Personal identity: complex or simple? [First paperback edition] 9781107014442, 1107014441, 9781107538924, 1107538920

Table of contents :
Introduction
Part I. Framing the Question: 1. Chitchat on personal identity David Barnett
2. In search of the simple view Eric T. Olson
3. Personal identity, indeterminacy, and obligation Ryan Wasserman
4. Personal identity and its perplexities Harold W. Noonan
Part II. Arguments for and against Simplicity: 5. How to determine which is the true theory of personal identity Richard Swinburne
6. Against simplicity Sydney Shoemaker
7. The probable simplicity of personal identity E. J. Lowe
8. Reply to E. J. Lowe Sydney Shoemaker
9. The non-descriptive individual nature of conscious beings Martine Nida-Rumelin
Part III. Reconsidering Simplicity: 10. Personal identity: a not-so-simple simple view Lynne Rudder Baker
11. Is 'person' a sortal term? Christian Kanzian
12. Materialism, dualism, and 'simple' theories of personal identity Dean Zimmerman
13. The morphing block and diachronic personal identity Hud Hudson
References
Index.

Citation preview

PERSONAL IDENTITY

We take it for granted that a person persists over time: when we make plans, we assume that we will carry them out; when we punish someone for a crime, we assume that she is the same person as the one who committed it. Metaphysical questions underlying these assumptions point toward an area of deep existential and philosophical interest. In this volume, leading metaphysicians discuss key questions about personal identity, including “What are we?,” “How do we persist?” and “Which conditions guarantee our identity over time?” They discuss whether personal identity is “complex,” whereby it is analyzable in terms of simpler relations such as physical or psychological features, or whether it is “simple,” i.e. something that cannot be analyzed in terms of more fundamental relations. Their essays offer an innovative discussion of this topic and will be of interest to a wide readership in metaphysics. georg gasser is a scientific researcher at the Department of Christian Philosophy, University of Innsbruck. He is the editor of Personal Identity and Resurrection (2010). matthias stefan is a scientific researcher at the Department of Christian Philosophy, University of Innsbruck. He has published papers on the ontological commitments of physicalism, personal identity and substance dualism.

PERSONAL IDENTITY: COMPLEX OR SIMPLE? edited by GEORG GASSER and MATTHIAS STEFAN

cambridge university press Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo, Delhi, Mexico City Cambridge University Press The Edinburgh Building, Cambridge cb2 8ru, UK Published in the United States of America by Cambridge University Press, New York www.cambridge.org Information on this title: www.cambridge.org/9781107014442 © Cambridge University Press 2012 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2012 Printed in the United Kingdom at the University Press, Cambridge A catalogue record for this publication is available from the British Library Library of Congress Cataloging in Publication data Personal identity : complex or simple? / edited by Georg Gasser and Matthias Stefan. p. cm. Includes bibliographical references and index. isbn 978-1-107-01444-2 1. Identity (Psychology) I. Gasser, Georg. II. Stefan, Matthias. BF697.P468 2012 126–dc23 2012012191 isbn 978-1-107-01444-2 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication, and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

To our parents for their love and support

Contents

page ix

List of illustrations List of contributors Acknowledgments

x xi

Introduction g e o r g g a s s e r a n d m a t t h i a s st ef a n

part i 1

19

framing the question

Chitchat on personal identity

21

david barnett

2 In search of the simple view

44

eric t. olson

3 Personal identity, indeterminacy and obligation ry an w as se rm an

4 Personal identity and its perplexities ha ro ld w. no on an

part ii

1

arguments for and against simplicity

5 How to determine which is the true theory of personal identity richard swinburne

6 Against simplicity

63 82 103 105 123

sy dn e y s h oe m a ke r

7 The probable simplicity of personal identity e. j. lowe

vii

137

Contents

viii 8 Reply to E. J. Lowe sy d ney sh o em ak er

9 The non-descriptive individual nature of conscious beings m a r t i n e n i d a -r u¨ m el in

part iii reconsidering simplicity 10 Personal identity: a not-so-simple simple view lynne rudder baker

11 Is “person” a sortal term? christian kanzian

12 Materialism, dualism, and “simple” theories of personal identity dean zimmerman

13 The morphing block and diachronic personal identity

156 157 177 179 192 206

h ud h ud so n

236

References Index

249 257

Illustrations

Figure 1 A taxonomy of views about personal identity over time. Figure 2 Different sums of person-stages.

ix

page 45 75

Contributors

lynne rudder baker is Distinguished Professor in the Department of Philosophy, University of Massachusetts (USA). david barnett is Associate Professor in the Department of Philosophy, University of Colorado (USA). hud hudson is Professor in the Department of Philosophy, Western Washington University (USA). christian kanzian is Associate Professor in the Department of Christian Philosophy, University of Innsbruck (AUT). e. j. lowe is Professor in the Department of Philosophy, Durham University (UK). martine nida-ru¨ melin is Professor of the Philosophy of Language, Philosophy of Mind and the Human Sciences, Department of Philosophy, Université de Fribourg (CH). harold w. noonan is Professor of Mind and Cognition in the Department of Philosophy, University of Nottingham (UK). eric t. olson is Professor in the Department of Philosophy, University of Sheffield (UK). sydney shoemaker is Emeritus Susan Linn Sage Professor of Philosophy, Cornell University (USA). richard swinburne is Emeritus Nolloth Professor of the Philosophy of the Christian Religion, University of Oxford (UK). ryan wasserman is Associate Professor and the Chair of the Department of Philosophy, Western Washington University (USA). dean zimmerman is Professor in the Department of Philosophy, Rutgers University (USA). x

