seems to presuppose that we can say what a person-stage is before we come to say what a person continuant (i.e. person) is...(1980: 168–9)

Clearly, this negative appraisal of the prospects of a neo-Lockean mereological account of persons does not concern the question of whether these accounts meet the demands of logical or theoretical adequacy. Rather, the question is whether they are compatible with the commonsensical, pretheoretical concept ‘person.’ It should be stressed that although the neo-Lockean approach is to a certain extent a revisionary approach to metaphysics, the neo-Lockeans are not eliminativists. They wish their contentions to be at least compatible with the commonsensical conception. Clearly, an alternative that attributes an essential role to the features of personhood that are emphasized by the neo-Lockeans, which nevertheless does not conflict with other commonsensical features of personhood, as does the neo-Lockean approach, should be (rationally) favored over the neoLockean positions on the basis of their own grounds. This alternative is the solution that Wiggins offers: We shall come more easily at the truth, I believe, if we dismiss the bold answer and supplant it with a much more modest suggestion. I shall call it the animal attribute view. This sees person as a concept whose defining marks are to be given in terms of a natural kind determinable, say animal, plus what may be called a functional or (as I shall prefer to say) systematic component. Perhaps

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

x is a person if and only if x is an animal falling under the extension of a kind whose typical members perceive, feel, remember, imagine, desire, make projects, move themselves at will, speak, carry out projects, acquire a character, as they age, are happy or miserable, are susceptible to concern for members of their own or like species...[note carefully these and subsequent dotes], conceive of themselves as perceiving, feeling, remembering, imagining, desiring, making projects, speaking...have, and conceive of themselves as having, a past accessible in experience-memory and a future accessible intention...,etc. On this account person is a non-biological qualification of animal, and, potentially at least, a cross-classification with respect to zoological classification across the grain, so to speak, of the evolution-based taxonomy...According to this view, a person is any animal that is such by its kind as to have the biological capacity to enjoy fully the psychological attributes enumerated; and whether or not a given animal kind qualifies is left to be a strictly empirical matter. (1980: 171–172)

Wiggins’s alternative is animalism. Persons are animals, and their identity is the identity of the animal that they are. However, the psychological characteristics that are emphasized by the neo-Lockean as essential for the concept ‘person’ are not left out from Wiggins’s account. Persons are animals of whatever kind that can have these characteristics, by virtue of the kind of animal that they are. ‘Person,’ in other words, is a non-biological qualification of ‘animal’ that is potentially a cross-classification of ‘animal.’ The features emphasized by the neo-Lockeans are essential qua capacities to the kind of things that persons are. However, they do not separately determine the identities of persons. Wiggins’s overall argument in Chapter 6 of Sameness and Substance therefore includes the following steps: 1. The faults of the charges of circularity and ineffectiveness against the Lockean positions are reinterpreted in a way that diminishes the significance of the neo-Lockean thought experiments (and therefore also of the impersonalizing Q-strategy regarding mental sates). 2. A sub-argument that clarifies why the best available neo-Lockean account, which conceives persons as Lesniewskian sums, is presented. The argument clarifies why this position is not compatible with the commonsensical, pretheoretical concept ‘person.’ 3. A third alternative (which is incompatible with neo-Lockeanism) is presented. The third alternative can clarify the essential role of the features related to human persons that are emphasized by the neo-Lockean without entailing the undesired results entailed by their positions. This argument involves a sub-argument of type B. The neo-Lockean might grant Wiggins’s rebuttal of the circularity charge. She may also share Wiggins’s

 Controversies and the Metaphysics of Mind

commitment to the demands of the commonsensical conception of persons. But the purpose of Wiggins’s argument is to convince all parties, including the neoLockeans, that the best available account of persons, which is faithful to the reasons that motivate neo-Lockeanism, and which meets the logical requirements, is incompatible with their main core contentions. The chain of debates on identity and personal identity does not end with Wiggins’s Sameness and Substance. The neo-Lockeans’ type of thought experiment, which Wiggins’s argument left out as insignificant to the account of personhood, continued to attract the imagination of philosophers of mind. The animalistic account of personhood is apparently chauvinistic (to use Block’s (1978) terms). It rules out the possibility that, for example, HAL, the super-computer from Space Odyssey, is a person. Since it conflicts with intuitions that we have regarding these possibilities, Wiggins’s position is debatable. 6. Conclusion In this part of the book, I examined the evolution of the chain of debates on personal identity from the early 1950s to the beginning of the 1980s. I have discerned three stages in this chain of debates, each of which seems to provide the unresolved puzzles that the next stage attempts to resolve. Philosophical thought experiments that involve imagined extraordinary cases play a key role in each of these stages. Questions regarding logical topics are often invoked in these debates with regard to disputes that deal with some factual matters. As noted above, the debates on the nature of personhood and personal identity did not end with Wiggins’s Sameness and Substance. In a sense, the first edition of Wiggin’s Sameness and Substance marks the birth of a new stage in the chain of the debates that mainly involves three positions: neo-Lockeanism, reductionism, and animalism. All of them can be linked to earlier positions with which they share some of their core contentions: Neo-Lockeanism has an affinity with Locke, reductionism can be traced to Hume, and animalism seems to be related to Aristotle. But the relevance of each of these positions to the present debates cannot be clarified merely by identifying them with their ancestors. As far as the present condition of each of these positions is concerned, they are shaped by the arguments presented in this chain of debates that made them controversial relevant alternatives. After the publication of Wiggins’s Sameness and Substance, further attempts were made to resolve the antinomic conflicts of the previous debates. New philosophers joined the debate. New strategies and tactical devices were contrived and employed. The novel accounts suggested by some of these

Chapter 9. Neo-Lockeanism, reductionism, and animalism 

philosophers aspired to overcome both old and new antinomies, although often in conflicting ways. I will not discuss these here. I leave the reader with the task of determining how these views were motivated by the preceding stages of this chain of debates.15

15. In 1984, Parfit published his Reasons and Person, which appears to be the most significant defense of reductionism. Shoemaker (1984, 1997a, 2004a, 2004b) continued to defend and to refine his neo-Lockean theory of personal identity. Wiggins (2001) and his followers continued to defend animalism. Olson (1997, 2007), Rovane (1998), and Rudder Baker (2000) present three of the most interesting novel approaches to the question regarding personal identity that are, to be sure, conflicting approaches. They evolved out of the earlier debates.

Conclusion My goal in this book was to examine the role of metaphysical controversies in the metaphysics of mind. Clearly, a comprehensive account of the nature of the metaphysics of mind includes far more than what is revealed by an examination of the relevant metaphysical controversies. Yet, I hope the examination of some of these controversies in Parts Two and Three of this book has clarified why controversies are so significant for progress in metaphysics, especially in the metaphysics of mind. Metaphysical inquiries address the most fundamental features of reality. Establishing a metaphysical contention cannot merely be based on experiments or other empirical sources. The uncovering of the nature of metaphysical concepts by examining imagined cases seems to be an endeavor wholly confined to the realm of the individual’s rational and reflective conceptual capacities. Yet, though this layer of metaphysical inquiries is clearly essential to their nature, it is not the only layer. If metaphysicians aim to reveal “the commonplaces of the least refined thinking; and ...yet the indispensable core of the conceptual equipment of the most sophisticated human beings,” (Strawson 1959: 10) it is also reasonable to suppose that the “massive central core of human thinking...has no history – or none recorded in histories of thought.” (ibid.) But metaphysical inquiries are always carried out in particular discursive and cognitive contexts. One is presumably motivated to conduct them because of the aspiration to reveal eternal truths that are not traceable to features of human nature, which do not depend on discursive contexts or cognitive contexts. And yet metaphysical inquiries are human affairs. They are at least partly shaped by the history of this discipline, even if they are not wholly determined by it. At each historical point, several conflicting metaphysical options regarding a given topic are supported by persuasive arguments. Establishing one’s own position must therefore be connected to criticizing relevant alternatives. My main intention in this book was to clarify what underlies this claim. As the chains of controversies discussed in Parts Two and Three clarified, it is impossible to grasp the significance of contentions made by philosophers from a later stage without considering how these contentions are related to puzzles and antinomies that preoccupied the champions of the earlier stages, the resolutions they offered for these puzzles and antinomies, and the new polemical conflicts

 Controversies and the Metaphysics of Mind

that they motivated. In my view, this is the most significant feature revealed by the controversy-based approach to metaphysics. Before concluding this book, I wish to call attention to two issues not discussed in it which seem to be relevant to its main topic. Firstly, my approach to the study of controversies in metaphysics and their role in the pursuit of metaphysical knowledge was confined to a unique type of controversy, namely, controversies between recognized experts of a given philosophical genre in metaphysics recognized as such by the research community to which they belong. In this book, I did not examine the nature of metaphysical controversies between philosophers belonging to “radically different” philosophical genres, approaches, or cultures. Similarly, I did not examine the nature of controversies between experts and those who have not yet been recognized as experts. The sociological and political character of the background of metaphysical controversies and the reasons that motivate certain polemical moves clearly include many more issues than those I have undertaken to discuss here. The positive epistemic role of controversies in metaphysics is more easily revealed if one focuses on controversies between recognized experts of a given philosophical genre. Also, the type of polemical exchange examined here allows one to bracket those factors that relate to human nature and human culture in order to point out more clearly what seems to have been missed by the current discourse, that is, that controversies are essential to the practice of metaphysics in the positive sense. Yet, the fact that some metaphysical polemical exchanges allow for this type of bracketing by no means entails that the other levels of the metaphysical discourse – the various levels of the sociological, political, and cultural nature of human beings – are irrelevant to the content of metaphysical contentions and metaphysical polemics. Nor is it implied that controversies have only a positive role in the pursuit of metaphysical knowledge. The topics that were not discussed in this book can be separated to a certain extent from those discussed in it. But they too deserve in-depth examination. Secondly, in this book, I examined the role of controversies in metaphysics. Yet, as I noted in Part One, it is difficult, if not impossible, to separate metaphysical from logical, modal, and semantic issues. This supports the idea that controversies are no less relevant to the development of other philosophical disciplines than to metaphysics. That is, a similar account of the role of controversies could be used in the attempt to chart the intellectual history of other philosophical disciplines. Indeed, there are reasons to believe that the same types of grounds that clarify the positive critical role that controversies have in metaphysics could, in fact, be extended to other branches of philosophy. An examination of the role of polemical dialogues in other branches of philosophy that follows the model presented in this book seems to me to be promising and worthwhile.

Conclusion 

The metaphysical debate on the nature of the mind continues. The neverceasing hopes of new philosophers to solve the metaphysical puzzles “once and for all” are not merely illusory hopes. Pursuing the fulfillment of these hopes steadily shapes and reshapes our concepts, the domain of conceivable possibilities, and the questions that require answers. What is now conceived as a vital philosophical enigma is inexplicable without revealing the traces of the philosophical efforts and failures that generated it. The history that shaped the current state of knowledge is not always visible to those that use the scheme of concepts provided to them by language, philosophical texts, and other cultural entities. The illusion that metaphysics has no history is not necessarily undesirable if one considers its results. In a sense, it is a useful illusion. It seems to cope with the goals of metaphysics. Practicing metaphysics “in person” must be conducted in isolation, as if from “nowhere.” Nothing can replace the direct contact between persons and reason. The history of metaphysics qua story of the cumulative communicative exchanges that are published and preserved is bound to be at the margins of the direct contact between persons and reason. It presupposes the attempt to uncover in isolation the most basic layers of reality. I hope that this book has not persuaded anyone to renounce the personal approach to the tasks of metaphysics. I consider myself a follower of this style of philosophizing. And yet, I hope that I have persuaded the reader that in order to practice metaphysics for its own sake, from this very solipsistic stance, one has also to practice metaphysics with others, with whom one often disagrees.