Acknowledgments

The origin of the present volume dates from a conference on personal identity at the conference center of Innsbruck University in Obergurgl, Austria, July 19–22, 2010, where preliminary drafts of the papers were discussed. Sydney Shoemaker’s, Hud Hudson’s and David Barnett’s papers were not presented at the conference, but were written especially for this volume. We wish to thank the following individuals for their role in the production of this volume. First, we owe a debt of thanks to Josef Quitterer, Daniel Wehinger and Monika Datterl for their help in organizing the conference. Dean Zimmerman and Mike Rea were kind enough to offer valuable advice on proceeding with this project after the conference. Our warmest thanks to Katherine Munn for her valuable suggestions for improving the readability of some contributions. We also thank Hilary Gaskin and Thomas O’Reilly of Cambridge University Press for all her help and thoughtful advice in preparing this book for publication. Finally, we would also like to thank all the contributors for their great patience during the completion of the volume. It does not happen as a matter of course that international philosophical conferences receive a great deal of funding. Therefore, we wish to express our deepest gratitude to the Austrian Science Fund (FWF), grant P20186– G14, whose generous support made it possible to organize the conference and to edit this volume. Furthermore, we are also greatly indebted to the Austrian Federal Ministry of Science and Research and to the Vice-Rector for Research of Innsbruck University for co-funding the accommodation expenses of the invited speakers.

xi

Introduction Georg Gasser and Matthias Stefan

the project One worthwhile task for philosophy is to give an overview of a whole domain of thought and to present the conceptual relationships that characterize it. The domain we have striven to portray in this introduction, on a quite general level with a broad brush, is the contemporary debate about personal identity over time. We proceed as follows: First, we specify the metaphysical question of personal identity tackled in this volume: namely, what makes a person P1 at t1 identical to a person P2 at t2? Second, we discuss views which analyze personal identity in terms of bodily and psychological relations. Problems associated with these theories have recently made a fourdimensional interpretation of such views quite popular. The following section presents this canny metaphysical alternative to traditional threedimensional views. Finally we discuss a rather neglected approach to personal identity over time, the so-called “simple view,” according to which personal identity does not consist in anything other than itself; it is simple and unanalyzable. Eric Olson once suggested that the simple view is poorly understood, and therefore deserves more attention than it has received so far (Olson 2010, section 3). A specific aim of this volume was to take up this suggestion. In the first section, “Framing the question,” the authors draw attention to the wider framework in which the question of personal identity is posed. They reveal some of the hitherto implicit background assumptions of the theories at hand, as well as the explanatory demands one should expect of the theories. The contributions of the second section, “Arguments for and against simplicity,” provide original in-depth analyses of arguments put forward in favor of and against the claim that personal identity is analyzable. The last section, “Reconsidering simplicity,” contains innovative and so far rather unnoticed arguments that might strengthen the case for the simple view. 1

2

georg gasser and matthias stefan the question of personal identity

In our ordinary non-philosophical moments we take it for granted in almost everything we do that persons persist over time: when we make plans, we assume that we will carry them out in the future. When we punish someone for a crime, we assume that she is the same person as the one who committed it. When we regret a misdeed, we assume that we are identical with the agent who performed it. These examples indicate that the assumption that personal identity is continuous is of particular importance for practices central to our lives, pertaining to both the treatment of ourselves and of others. Furthermore, to know an entity’s identity conditions is to know what kind of entity it is; a fortiori, this is true of human beings: if you want to figure out who you are essentially, you must ask which conditions guarantee your identity over time. Philosophers often refer to this problem as “the” question of personal identity. But this is misleading; the question of personal identity is not a single problem (see e.g. Rorty 1976). It is important to distinguish between various meanings of “personal identity.” For our purposes we identify four: (i) Biographical (or narrative) identity: Who am I? This question asks how an individual understands and defines herself in light of her values, convictions, and aims. “Identity” in this context is a normative or evaluative concept which incorporates an individual’s selfunderstanding and her broader life-plan. (ii) Personhood: What are the conditions for personhood? This question seeks conditions that make something a person as opposed to a nonperson. It calls for necessary and sufficient conditions for personhood, such as being an intelligent, conscious and feeling agent. (iii) Metaphysical nature: To which metaphysical category do human persons belong? Possible answers include the claims that human persons are temporary stages of human organisms, thinking substances (souls), collections of temporal parts, or bundles of mental and physical states. (iv) Diachronic (personal) identity: What makes it the case that a person X at t1 is identical with a person Y at t2? This question seeks the persistence conditions of persons: that is, what it takes for the same person to exist at different times. Note that the question about epistemic criteria for determining the persistence conditions for persons must be distinguished from (iv). The question of what it takes for a person to persist over time is different from the question of how to find out whether a person at one time is identical to a person at another time. Epistemic criteria for recognizing personal identity

Introduction

3

over time must not be confused with criteria for identity itself. Generally, of course, epistemic criteria – such as the continuity of psychological and physical features – are reliable signs for tracking personal identity over time. It is easy to imagine, however, that epistemic criteria for metaphysical identity come apart from identity itself. Consider the case of physical disfigurement after a serious accident: in such a case a person can no longer be recognized by her physical appearance, but this does not imply that there are two different persons, one before and one after the accident. This volume is about question (iv). An adequate answer to the question of the diachronic identity of human persons can be presented schematically in the following way: “If x and y are things of kind K, then x is identical with y if and only if x and y stand in the relation Rk to one another” (Lowe 2000, p. 272). If you ask yourself whether the woman at the mall is the girl you knew in high school, then you ask yourself whether you refer to the same human person twice or refer once to each of two different persons. Although the question of diachronic identity is related to questions (i) to (iii), keeping them separate will help avoid confusion. Still, especially question (iii) will play a role in answering question (iv) because the metaphysical nature of an entity determines its persistence conditions. the debate about personal identity One way to introduce the contemporary debate about personal identity is to distinguish two basic kinds of account of personal identity in the sense of question (iv): the complex and the simple view. The complex view analyzes personal identity in terms of simpler relations. The fact that a person persists over time is nothing more than some other facts which are generally spelled out in either biological or psychological terms, or both. That is, the complex view takes talk about “what personal identity consists in” literally. It aims to provide necessary and sufficient conditions for personal identity, thereby reducing it to the holding of basic biological or psychological relations. Whenever these relations obtain, personal identity obtains. The simple view of personal identity, by contrast, denies that a person’s identity through time consists in anything but itself. Biological and psychological continuity may be regarded as epistemic criteria for diachronic identity, but they are neither necessary nor sufficient conditions for personal identity. There are no non-circular, informative necessary and sufficient conditions for personal identity: personal identity consists in nothing other than itself.