References Allison, E.H. 1973. The Kant-Eberhard Controversy. Baltimore, MD: The Johns Hopkins University Press. Allison, E.H. 2004. Kant’s Transcendental Idealism. An Interpretation and Defense (Revised and enlarged ed.). New Haven, CT: Yale University Press. Baker, L.R. 1997. “Why constitution is not identity?”. The Journal of Philosophy 94 (12): 599–621. Baker, L.R. 2000. Persons and Bodies. Cambridge: Cambridge University Press. Balog, K. 2001. “Commentary on Frank Jackson’s From Metaphysics to Ethics.” Philosophy and Phenomenological Research 62 (3): 645–652. Bealer, G. 1996a. “A priori knowledge and the scope of philosophy.” Philosophical Studies 81 (2): 121–142. Bealer, G. 1996b. “On the possibility of philosophical knowledge.” Noûs 30 Supplement: Philosophical Perspectives 10: 1–34. Bealer, G. 2002. “Modal epistemology and the rationalist renaissance.” In T. Szabó- Gendler and J. Hawthorne (eds), 71–125. Beiser, C.F. 1987. The Fate of Reason. Cambridge, MA: Harvard University Press. Bennett, J. 1974. Kant’s Dialectic. Cambridge: Cambridge University Press. Block, N. 1978. “Troubles with functionalism.” in C.W. Savage (ed), Minnesota Studies in the Philosophy of Science 1: 261–325. [Reprinted with revisions in Block, 2007.] Block, N. 1980. “Are absent qualia impossible?.” The Philosophical Review 89 (2): 257–274. [Reprinted with revisions in Block 2007.] Block, N. 1990. “Inverted earth.” Philosophical Perspectives 4: 53–79. [Reprinted in Block 2007.] Block, N. 2003. “Mental paint.” In M. Hahn and B. Ramberg (eds), Reflections on the Philosophy of Tyler Burge. Cambridge: Cambridge University Press, 165–200. [Reprinted in Block 2007.] Block, N. 2007. Consciousness, Function, and Representation. Cambridge, MA: The MIT Press. Block, N. and Fodor, J. 1972. “What psychological states are not.” The Philosophical Review 81 (2): 152–181. Block, N., Flanagan, O., and Güzeldere, G. (eds). 1997. The Nature of Consciousness. Cambridge, MA: The MIT Press. Block, N. and Stalnaker, R. 1999. “Conceptual analysis, dualism, and the explanatory gap.” The Philosophical Review 108 (1): 1–46. Bonjour, L. and Sosa, E. 2003. Epistemic Justification. Oxford: Blackwell. Byrne A. 2002. “Something about Mary.” Grazer Philosophical Studies 63: 213–72. Chalmers, D. 1996. The Conscious Mind. Oxford: Oxford University Press. Chalmers, D. 1999. “Materialism and the metaphysics of modality.” Philosophy and Phenomenological Research 59 (2): 473–496. Chalmers, D. 2002. “Does conceivability entail possibility?”. In T. Szabó Gendler and J. Hawthorne (eds), 145–200.

 Controversies and the Metaphysics of Mind Chalmers, D. 2004. “Phenomenal concepts and the knowledge argument.” In P. Ludlow, Y. Nagasawa, and D. Stoljar (eds), 269–297. Chalmers, D. 2006. “The foundations of two dimensional semantics.” In M. García-Carpintero and J. Macià (eds), 55–140. Chalmers, D. and Jackson, F. 2001. “Conceptual analysis and reductive explanation.” The Philosophical Review 110 (3): 315–360. Churchland, P. 1981. “Eliminative materialism and the propositional attitudes.” The Journal of Philosophy 78 (2): 67–90. Churchland, P. 1985. “Reduction, qualia and the direct introspection of brain states.” The Journal of Philosophy 82 (1): 8–28. Churchland, P. 1997. “Knowing qualia: A reply to Jackson.” In N. Block, O. Flanagan, and G. Güzeldere (eds), 571–577. Churchland, P. 2004. “Philosophy of mind meets logical theory: Perry on neo-dualism.” Philosophy and Phenomenological Research 58 (1): 199–206. Cummins, R. 1983. The Nature of Psychological Explanation. Cambridge, MA: The MIT Press. Dancy, J. 1997. Reading Parfit. Oxford: Blackwell. Dascal, M. 1998a. “Types of polemics and types of polemical moves.” In S. Čmejrková, J. Hoffmannová, O. Müllerová, and J. Svetlá (eds), Dialogue Analysis VI (= Proceedings of the 6th Conference, Prague 1996), vol.1. Tübingen: Max Niemeyer, 15–33. Dascal, M. 1998b. “The study of controversies and the theory and history of Science.” Science in Context 11 (2): 147–154. Dascal, M. 2005a. “Introduction to controversies and subjectivity.” In P.L. Barrota and M. Dascal (eds), Controversies and Subjectivity. Amsterdam: John Benjamins, 1–29. Dascal, M. 2005b. “The balance of reason.” In D. Vanderveken (ed), Logic, Thought and Action. Dordrecht: Springer: 27–47. Dascal, M. 2008a. “Dichotomies and types of debate.” In F.H. van Eemeren and B. Grassen (eds), Controversy and Confrontation: Relating Controversy Analysis with Argumentation Theory. Amsterdam: John Benjamins, 27–49. Dascal, M. 2008b. “Leibniz’s two-pronged dialectic.” In M. Dascal (ed), Leibniz: What Kind of Rationalist?. Dordrecht: Springer, 37–72. Dascal, M. 2008c. “Introductory essay”. In M. Dascal (ed), G.W. Leibniz: The Art of Controversies. Dordrecht: Springer, xix-lxxii. Davies, M. and Humberstone, L. 1980. “Two notions of necessity.” Philosophical Studies 38: 1–30. Dennett, D. 1991. Consciousness Explained. London: Penguin Books. Dennett, D. 2005. Sweet Dreams. Cambridge, MA: The MIT Press. DePaul, M. and Ramsey, W. 1998. Rethinking Intuition. Lanham, MD: Rowman and Littlefield. Eberhard, J.A. 1789a. “Ueber die Schranken der Menschlichen Erkenntniß.” Philosophisches Magazin (Erster Band): 5–29. Eberhard, J.A. 1789b. “Ueber die logische Wahrheit.” Philosophisches Magazin (Erster Band): 150–173. Eberhard, J.A. 1789c. “Die Unterscheidung der Urtheile in analytische und synthetische.” Philosophisches Magazin (Erster Band): 307–332. Eberhard, J.A. 1790. “Weitere Ausführung der Untersuchung über die Unterscheidung der Urteile in analytische und synthetische.” Philosophisches Magazin (Zweiter Band): 285–315. Eberhard, J.A. 1791. “Die analytischen und synthetischen Urtheile.” Philosophisches Magazin (Dritter Band): 280–303.

References  van Eemeren, F.H. and Grootendorst, R. 1984. Speech Acts in Argumentative Discussions. Dordrecht: Foris Publication. van Eemeren, F.H. and Grootendorst, R. 1992. Argumentation, Communication, and Fallacies. Hillsdale, N.J.: Lawrence Erlbaum Association. van Eemeren, F.H. and Grootendorst, R.. 2004. A Systematic Theory of Argumentation. Cambridge: Cambridge University Press. Feigl, H. 1958. ‘The “Mental” and the “Physical”.” In H. Feigl, M. Scriven, and G. Maxwell (eds), Concepts, Theories and the Mind-Body Problem. Minneapolis, Minnesota Studies in the Philosophy of Science 2: 370–497. Freeman, B.J. 2005. Acceptable Premises. Cambridge: Cambridge University Press. Freudenthal, G. 1998. “Controversy.” Science in Context 11 (2): 155–160. Frogel, S. 2005. The Rhetoric of Philosophy. Amsterdam: John Benjamins. García-Carpintero, M. and Macià, J. (eds). 2006. Two-Dimensional Semantics. Oxford: Clarendon Press. Godden, D.M and Walton, D. 2007. “A theory of presumption for everyday argumentation.” Pragmatics & Cognition 15(2): 313–346. Graham, G. and Horgan, T. 2000. “Mary Mary, quite contrary.” Philosophical Studies 99: 59–87. Graham, G. and Horgan, T. 2005. “Mary Mary Au Contraire: Reply to Raffman.” Philosophical Studies 122: 203–212. Grice, P. 1941. “Personal identity.” Mind 50: 330–350. Grier, M. 2001. Kant’s Doctrine of Transcendental Illusion. Cambridge: Cambridge University Press. Hamblin, C.L. 1970. Fallacies. London: Methuen. Hamblin, C.L. 1971. “Mathematical models of dialogue.” Theoria 37: 130–155. Harman, G. 1990. “The intrinsic quality of experience.” Philosophical Perspectives 4: 31–52. Horgan, T. 1984. “Jackson on physical information and qualia.” Philosophical Quarterly 34(135): 147–152. Jackson, F. 1982. “Epiphenomenal qualia.” Philosophical Quarterly 32 (127): 127–136. Jackson, F. 1986. “What Mary didn’t know.” Journal of Philosophy 83 (5): 291–295. [Reprinted in N. Block, O. Flanagan, and G. Güzeldere (eds), 567–570.] Jackson, F. 1998. From Metaphysics to Ethics. Oxford: Clarendon Press. Jackson, F. 2001. “Responses.” Philosophy and Phenomenological Research 62 (3): 653–664. Jackson, F. 2003. “From H2O to water: The relevance to a priori passage.” In H. Lellehammer and G. Rodriguez-Pereyra (eds), Real Metaphysics. London: Routledge, 84–97. Jackson, F. 2004. “Mind and illusion.” In P. Ludlow, Y. Nagasawa, and D. Stoljar (eds), 421–440. Johnston, M. 1987. “Human beings.” The Journal of Philosophy 84(2): 59–83. Kant, I. Kants gesammelte Schriften. Akademie edition (Berlin, 1900—). Kant, I. 1998. Critique of Pure Reason. P. Guyer and A.W. Wood (eds, trans). Cambridge: Cambridge University Press. Kant, I. 1997. Lectures on Metaphysics. K. Ameriks and S. Naragon (eds, trans). Cambridge: Cambridge University Press. Kant, I. 2002. Theoretical Philosophy after 1781. H. Allison and P. Heath (eds), G. Hatfield, M. Friedman, H. Allison, and P. Heath (trans). Cambridge: Cambridge University Press. Kaplan, D. 1979. “Dthat.” In P. Cole (ed), Syntax and Semantics. New York: Academic Press. Kaplan, D. 1989. “Demonstratives.” In J. Almog, J. Perry, and H. Wettstein (eds), Themes from Kaplan. New York: Oxford University Press, 481–563. Kim, J. 2005. Physicalism or Something Near Enough. Princeton, NJ: Princeton University Press.

 Controversies and the Metaphysics of Mind Kripke, S. 1971. “Identity and necessity.” In M.K. Munitz (ed), Identity and Individuation. New York: New York University Press, 135–164. Kripke, S. 1980. Naming and Necessity. Cambridge, MA: Harvard University Press. Levine, J. 2004. “The evidential status of philosophical intuition.” Philosophical Studies 121(3): 193–224. Levine, J. 1983. “Materialism and qualia: The explanatory gap.” Pacific Philosophical Quarterly 64: 354–361. Levine, J. 1993. “On leaving out what it is like.” In M. Davis and G. Humphreys (eds), Consciousness: Psychological and Philosophical Essays. Oxford: Blackwell, 121–136. [Reprinted in N. Block, O. Flanagan, and G. Güzeldere (eds). 1997, 543–555.] Levine, J. 1998. “Conceivability and the metaphysics of mind.” Noûs 32 (4): 449–480. Levine, J. 2001. Purple Haze: The Puzzle of Consciousness. Oxford: Oxford University Press. Lewis, D. 1976. “Survival and identity.” In A. Rorty (ed), 17–40. [Reprinted in Lewis 1983a.] Lewis, D. 1983a. Philosophical Papers, Volume I. Oxford: Oxford University Press. Lewis, D. 1983b. “Postscript to ‘Mad Pain and Martian Pain’.” In D. Lewis 1983a, 130–132. Lewis, D. 1983. “New work for a theory of universals.” Australasian Journal of Philosophy 61: 343–377. [Reprinted in Lewis 1999.] Lewis, D. 1988. “What experience teaches.” Proceedings of the Russellian Society 13: 29–57. [Reprinted in Lewis 1999.] Lewis, D. 1999. Papers in Metaphysics and Epistemology. Cambridge: Cambridge University Press. Loar, B. 1990. “Phenomenal states.” Philosophical Perspectives 4: 81–108. Loar, B. 1997. “Phenomenal states.” In N. Block, O. Flanagan, and G. Güzeldere (eds), 597–616. Loar, B. 1999. “David Chalmers’s The Conscious Mind.” Philosophy and Phenomenological Research 59 (2): 465–472. Locke, J. 1975. An Essay Concerning Human Understanding. Oxford: Clarendon Press. Ludlow, P., Nagasawa, Y., and Stoljar, D. (eds). 2004. There is Something About Mary. Cambridge, MA: The MIT Press. Maaß, J.G.E. 1789. “Ueber den Höchsten Grundsatz der synthetischen Urtheile; in Beziehung auf die Theorie von der mathemathischen Gewißheit.” Philosophisches Magazin (Erster Band): 186–231. Maaß, J.G.E. 1791. “Der Satz des zureichenden Grundes.” Philosophisches Magazin (Dritter Band): 173–203. McGinn, C. 1991. The Problem of Consciousness. Oxford: Basil Blackwell. McGinn, C. 1999. The Mysterious Flame. New York: Basic Books. Nemirow, L. 1980. “Review of Thomas Nagel, Mortal Questions.” Philosophical Review 89 (3): 473–477. Noonan H. 1989. Personal Identity. London: Routledge. Olson, E.T. 1997. The Human Animal. Oxford: Oxford University Press. Olson, E.T. 2007. What Are We?. Oxford: Oxford University Press. Papineau, D. 1993. “Physicalism, consciousness, and the antipathetic fallacy.” Australasian Journal of Philosophy 71:169–83. Papineau, D. 1998. “Mind the gap.” Philosophical Perspectives 12: 373–388. Papineau, D. 2002. Thinking about Consciousness. Oxford: Clarendon Press. Parfit, D. 1971. “Personal Identity.” The Philosophical Review 80 (1): 3–27. Parfit, D. 1976. “Lewis, Perry, and what matters.” In A. Rorty (ed), 91–107. Parfit, D. 1984. Reasons and Persons. Oxford: Clarendon Press. Perelman C. 1982. The Realm of Rhetoric. Notre Dame, IN: University of Notre Dame Press.