4

georg gasser and matthias stefan the biological approach

A natural idea is that a human person’s identity over time consists in biological persistence conditions either of the entire organism or of particular parts of it − for instance the brain. The former approach has become known as animalism (Olson 1997b); the latter might be called the brainbased approach (see e.g. Nagel 1986, p. 40, sketches this view as the empirical hypothesis of the self’s true nature). Animalism assumes that the biological functioning of the human organism − that is, the persistence of the unity and interaction of metabolic processes − is essential for human beings to persist. Accordingly, a person’s identity is no different from the identity of other living things like horses or mosquitoes. Her persistence does not consist in the preservation of the same matter but rather in the preservation of the same organizational biological form, since the matter constituting the organism is continually replaced. For animalism, the identity of a functioning organism and the identity of the person constituted by this organism do not necessarily go hand in hand. Take, for instance, an irreversibly comatose patient. If we assume that the actual performing of higher brain functions is a necessary condition for an individual’s mental state, and if these brain functions are absent in a comatose patient, then the patient is a living and functioning organism, but she is not a person, because this presupposes that she enjoys a mental life of some complexity. According to the biological approach, we are not essentially human persons, but rather human organisms or animals. We can lose the status of personhood while remaining us, because our human organism can continue to be alive. A well-known argument for animalism is the “thinking animal problem.” It starts with the insight that human animals exist. Wherever a human person is, there seems to be a human animal too: wherever you sit, a human animal sits too; whenever you work, so does an animal; whenever you are thirsty, an animal is too. The animal is most intimately related to you, so that it is difficult to tell the difference between you and it. Olson writes: In fact the animal seems to be mentally exactly like you: every thought or experience of yours appears to be a thought or experience on the part of the animal. How could you and the animal have different thoughts? But if the animal thinks your thoughts, then surely it is you. You could hardly be something other than the thing that thinks your thoughts. (Olson 2007, p. 29)

How might one respond to this argument? One response is to deny that there are animals. A second is to say that there is an animal where you are

Introduction

5

but that it is only you and not the animal that enjoys a mental life. Animals are living but non-thinking beings. Finally, one could respond that there is an animal where you are, which has the same thoughts as you, but that nevertheless you are not identical with it. You share a mental life with an animal but not a metaphysical nature, because you are a human person whereas the animal is not. The first option is implausible. There are no good reasons to reject the existence of animals. The second option dissolves into a kind of dualism: one might argue that animals qua biological organisms are unable to think but that their soul is the entity endowed with a mental life. A materialist alternative to this option could instead identify human brains as thinking beings in contrast to animals. Accordingly, a mental life can be attributed only to a specific part of the human animal – the brain – rather than to the animal as such. The third alternative results in what is called the “too many thinkers problem.” If we assume that the human animal enjoys the same mental life as the human person but that they are nonetheless two distinct entities, then apparently the human animal and the human person coincide spatially. Even though ordinary parlance refers to just one entity, there are in fact two. This proposal thus fails to solve the problem of coincidence. If there is a human animal thinking your thoughts as well as you as a human person thinking your thoughts, then why should you assume that you are the human person and not the thinking human animal? There is a serious epistemological problem because you are not in a position to tell which of the two entities you are. In addition, someone might wonder why a thinking human animal should count as an animal and not as a person. What keeps the human animal from being a human person? If this argument for animalism is sound, there is strong reason to think that we are, at least during certain episodes of our existence, essentially animals capable of thought, unless human persons are identical either to mental substances or to brains. This conclusion pertains to the metaphysical nature of human beings, but is closely related to the question of conditions for personal identity: if we are biological organisms, then our identity conditions are those of biological organisms − that is, our identity consists in the continuity of a living body over time. An alternative to animalism is the brain-based approach. It claims that there are certain biological conditions for personal identity which, however, fall short of the entire organism. It starts from the assumption that one part of the body − the brain (or certain parts thereof ) − is of particular importance because it produces the mental life characteristic of being a