References  Perelman C. and Olbrechts-Tyteca, L. 1969. The New Rhetoric. Notre Dame, IN: University of Notre Dame Press. Perry, J. 1972. “Can the Self Divide?”. The Journal of Philosophy 69 (16): 463–488. [Reprinted in Perry 2002.] Perry, J. 1976. “Review of B. Williams Problems of the Self.” The Journal of Philosophy 73 (13): 416–428. Perry, J. (ed). 1975. Personal Identity. Berkeley, CA: University of California Press. Perry, J. 1976. “The importance of being identical.” In A. Rorty (ed), 67–90. [Reprinted in Perry 2002.] Perry, J. 2001. Knowledge, Possibility, Consciousness. Cambridge, MA: The MIT Press. Perry, J. 2002. Identity, Personal Identity, and the Self. Indianapolis, IN: Hackett Publishing Company. Place, U.T. 1956. “Is consciousness a brain process?.” British Journal of Psychology 47: 44–50. Popper, K. 1963a. “On the status of science and metaphysics.” In K. Popper 1963c. Popper, K. 1963b. “The demarcation between science and metaphysics.” In K. Popper 1963c. Popper, K. 1963c. Conjectures and Refutations. New York: Harper. Quine, W.V.O. 1953. “Two dogmas of empiricism.” In From a Logical Point of View. Cambridge, MA: Harvard University Press, 20–46. Quinton, A. 1962. “The soul.” The Journal of Philosophy 59 (15): 393–409. Raffman, D. 2005. “Even zombies can be surprised: A reply to Graham and Horgan.” Philosophical Studies 122: 189–202. Rescher, N. 1977. Dialectic. Albany: State University of New York Press. Rescher, N. 2006. Presumption and the Practices of Tentative Cognition. Cambridge: Cambridge University Press. Robinson, H. 1993. “Dennett on the knowledge argument.” Analysis 53: 174–177. [Reprinted in P. Ludlow, Y. Nagasawa, and D. Stoljar (eds), 69–73.] Rorty, A. (ed). 1976. The Identities of Persons. Berkeley, CA: University of California Press. Rorty, R. 1980. Philosophy and the Mirror of Nature. Princeton, NJ: Princeton University Press. Rovane, C. 1998. The Bounds of Agency. Princeton, NJ: Princeton University Press. Russell, B. 1910. “Knowledge by acquaintance and knowledge by description.” In his Mysticism and Logic. London: Unwin Books, 152–167. Saner, H. 1973. Kant’s Political Thought. E.B. Ashton (trans). Chicago: The University of Chicago Press. Senderowicz, Y. 1998. “Facing the bounds of tradition: The Kant – Eberhard controversy.” Science in Context 11 (2): 205–228. Senderowicz, Y. 2004. “Figurative synthesis and synthetic a priori knowledge.” The Review of Metaphysics 57: 755–785. Senderowicz, Y. 2005. The Coherence of Kant’s Transcendental Idealism. Dordrecht: Springer. Shoemaker, S. 1959. “Personal identity and memory.” The Journal of Philosophy 56 (22): 868–882. Shoemaker, S. 1963. Self-Knowledge and Self-Identity. Ithaca, NY: Cornell University Press. Shoemaker, S. 1968. “Self-reference and self-awareness.” The Journal of Philosophy 65 (19): 555–567. [Reprinted in Shoemaker 2003.] Shoemaker, S. 1970a. “Wiggins on identity.” The Philosophical Review 79 (4): 529–544. Shoemaker, S. 1970b. “Persons and their pasts.” American Philosophical Quarterly 7 (5): 269– 285. [Reprinted in Shoemaker 2003.] Shoemaker, S. 1975. “Functionalism and qualia.” Philosophical Studies 72: 291–315. [Reprinted in Shoemaker 2003.]

 Controversies and the Metaphysics of Mind Shoemaker, S. 1981. “Absent qualia are impossible.” The Philosophical Review 90 (4): 581–599. Shoemaker, S. 1997a. “Self and Substance.” Philosophical Perspectives 11: 283–304. [Reprinted in Shoemaker 2003.] Shoemaker, S. 1997b. “Parfit on Identity.” In J. Dancy (ed), 135–148. Shoemaker, S. 2003. Identity, Cause, and Mind (Expanded Edition). Oxford: Oxford University Press. Shoemaker, S. 2004a. “Brown-Brownson revisited.” The Monist 87 (4): 573–593. Shoemaker, S. 2004b. “Reply to Wiggins.” The Monist 87 (4): 610–613. Shoemaker, S. and Swinburne, R. 1984. Personal Identity. Oxford: Basil Blackwell. Smart, J.J.C. 1959. “Sensations and brain processes.” Philosophical Review 69 (4): 141–156. Smart, J.J.C. 2007. “The identity theory of mind.” In Stanford Encyclopedia of Philosophy. http://plato.stanford.edu/entries/mind-identity/ Soams, S. 2005. Reference and Description. Princeton, NJ: Princeton University Press. Stalnaker, R. 1978. “Assertion.” In P.Cole (ed), Syntax and Semantics 9: Pragmatics. New York: Academic Press, 315–332. Stalnaker, R. 2001. “Metaphysics without conceptual analysis.” Philosophy and Phenomenological Research 62 (3): 631–636. Stalnaker, R. 2008. Our Knowledge of the External World. Oxford: Clarendon Press. Stoljar, D. and Nagasawa, Y. 2004. “Introduction to There’s Something About Mary.” In P. Ludlow, Y. Nagasawa, and D. Stoljar (eds), 1–36. Strawson, P.F. 1990. Individuals. London: Routledge. [First edition 1959] Swoyer, C. 1999. “How ontology might be possible: Explanation and inference in metaphysics.” Midwest Studies in Philosophy 23: 100–131. Szabó-Gendler, T. and Hawthorne, J. (eds). 2002. Conceivability and Possibility. Oxford: Clarendon Press. Thompson, E. 1995. Color Vision. London: Routledge. Thorin, A. and Walter, S. 2007. Phenomenal Concepts and Phenomenal Knowledge. Oxford: Oxford University Press. Tye, M. 1995. Ten Problems of Consciousness. Cambridge, MA: The MIT Press. Tye, M. 2000. Consciousness, Color and Content. Cambridge MA: The MIT Press. Tye, M. 2009. Consciousness Revisited. Cambridge MA: The MIT Press. Ullmann-Margalit, E. 1983. “On presumption.” The Journal of Philosophy 80 (3): 143–163. Walton, D. 1984. Logical Dialogue-Games and Fallacies. Lanham, MA: University Press of America. Walton, D. 1998. The New Dialectic. Toronto: University of Toronto Press. Wiggins, D. 1967. Identity and Spatio-Temporal Continuity. Oxford: Basil Blackwell. Wiggins, D. 1980. Sameness and Substance. Oxford: Basil Blackwell. Wiggins, D. 2001. Sameness and Substance Renewed. Cambridge: Cambridge University Press. Wiggins, D. 2004a. “Reply to Shoemaker.” The Monist 87 (4): 594–609. Wiggins, D. 2004b. “Reply to Shoemaker’s reply.” The Monist 87 (4): 614–615. Wilkes, K. 1988. Real People: Personal Identity without Thought Experiments. Oxford: Clarendon Press. Williams, B. 1956–7. “Personal Identity and individuation.” Proceedings of the Aristotelian Society, New Series, 57: 229–252. [Reprinted in Williams 1973.] Williams, B. 1961. “Mr. Strawson on Individuals.” Philosophy 36 (138): 309–332. [Reprinted in Williams 1973.]

References  Williams, B. 1970. “The self and the future.” The Philosophical Review 79 (2): 161–180. [Reprinted in Williams 1973.] Williams, B. 1973. Problems of the Self. Cambridge: Cambridge University Press. Yablo, S. 1993. “Is conceivability a guide to possibility?.” Philosophy and Phenomenological Research 53 (1): 1–42.

Index A ability hypothesis 6, 101, 111 Allison, E.H. 9, 18, 21, 24 antinomy 7, 9–11, 13, 17, 19, 27–28, 30–32, 36, 56, 75, 97, 103, 107–109, 153, 155, 165, 167, 192, 196–197, 201, 203, 212 a priori entailment 136–137, 143–144, 149, 152–153 explanation 141, 143, 145–146, 150 judgment 29 knowledge 2, 7, 19–20, 25–26, 71–72, 84, 86, 112, 125, 129–137, 140, 142, 149, 152, 154, 156–157, 184, 186 a priority 84, 149–150 argument dialectical 9, 11–12, 36, 72, 168 type A 54, 143 type B 55, 83, 87, 95, 116, 201, 204–205 argumentation 48–50 theory 40 probative 42 Armstrong, D. 187 Austin, J. L. 1 B Baker, L.R 183, 219 Balog, K. 133 Bealer, G. 59–60, 62, 67 Bennett, J. 32 Block, N. 53, 103, 116, 125–127, 133, 138, 140–153, 155–158, 218 bodily continuity 12, 64, 163, 173, 178, 181–184, 187–188, 197, 201–202, 213–214 identity 161–162, 169–172, 175–176, 178–179, 182–183, 190, 199

body exchange 13, 170, 175, 177–178, 184, 190, 192, 198–200, 203–204, 215 Butler, J. 171, 212 Byrne, A. 94 C Carnap, R. 154 causal efficacy of conscious states 103, 107–108 Chalmers, D. 2, 44, 73–74, 105, 117–122, 125–126, 131, 136–141, 143–158 chain of controversies 12, 49, 118, 125, 158 Churchland, P. 95, 97–104, 107, 110, 113, 131 common sense 29–30, 103, 163, 188, 198, 206 see also commonsensical intuition concept theoretical 112, 120–121 indexical 115 recognitional 112–113, 115, 120–121 phenomenal 97, 109–115, 117, 120–122 physical-functional 110, 112–113, 120–122 conceptual analysis 3–4, 6, 60, 126, 130– 135, 138–139, 141, 143–144, 146, 149, 184, 186, 210 innovations 9, 108–109, 192 see also non-conceptual content conceivability 22, 30–31, 53, 88, 91, 117–119, 125, 127, 129–131, 136, 141–143, 154, 177–178, 190, 210, 215 conscious experience 90, 97, 106–108, 113–114, 119, 122–123, 125, 129, 144, 165

consciousness 12, 47, 65, 75, 82, 97–98, 103, 106–108, 110, 112–116, 118–119, 122–123, 125, 127, 142–143, 149, 158, 163, 164–166, 168–169, 172– 173, 178, 186, 189, 209 see also phenomenal consciousness content see non-conceptual content; representational content contingent modes of presentation 112–113, 120–121 controversy 9–10, 18, 20–22, 25, 30–31, 35, 37, 39, 45, 47–49, 51–55, 66, 113–114, 116–118, 121, 125–126, 136, 141, 153, 158, 205, 211–212, 222 Cummins, R. 128 D Dascal, M. ix, 33, 37–39, 46 Davies, M. 134 Dennett, D. 74, 104–107, 110 discussion 37, 39–40, 48 dispute 2, 19, 21–22, 26, 28, 37, 39, 46, 52 E Eberhard, J.A. 18, 21 van Eemeren, F.H. 47–50, 54 epiphenomenalism 91–92, 94–95, 103–107, 131, 145 epistemic possibility 131, 156–158 explanation 26, 55–56, 127–128, 146, 152–153 causal 24, 130 empirical 72 physical 94 psychophysical 81, 122 reductive 98, 117, 138–139, 143, 145, 148, 150 scientific 6, 24, 26, 97, 127, 129, 138, 141, 143–144, 151, 158