6

georg gasser and matthias stefan

person. Damage to the brain can result in personality changes and amnesia, and in severe cases can utterly obliterate the capacity for higher cognitive functions, whereas damage to the limbs, stomach or other organs has no such dramatic effects. Admittedly, we might undergo a personality change as a consequence, but these events would not affect the brain’s capacity for maintaining a mental life. You can even imagine that your brain could be removed from your body and preserved in a functional state by a complex machine. Assuming that your brain works as well after as before its disembodiment, you might still be considered the same person under these artificial conditions. A thought experiment by Sydney Shoemaker (1963) underpins this line of thought. Imagine that the brain of one person, Brown, is removed and transplanted into the body of another person, Robinson. The resulting person, Brownson, has the body of Robinson but the brain of Brown and thus Brown’s whole psychological makeup. Most of us would be inclined to say that the newly created person Brownson is identical with the former person Brown. Derek Parfit formulates this intuition as follows: “Receiving a new skull and a new body is just the limiting case of receiving a new heart, new lungs, new arms, and so on” (Parfit 1984, p. 253). These considerations favor the view that the functioning brain (or a certain part thereof) needs to persist for the human person to persist, but that the entire organism is not required to do so because only a functioning (and appropriately stimulated) brain, and no other organs, is needed for producing an individual’s mental life. Note that the brain-based approach places so much emphasis on the brain because the latter sustains one’s mental life. Herein lies a major difference between the brain-based view and animalism: for animalism, the fact that a human organism can enjoy a mental life is of no importance in deciding identity over time. For the brainbased approach, by contrast, the brain is essential for guaranteeing one’s identity over time because it guarantees the continuity of one’s mental life. the psychological approach As the thought experiment of the brain transplantation between Robinson and Brown shows, some intuitions motivate a link between personal identity and psychological continuity rather than between personal identity and the identity of the brain. Such intuitions are encouraged by another thought experiment involving what Shoemaker (e.g. 1984, p. 108) called a “brain state transfer device”: this device reads the states of a person P1’s brain, writes them into the brain of person P2, and then destroys P1’s brain,

Introduction

7

turning P1’s organism irreversibly into a human vegetable. What becomes of P1? Is she identical to the human vegetable or did she just acquire a new organism thanks to the brain state transfer device? Pace the brain-based approach, P1 ceases to exist once her brain is destroyed. The line of thought pursued in the thought experiment, however, provokes the question whether we should say that personal identity consists in the continuing functioning of the human brain at all. We could imagine that no brain is needed anymore and that anything would do the job as long as it sustains one’s psychological life. If psychological states and their continuity are the mark of personal identity rather than the continuity of the brain (or some other biological fact), then P1 continues to exist by acquiring P2’s body. This approach to analyzing personal identity amounts to the claim that a human person’s persistence consists in a particular constellation of psychological relations over time. It is known as the psychological approach. An advocate of some version of the psychological approach must specify what kind of psychological relations are necessary and sufficient for human persons to persist. Philosophers generally try to spell out psychological continuity in terms of causal connections between earlier and later psychological states, such as remembering earlier experiences, forming and carrying out intentions, and holding beliefs over time. Sydney Shoemaker characterizes this approach this way: Reverting to the “person-stage” terminology, two person-stages will be directly connected, psychologically, if the later of them contains a psychological state (a memory impression, personality trait, etc.) which stands in the appropriate relation of causal dependence to a state contained in the earlier one; and two stages belong to the same person if and only if . . . they are connected by a series of stages such that each member of the series is directly connected, psychologically, to the immediately preceding member. (Shoemaker 1984, p. 90)

On this view, whether a person at t1 is the same as a person at t2 depends on what constitutes an “appropriate” relation of causal dependence between psychological states – however “appropriateness” is specified in detail. We have been led by a series of considerations from animalism, via the brain-based approach, to the psychological approach to personal identity. Whichever version of the complex view one prefers, either way the biological and psychological approaches agree that a person’s identity over time is definable in terms of something other than itself. They differ merely in spelling out what that something is.

8

georg gasser and matthias stefan two problems for complex approaches

The general claim of any complex approach is that a person’s identity over time can be analyzed into necessary and sufficient components other than identity over time itself. This claim comes with problems of its own. This section briefly examines two of these problems, the problem of graduality and the problem of fission. First, the problem of graduality (Noonan 1989, pp. 128–48). If personal identity consists in simpler items such as bodily and/or psychological relations, then we can imagine situations in which these relations admit of degree (Lewis 1976). This leads to the question of what is the threshold that demarcates personal identity from non-identity. Imagine that in a series of operations the parts of P1’s body are discarded and replaced with parts of P2’s body, until all parts of P1’s original body have been replaced with parts of P2’s body. With each operation some of P1’s psychological states are also lost and are replaced by psychological states of P2. By the end, P1’s entire body and consequently all of her psychological states are exchanged with P2’s body and her psychological states so that P1 is identical to P2 in terms of her body and psychology. It is very likely that, after only a few operations, P1 still exists with just minor changes in her bodily and psychological states. At the end of this series of operations, however, one might have the intuition that P1 does not exist anymore but was replaced with P2. The crucial question is: is there a way to indicate the sort of relations needed between two person-stages of P1 so that changes in these relations do not result in her dropping out of existence by being replaced with P2? Is there a precise threshold demarcating P1’s persistence in time? As can easily be imagined, it is unclear how to specify the exact threshold demarcating P1’s existence. It may well always be possible to present examples of deviant continuity relations which leave it undetermined whether or not two person-stages are continuous. It is very likely that complex views go hand in hand with bodily or psychological continuity relations that admit of degree, and it is hard, if not impossible, to specify what degree must obtain to guarantee personal identity. Thus the sort of continuity relations that personal identity consists in is elusive. One could argue that this might only be an epistemic problem for finite minds such as ours, because human bodies and human psychology are extremely complex phenomena. An omniscient being, by contrast, might be able to specify the appropriate sort of bodily and psychological continuity needed for a human person to persist. But who can tell? There are no

Introduction

9

points of reference that a clear demarcation exists. This suggestion is merely speculative and we will not pursue it here any further. Apart from the insight that complex views might provide us with a less specific account of personal identity than we would wish for, there is a more serious objection to these views. Opponents of the complex view point to an intuition to the effect that it does not make sense to think that personal identity can be gradual. They say: whatever tomorrow brings, there is the strong intuition that either P1 will exist or will not; of all the people existing, either she will be one of them or none will be identical with P1. Either of these two states of affairs will obtain and, so the claim goes, the view must be rejected which admits of degrees between these two states. If personal identity is determinate, it cannot be the case that there is just P1 at t1 and just P2 at t2 while being indeterminate whether P1 is identical to P2. Maybe no one can tell whether P1 is identical to P2, but nevertheless the statement about P1’s identity is either true or false. One way to challenge this line of argument is by appeal to well-known puzzle cases such as Parfit’s club (1984, p. 213). Imagine that, for some years, a club exists, then that the regular meetings cease, and a few years later that the members of the same club start to meet again with the same rules and the same name. Someone could ask: have the members set up a new club similar to the old one? Or do they continue the same club which exists intermittently? It seems reasonable to argue that these questions can be answered conventionally. Points can be made in favor of the thesis that the old and the new club are identical, and other points can be made in favor of the alternative thesis that they are not identical. Depending on the conditions of identity one accepts, one can legitimately hold either thesis, for there are no right or wrong answers. It seems obvious that once it is settled which conditions of identity are accepted, it does not make sense to continue to argue about the correct answers for this case. But can we proceed in this way when it comes to human persons? Defenders of the thesis that personal identity is determinate claim it is not absurd to ask whether a correct answer can be given. On the contrary, it would be absurd if a human person, facing this question for herself, thought it sensible to consult a general meeting or a law court. It does not make sense, the argument goes, to assume that (under certain circumstances) a general meeting or a law court could simply decide whether or not a person at one time is identical to a person at another. In contrast to the case of the club, there is a pertinent intuition that whatever decision is made, the possibility remains open that it would be the wrong one. Any decision is made under risk, that is, because things could differ from the way the court

10

georg gasser and matthias stefan

decides. This intuition is particularly strong when one considers the matter from a first-person perspective. David H. Lund, for instance, writes: I am unable to imagine being involved in circumstances under which I am neither fully admitted nor fully excluded. The experiences occurring under these circumstances would have to be something for me if I am to be involved in them at all, but the suggestion that they would be something for me even though it is indeterminate as to whether I am having them seems simply unintelligible when one takes the first-person perspective and reflects upon what it is to have experience. (Lund 2005, p. 229)

Second, the problem of fission (Noonan 1989, pp. 149–68). The complex view does not rule out the possibility that the necessary and sufficient relations for personal identity over time are not one-to-one but one-tomany (Shoemaker 1963; Williams 1970). Imagine the scenario of a person P1 splitting up at t1 into two subsequent distinct persons at t2, P2 and P3. They stand to the original person symmetrically in the identity-defining relation. Both P2 and P3 are fully continuous with P1 in terms of the latter’s psychological and biological characteristics. Neither from the outside nor the inside can it be determined which of the two successors is identical to P1. Although the same identity-defining relations obtain between P1 and P2 as well as between P1 and P3, P1 cannot be identical with both successors at the same time because P2 and P3 are numerically different persons. The proponent of the complex view must have a story to tell about such scenarios. One option is for her to reject the presupposition which is often called the “only x and y principle.” This principle says that two persons, P1 at t1 and P2 at t2, are identical iff (a) they stand in the appropriate internal relationship and (b) there is no other competing person, P3 at t2, who stands in the same relationship with P1 as P2 at t2 (see e.g. Nozick 1981, pp. 29–47). In other words, whether P1 and P2 are identical depends on whether a process of fission occurs. If fission occurs, identity does not obtain; if fission does not occur, identity obtains. However, this line of argument seems to imply that questions of personal identity do not depend only on facts internal to the relation between a person existing at different times. Instead, whether or not P1 exists after t1 depends on the seemingly secondary issue of whether or not another person similar to P1 will exist at t2. The adoption of the “only x and y principle” has been criticized as counterintuitive: what else could a person’s and her successor’s being identical depend on other than facts about themselves? In addition, the “only x and y principle” appears ad hoc, its only ostensible

Introduction

11

purpose being to exclude fission cases from the discussion about personal identity. A related solution to this puzzle could propose that what matters to P1’s survival is not personal identity but simply that someone stands in the appropriate continuity relations. If more than one person does so, then this is no cause for concern but a sign that one’s continuing existence should not be understood in terms of a personal identity rather than an (appropriate) continuity relation. It is not identity that matters, but continuity. As long as at least one person exists in the future who will be related to P1 in the way in which P1 is related to her past selves, P1 continues to exist. Fission represents no particular difficulty to this approach because, unlike the normal one-toone case, in such a scenario the person continues to exist “twice over” and she has no reason not to value the existence of her successors as much as her own existence before fission took place. Since it is not identity that matters in survival, no violation of transitivity of the identity relation takes place in such fission scenarios (Parfit 1984, pp. 245–80). The bottom line of the discussion so far is that no complex account of personal identity is wholly satisfactory. One way to deal with this result is to assume that our intuitive beliefs concerning personal identity are somehow confused or even inconsistent. Another possibility is to assume that the thought experiments lead us astray because they do not represent real possibilities in our world (Wilkes 1988, pp. 8–18). A third possibility is to propose an alternative metaphysical framework which can account for these puzzle cases. So far we have considered only a three-dimensional metaphysical framework. It seems, however, that four-dimensionalism fares better with the problems discussed than its three-dimensionalist alternative.

four-dimensionalism Theorists of personal identity generally hold that a human person exists at different times. How she persists, however, is a matter of metaphysical dispute. Three-dimensionalism claims that a human person, like any other material object, is wholly present at each time that she exists and has no temporal parts. Four-dimensionalism, instead, holds that human persons not only have spatial parts, but successive temporal parts as well. They are temporally extended composites filling up regions of space-time. Ted Sider writes: My spatial parts extend through time like I do. We call them spatial parts because they are smaller than I, spatially speaking; they are “cut out of” me along a spatial dimension. Reverse time and space in this description and we obtain a description

12

georg gasser and matthias stefan

of my temporal parts, which extend through space like I do but are smaller than I, temporally speaking; they are what you get by slicing me along a temporal dimension. (Sider 2001, p. 2)