 Controversies and the Metaphysics of Mind explanatory gap 12, 75, 125–127, 129–131, 136, 138, 141–144, 146, 156–157 F Feigl, H. 83 first-person viewpoint 103 Fodor, J. 187 Freeman, B.J. 40 Frege, G. 154–156 Freudenthal, G. 53 Frogel, S. ix, 51 G García-Carpintero, M. 126 Godden, D.M 40 Graham, G. 93, 111 Grice, P. 209 Grier, M. 24 Grootendorst, R. 47–50, 54 H Hamblin, C.L. 46 Harman, G. 115 Hawthorne, J. 126 Hegelianism 10, 46, 89, 162 Horgan, T. 93, 110–111, 130 Humberstone, L. 134 Husserl, E. 161 I identity see personal identity; relative identity; self-identity illusion 5, 20, 51, 75, 122–123, 162, 170, 177, 193, 223 intension primary 118–120, 139–140, 147–148, 154 secondary 117, 119, 139, 149, 157 intuition 60–65, 69, 71–78, 87, 90, 92–94, 97, 99, 103–104, 104, 106–107, 127, 131, 133, 144, 146, 148, 152, 162–163, 169, 175, 182–183, 196, 199–201, 208, 210, 218 a priori 7, 67 causal efficacy 106–110, 114, 118, 130 commonsensical 210 empirical 7, 21 knowledge 60, 82, 95, 106–108, 109, 114, 130 Lockean 198, 200, 214

logical 73, 198 modal 87–88 metaphysical 2, 8–9, 11, 25–26, 30, 32, 36, 38, 45–46, 57, 59–60, 73, 76, 95, 121, 125, 163, 198 of distinctness 123 philosophical 67 rational 66 self-consciousness 163 bodily-continuity 163 semantic 118 sensible 22 J Jackson, F. 2, 5–6, 12, 35, 44, 60–61, 66–68, 73–75, 81–82, 91–95, 97–114, 116–117, 122– 123, 125–126, 129–141, 143–146, 148–153, 155–156 Johnston, M. 197 K Kant, I. 7–9, 11, 17–33, 35, 44, 57, 60, 65, 73–75, 97, 103, 153, 161, 188 Kantian 6, 11, 17, 25–26, 31, 47, 154, 162, 201 Kantianism 46–47 Kaplan, D. 134, 139 Kim, J. 126 knowledge argument 5–6, 12, 39, 42, 60, 66–68, 74, 79, 81–82, 91–92, 94–95, 98–116, 118, 122, 126, 129, 131, 133, 138, 157 by acquaintance 99–102 discursive 99, 101–102 see also a priori; phenomenal Kripke, S. 12, 82–89, 107, 109, 111–112, 114, 116, 119, 125–126, 128, 134, 136, 142, 143, 154 L Leibniz, W.G. 21, 32, 39 Leibniz law 99, 101, 181 Levine, E.J. 108, 121, 125–120, 136, 139–143 Levine, M.J. 59, 74 Lewis, D. 13, 100, 101, 111, 132, 137, 187, 190, 195, 198, 206–207, 215, 216 Loar, B. 109–122, 136

Locke, J. 66, 161–163, 170–171, 173–174, 178, 181–182, 187, 192, 198, 200–201, 203, 209, 212–214, 215, 218 neo-Lockeanism 12–13, 46, 187, 192, 196–205, 209, 211–212, 214–219 logical necessity 117, 118, 139 supervenience 117–118, 138–139 M Maaß, J.G.E. 18 Macià, J. 126 materialism 99, 117–119, 121, 123, 125, 130, 152, 162 McGinn, C. 75, 108, 123 memory criterion 178, 181, 183–184, 188, 190–191, 197, 213–214 mental causation 81, 103–104 metaphysical supervenience 117–118 necessity 117–118, 136 possibility 118, 128, 131, 158 metaphysics descriptive 3–4, 12, 163 revisionary 159, 181, 187–188, 196 serious 60, 126, 131 mind-body dualism 82, 87–89, 91, 99–100, 102, 108, 111, 117, 122, 125, 129, 131–132, 138, 140–141, 152, 175 problem 73, 81, 82, 118, 123, 131 N Nagasawa, Y. 106 Nemirow, L. 100, 111 non-conceptual content 115–116 Noonan H. 196 O Olson, E.T. 12, 219 P PANIC theory 114–115 Papineau, D. 114, 122–123, 146, 152 paradox of phenomenal consciousness 107–108, 114

Index  Parfit, D. 12–13, 196–198, 202–206, 210–212, 219 personal identity 2, 5, 12–13, 64, 158–159, 161–164, 169–179, 181–184, 187–188, 190–193, 195–208, 210–212, 214–215, 218–219 personhood 72, 161, 169–170, 179, 181, 186, 210–212, 215, 216, 218 phenomenal beliefs 106 concepts 97, 109–115, 117, 120–122 consciousness 97, 113–116, 118 knowledge 106 states 114, 115–116, 120, 127 see also properties; paradox of phenomenal consciousness Place, U.T. 83 physical information 67, 92–95, 100–101, 106, 110–111, 130–131, 133, 138, 146, 165, 175, 179, 189 physicalism 44, 81–82, 88, 92, 94, 105, 111, 114, 116–117, 127, 131–132, 134–138, 140, 143, 145, 148 Popper, K. 18, 25 presumption 39–46, 48–51, 54–57, 60, 63, 65, 68, 74, 77, 82, 87, 95, 104, 191–192, 197 property epiphenomenal 94, 102, 105, 131 objective 23, 102 phenomenal 2, 81, 94, 102, 105–106, 111–113, 115–116, 131 physical 2, 44, 90, 93–95, 100, 102, 107, 110, 113, 118, 130–131, 138, 145 psychological continuity 12, 64, 184, 187, 192, 202, 214 Putnam, H. 137, 187 Q qualia 5, 25, 81, 91–92, 95, 97, 98, 102–103, 107, 110, 112, 125, 127, 129, 131, 136