According to this approach, persons have temporal parts located at different times, in a way that is analogous to having spatial parts located at different places. A person who exists throughout the decade between 1990 and 2000 does so by having connected temporal parts at every time between 1990 and 2000. She extends along not three but four dimensions, being “spread out” over a region of space-time. Persons, like material objects, are conceived as four-dimensional space-time worms whose single parts are united by relations which – depending on the account one favors – can be specified in various ways. John Perry (2008, pp. 7–12), for instance, underlines that the unity relation is of crucial importance for understanding personal identity because it determines which parts belong to the same entity. The identity relation, instead, indicates whether one or two (or more) distinct entities are present. Thus, it has to be kept distinct from identity over time even though both relations are themselves closely related (see e.g. Perry 2008, pp. 7–12). Perry clarifies the distinction between these two relations with the following example: If we want to learn about a baseball game we need to know when events in the game are parts of one single baseball game. If we just knew about the parts of the game, but not about when they belong to one single game rather than to two different ones, we would not possess the concept of a baseball game. Thus, we are looking for the unity relation between the single events of the one baseball game. Unity is a relation between the events of one game specifying which events belong to the same game. When different events or parts stand in the unity relation, they are events or parts of one single entity. Identity, by contrast, specifies whether game A is numerically the same as game B. The unity relation can be understood in various ways. You could have psychological continuity, bodily continuity or whatever unifies spatiotemporal parts into one single person. Depending on one’s preferred account of personal identity, variants of bodily theories of personal identity are as correct as variants of psychological ones. Different accounts of personal identity over time simply define persons differently: whereas one theory employs bodily connections between temporal parts to pick out “bodily” persons, the other theory employs psychological connections to pick out “psychological” persons. Facts about personal identity over time are up to us to determine, since it is up to us to decide which kind of “person” to refer to. Those who employ different unity relations may not

Introduction

13

disagree about the correct approach to personal identity at all. Rather, they may refer to different kinds of persons − persons being composed of personstages related to one another by bodily continuity and persons being composed of person-stages related to one another by psychological continuity. With this ontological framework, four-dimensionalism gets a grip on the problems of fission and graduality (see e.g. Perry 1972). As noted, fission violates the transitivity of identity. Not so the unity relation, which is not transitive. If x and y are parts of A and y and z are parts of B, then this does not imply that x must be a part of B and z be a part of A. Different objects can share some but not all of their parts. The same holds for temporal parts of persons: if two persons share temporal parts P1 and P2 at t1 and t2, then this does not imply that they share their temporal parts P3 and P4 at t3 as well: imagine there is one person composed of the temporal parts P1, P2 and P3, and another composed of P1, P2 and P4, but there is no person composed of P2, P3 and P4. So there is one single person with the temporal parts P1, P2 and P3, and another with the temporal parts P1, P2 and P4. These two persons are numerically distinct: they share temporal parts P1 and P2, and they therefore cannot be distinguished during the times t1 and t2. Fission does not pose a particular problem to a four-dimensionalist account, because there is not one single person dividing into two but two persons parting ways after they have coincided for some time. Four-dimensionalism suggests a solution to the problem of graduality, too (Noonan 1989, pp. 140–8). Graduality is understood as the result of semantic underdetermination and not a result of metaphysical puzzles. Once it is agreed that all there is are spatio-temporal parts, then there is no open ontological question. We only have to specify our concept of the human person: that is, we have to specify the unity relation we pick out with this concept, and then we can state precisely which spatio-temporal parts belong to the same (kind of) person. The upshot of this discussion is that four-dimensionalism’s liberal ontology of objects provides a powerful metaphysical framework for solving the problems that plague complex approaches within a three-dimensionalist framework. Of course, not everybody sees this liberal ontology as a virtue. Some philosophers consider its cost to be prohibitive. There is no space to dwell on this discussion here. It suffices to note that, by construing any filled region of space-time as an object, for some philosophers fourdimensionalism countenances far too many objects which are not robust enough to make sense of our ordinary understanding of an object’s persistence (see e.g. Baker 2007d, pp. 199–217).

14

georg gasser and matthias stefan the simple view

Some philosophers reject all the accounts of personal identity presented so far. They consider them fatally flawed and as a consequence take themselves to be justified in assuming that there are no informative non-circular conditions for personal identity. According to them, personal identity is something so fundamentally basic that it cannot be accounted for in more basic terms such as biological and psychological relations or temporal parts. E. J. Lowe, for instance, puts this account as follows: “The persistence of at least some sorts of things must . . . be primitive or ungrounded, in that it can consist neither in relationships between non-persisting things nor in the persistence of other sorts of things” (Lowe 1988, pp. 77–8). As a simple relation, personal identity does not have other relations as proper parts, and what has no proper parts cannot obtain partly, as this would presuppose that only some parts exist and others do not. Accordingly, personal identity either obtains or does not obtain, and as such it does not admit of degree. Taking personal identity to be ontologically basic does not imply any form of skepticism about the reality of persons and their identity over time. It seeks merely to distinguish between epistemic criteria and conditions for identity over time. Biological and psychological relations are epistemic criteria for justifying the assumption of personal identity, but they are not truth conditions for its obtaining. Even if one knew everything about a person’s psychological and bodily relations, the question of personal identity would remain open. According to the simple view, it is metaphysically possible for there to be two worlds which are identical in their physical and psychological details, except for the distribution of personal identity. Of course, under ordinary circumstances psychological and bodily relations can doubtlessly be considered to be reliable signs for the obtaining of personal identity. Nevertheless this does not justify conflating them with conditions for the obtaining of identity over time because in the light of the simple view there is exactly one such a truth condition, which is the obtaining of personal identity itself. The claim that personal identity is simple implies that it cannot be analyzed because it is not possible to appeal to other entities of a suitable kind for formulating in a non-circular way what personal identity consists in. This does not imply that the simple view is entirely uninformative. The discovery − if it is a discovery − that personal identity is not analyzable is a kind of progress in understanding it. In addition, it should be clear by now why the simple view is mainly a negative thesis in light of which the most

Introduction

15

one can realistically hope to show is that its rivals come with prohibitive costs of their own. Arguments for the simple view may be grouped into arguments from intuition, epistemic arguments and ontological arguments. Arguments from intuition appeal to the intuitive force of the simple view by considering the various counterintuitive problems, such as fission and graduality, that burden the complex view (Chisholm 1976, pp. 110–12; Swinburne 1984, pp. 13–19). The simple view provides a clear answer to the problem of fission: since personal identity is a simple fact, either it obtains or it does not. In the fission scenario the simple view offers three possibilities: P1 is identical with P2 and not with P3, P1 is identical with P3 and not with P2, or P1 ceases to exist and is identical with neither P2 nor P3. Still, however, there is a fact of the matter about which possibility is realized, and therefore there is no need for the simple view to draw on conventionalism to decide whether identity obtains. The same is true in the case of graduality: there must be a definite borderline that demarcates identity from non-identity, because identity, being a simple fact, cannot obtain partly. One peculiar implication of the simple view is that we may well be ignorant about how identity is actually distributed. It may be that we simply cannot tell for certain whether P1 is identical with P2, with P3, or with neither. Though this might be unfortunate, it is an epistemological and not an ontological problem. Human beings are not omniscient, and facts about personal identity in fission and gradual cases might be among the things we do not know. The second type of argument for the simple view appeals to the idea that knowing everything about bodily and psychological properties and their relations would still leave open the question of personal identity, because this question must be answered from a first-person perspective. This type of argument starts from the claim that personal identity is conceivable in the absence of psychological and bodily relations, and moves on to its metaphysical possibility (e.g. Madell 1981, pp. 78–106). Accordingly, there are two conceivable possibilities: first, I might have had a totally different life: for example, living in a different century with a different body and a different psychological makeup. I might, for instance, have lived in seventeenth-century France, been born of different parents and had different memories, intentions, desires and so forth. Nevertheless, so the argument goes, from my first-person perspective it is still conceivable that it is me that lives that life. A similar argument points to changes of body and psychology (see e.g. Swinburne 1984, pp. 22–3). I can conceive of myself as having your body and psychology and of you as having mine. I could also imagine that I might not have existed, but that instead someone else exists

16

georg gasser and matthias stefan

with the same life and body that I actually have. If these scenarios really are metaphysical possibilities, then the obtaining of psychological or of bodily relations is neither necessary nor sufficient for personal identity: there is a possible world in which I exist without the bodily and psychological properties that I actually have, and another in which the bodily and psychological properties I actually have belong to another person. It should be noted that this is an argument against the claim that synchronic personal identity is analyzable. However, synchronic and diachronic identity are not different in kind. Rather, they are instances of the same relation as it obtains either at one time or at different times. Accordingly, if synchronic personal identity cannot be analyzed, then we have every reason to assume that diachronic personal identity cannot be analyzed either. There is another epistemic argument closely related to this one. It argues that whether one has a conceptual grasp of personal identity over time has nothing to do with whether one has knowledge of psychological and bodily relations. In other words, we understand clearly what it means for a person at t1 to be identical with a person at t2 without knowing any bodily and psychological criteria for personal identity (see e.g. Nida-Rümelin 2010). This becomes clear on consideration of epistemically underdetermined cases such as fission scenarios: if there is perfect symmetry between P1 and P2 as well as between P1 and P3, then there are no bodily or psychological criteria available to decide which of the two successors at t2, P2 or P3, is identical with P1 at t1. Nevertheless, one can clearly conceive of the difference between the case in which P2 is identical with P1 and the case in which P3 is identical with P1. In the former case P2 would have P1’s first-person perspective, whereas in the latter case P3 would have it. There is, according to this argument, a conceptual difference between bodily and psychological relations and identity. As a human person, each one of us is able to take the perspective of P1 and from this viewpoint she knows, independently of considerations about the bodily and psychological relations obtaining between P1 and P2 and P1 and P3, what must be the case for herself as P1 to be identical to P3 in contrast to P2 (or vice versa). What is claimed is that one has a clear conceptual grasp of personal identity independently of one’s knowledge of psychological and bodily relations. Nida-Rümelin corroborates her conceptual reflections by arguing that one cannot reasonably give up this understanding of personal identity, since it is deeply rooted in our capacity for first-person thought. She argues that there is good reason to interpret the findings of this conceptual analysis in a straightforwardly realist sense – as long as there are no strong reasons against

Introduction

17

it. There is a difference between bodily and psychological criteria for personal identity and personal identity itself. Although these epistemic arguments resemble each other, they are distinct: the first claims that it is metaphysically possible for a person to have a totally different psychological and bodily composition and still be herself (and the other way around). The second argument says that, even in epistemically underdetermined cases, we have a clear conception of what must be the case for there to be a definite identity relation. A third type of argument for the simple view is ontological (Lowe 1988; Lowe 1994). It aims to show that persons are simple individual substances, and that as such they do not have proper parts. Entities without proper parts cannot have non-circular identity conditions, for then the latter would have to refer to the identity conditions of those entities constituting the entity in question. This would imply, however, that the entity has parts and as such would not be simple. Without delving into the details of the argument for the simplicity of persons (e.g. Lowe 2000, pp. 15–21; Barnett 2010, pp. 161–74), it suffices to note for present purposes that, once one accepts that persons have no parts, it seems plausible to assume that their identity conditions cannot be spelled out in terms different from personal identity itself.