q-strategy 196, 198, 202, 204, 215, 217 quasi-memory 187–188, 194–196, 202 Quine, W.V.O. 1–2 Quinton, A. 209 R Raffman, D. 111 Ramsey, W. 59 real possibility 22, 54, 76 relative identity 13, 181, 182–183, 188–189, 192 relevant alternative 8–11, 75–77, 104, 126, 140–141, 204, 218, 221 representational content 115 Rescher, N. 35, 39, 40–45 Robinson, H. 105–106, 173, 175, 199 Rorty, R. 2 Rovane, C. 219 Russell, B. 99 Ryle, G. 1 S Saner, H. 9 scientific explanation 6, 12, 24, 97, 118, 125, 127–129, 138, 141, 143–144, 149, 151, 158 reduction 6, 125–126, 128, 144 self-consciousness 12, 161–163, 169–170, 178, 204, 213 self-identity 170, 173–174, 176, 182, 196, 199, 204, 213 Senderowicz, Y. 18–19, 21–22 Shoemaker, S. 12–13, 66, 103, 106, 162–163, 173–179, 182–185, 187–199, 202, 203–205, 207, 209, 213–215, 219 Smart, J.J.C. 1959. 83–85, 87 Soams, S. 126 spatiotemporal continuity 181, 184 Stalnaker, R. 2, 94, 125–126, 133–134, 137–138, 140–153, 155–158 Stoljar, D. 106

Strawson, P.F. 3–5, 12, 162–170, 172, 174, 178, 182, 187, 189, 191, 193, 195, 208, 215, 221 strong necessity 117, 120–121 Swoyer, C. 65 T Thompson, E. 93 Thorin, A. 126 Third Person viewpoint 106, 176 Thought experiment 4, 11, 13, 25–26, 32, 35–36, 45, 59–60, 64, 66–69, 72, 74, 93–94, 102, 104–105, 111, 129, 133, 162–164, 171–173, 178, 184–186, 188, 190, 193–195, 199, 201, 203, 207, 217–218 transcendental idealism 7–8, 21, 29–30, 57, 188 Illusion 7, 24, 32, 74–75 two-dimensional semantics 12, 97, 106, 117, 118, 119, 122, 125, 132, 136, 138–139, 141, 143, 146, 152–156 framework 126, 147, 149, 152–155 two-dimensionalism 117–119, 121, 126, 140–141, 147–149, 153–156, 158 Tye, M. 114–116, 122–123, 133 U Ullmann-Margalit, E. 40–41 W Walton, D. 38, 40, 46–47, 49 Walter, S. 126 Wiggins, D. 12–13, 181–196, 198, 203, 207, 211–219 Wilkes, K. 68 Williams, B. 161–163, 169–170, 172–173, 175, 177–178, 182–183, 187, 195, 197, 199–204, 211–212 Wittgenstein, L. 1, 176 Y Yablo, S. 59–60, 62, 69–70

In the series Controversies the following titles have been published thus far or are scheduled for publication: 8 7 6 5 4 3 2 1

SENDEROWICZ, Yaron M.: Controversies and the Metaphysics of Mind. 2010. xi, 235 pp. DASCAL, Marcelo (ed.): The Practice of Reason. Leibniz and his Controversies. 2010. xvi, 359 pp. EEMEREN, Frans H. van and Bart GARSSEN (eds.): Controversy and Confrontation. Relating controversy analysis with argumentation theory. 2008. xiii, 278 pp. WALTON, Douglas: Dialog Theory for Critical Argumentation. 2007. xviii, 308 pp. DASCAL, Marcelo and Han-liang CHANG (eds.): Traditions of Controversy. 2007. xvi, 310 pp. FROGEL, Shai: The Rhetoric of Philosophy. 2005. x, 156 pp. EEMEREN, Frans H. van and Peter HOUTLOSSER (eds.): Argumentation in Practice. 2005. viii, 368 pp. BARROTTA, Pierluigi and Marcelo DASCAL (eds.): Controversies and Subjectivity. 2005. x, 411 pp.

Controversies and the Metaphysics of Mind

Metaphysics and the Philosophy of Mind

Review of 'The Metaphysics of Mind'

Essays in the Metaphysics of Mind

The Evolution of Mind: Fundamental Questions and Controversies

Personal Agency: The Metaphysics of Mind and Action

Debating Dispositions: Issues in Metaphysics, Epistemology and Philosophy of Mind

Descriptive versus Revisionary Metaphysics and the Mind-Body Problem

Baboon Metaphysics: The Evolution of a Social Mind

The Metaphysics of Hyperspace

The Metaphysics of Capitalism

The Possibility of Metaphysics

The Metaphysics of Meaning

The God of Metaphysics

The Metaphysics of Knowledge

The Metaphysics of Knowledge

The Metaphysics of Concepts

The Metaphysics of Nonfactualism

The Nature of Metaphysics

The Necessity of Metaphysics

The Rhetoric of Philosophy (Controversies)

The Rhetoric of Philosophy (Controversies)

Nietzsche and the Metaphysics of the Tragic

Historical Controversies And Historians

Metaphysics and the Philosophy of Mind: 2 (The Collected philosophical papers of G.E.M. Anscombe)

The Metaphysics of the Incarnation

The Metaphysics of the School

The Metaphysics of the Tractatus

Metaphysics and the Philosophy of Mind: 2 (The Collected philosophical papers of G.E.M. Anscombe)

Controversies and Subjectivity

Controversies and the Metaphysics of Mind

Metaphysics and the Philosophy of Mind

Review of 'The Metaphysics of Mind'

Essays in the Metaphysics of Mind

The Evolution of Mind: Fundamental Questions and Controversies

Personal Agency: The Metaphysics of Mind and Action

Debating Dispositions: Issues in Metaphysics, Epistemology and Philosophy of Mind

Descriptive versus Revisionary Metaphysics and the Mind-Body Problem

Baboon Metaphysics: The Evolution of a Social Mind

The Metaphysics of Hyperspace

The Metaphysics of Capitalism

The Possibility of Metaphysics

The Metaphysics of Meaning

The God of Metaphysics

The Metaphysics of Knowledge

The Metaphysics of Knowledge

The Metaphysics of Concepts

The Metaphysics of Nonfactualism

The Nature of Metaphysics

The Necessity of Metaphysics

The Rhetoric of Philosophy (Controversies)

The Rhetoric of Philosophy (Controversies)

Nietzsche and the Metaphysics of the Tragic

Historical Controversies And Historians

Metaphysics and the Philosophy of Mind: 2 (The Collected philosophical papers of G.E.M. Anscombe)

The Metaphysics of the Incarnation

The Metaphysics of the School

The Metaphysics of the Tractatus

Metaphysics and the Philosophy of Mind: 2 (The Collected philosophical papers of G.E.M. Anscombe)

Controversies and Subjectivity

Recommend Documents