conclusion The section on the simple view completes our presentation of the contemporary discussion about personal identity over time, a relation for which manifold approaches try to account. It is mostly taken for granted that personal identity consists in simpler relations of whatever kind, and that the major task is to specify in detail when personal identity obtains and when it does not. Against the background of this discussion, this volume tackles the more fundamental question first − whether personal identity is analyzable into simpler relations at all. Thus Personal Identity: Complex or Simple? contributes to a better understanding of what it means for human persons to be identical over time.

part i

Framing the question

chapter 1

Chitchat on personal identity David Barnett

“Kid looks like you,” says Jitney. “Kid looks like you,” says Cletus. Jitney and her grown twin brother, Cletus, are cleaning out their mother’s attic. Cletus has found a photograph of a child with a squirrel in one hand, a meatball in the other, and a nametag that reads “Kid.” Cletus and Jitney mull over the photo from the comfort of two ragtag armchairs. “I vaguely remember one of us being called ‘Kid’,” says Jitney. “I might be Kid,” says Cletus. “Kid’s body was different from yours, Cletus. Kid’s body was small and youthful, with flawless skin. Your body is big and old, with spots.” “I might be Kid.” “Kid’s mind was youthful and innocent. Kid lived in the moment, with thoughts of slingshots and bubblegum. You live in the past and future.” “I might be Kid, Jitney.” “You might be, Cletus. After all, you’re just that kind of a guy.” “What do you mean by that?” “You’re the sort of thing that could survive significant physical and mental changes.” “So are you, Jitney.” “Thanks, Cletus.” “Just how significant do you reckon these changes could be?” “That depends on just what sort of thing you are, Cletus.” “How about one of them rapid, complete, body swaps, like in the movie Freaky Friday? Could I survive one of them?” “Not if you are your body. And not if you are one of its organs – say, its brain. If you’re either of those sorts of things, you could not survive having one living body, with one set of organs, one moment, and a numerically distinct living body, with an entirely different set of organs, the next.” 21

22

david barnett

“So what sort of thing am I?” “That’s no easy question, Cletus.” “I’ve got all day and plenty of gin. I’ll think real hard.” “A full day of thinking by you might not be enough. In fact, an eternity of thinking by someone even smarter than you might not do.” “Why’s that?” “Because the question of your nature might not be the sort of question that can be answered by thinking alone. Take one candidate answer: that you are a human brain. Thinking alone can’t even verify that there exist such things as brains, so it surely can’t verify that you are a brain. If you’re a brain, the only way you can know that you’re a brain is to get out of that chair and take a look in that thick skull of yours.” “Already done that.” “And?” “Doctors say there’s a brain.” “What else did they say?” “Doctors say what the brain does seems a pretty good sign of what I’m about to do, what I’m thinking and what I’m feeling.” “Well, then, maybe you are that brain.” “Maybe I’m just a prisoner of that brain, made to be how the brain makes me be.” “Did the doctors find anything else in that skull?” “Nope.” “Then maybe you’re that brain.” “Maybe I’m invisible.” “Right. And maybe you’re a squad of cheerleaders.” “I’m not that kind of a guy, Jitney.” “I was being sarcastic. Is it any less absurd that you might be invisible than that you might be a squad of cheerleaders?” “My being invisible may seem unlikely, but it doesn’t seem impossible. My being a squad of cheerleaders seems unlikely because it seems impossible.” “Look at you, Cletus, a real philosopher, making distinctions.” “And look at us, Jitney: sittin’ here, sippin’ on gin, thinkin’ hard, and we’re two steps closer to knowing what I am.” “How’s that?” “As you said, I’m a real philosopher, a distinction maker, a thinking thing. That’s one. And by your own sarcastic tongue, whatever the doctors say, I’m no squad of cheerleaders. That’s two.”

Chitchat on personal identity

23

“Fine. You’re a thinker and you’re no squad of cheerleaders. But that leaves more than a few options open, including your being a brain – and you’re not going to settle whether you’re a brain without listening to the doctors.” “Maybe the doctors are wrong. Maybe that oversized camera of theirs is broke, and inside my head is something other than a brain.” “Right. And maybe there are no such things as doctors. Maybe your senses have been deceiving you your entire life. Maybe you are, and always have been, dreaming. Maybe I’m just a figment of your imagination.” “Good point, Jitney. You might be that kind of a gal. And I just might be that kind of a guy.” “Sarcasm, Cletus. While it may be less likely that your senses have been deceiving you your entire life than that your doctors are deeply mistaken about what’s in your head, neither idea is worth taking seriously – both are absurd.” “You mean impossible?” “No. I mean very unlikely.” “Then we’ll throw ’em together with the idea of my being invisible; all three seem possible, but unlikely. And we’ll keep ’em separate from the idea of my being a squad of cheerleaders, which seems impossible and thus unlikely. Now, from these chairs, with this gin, let’s get thinking. Let’s take every possibility seriously. Let’s not trust the doctors or even our sensations, past or present. We’ll forget about what’s actually going on out there and think hard on what is and isn’t possible. Then maybe – just maybe – we’ll narrow down the range of things I might be, so that we know somethin’ worth knowing about the sorts of changes I could survive.” “But Cletus, just as things might not be as they seem with your sensory experience, they might not be as they seem with your intellect: what seems to you to be possible might be impossible, and what seems to you to be impossible might be possible.” “Fair enough.” “So why trust your intellectual seemings but not your sensory experience?” “Because my sensory experience takes a back seat to my intellectual seemings. No sensory experience is going to convince me that I am a bunch of cheerleaders. Why? Because it seems impossible, in the first place, for me to be a bunch of cheerleaders. So what seems possible or impossible limits what, by the lights of my sensory experience, seems actual.” “Are you claiming that what you can learn from your armchair, through hard thinking about what’s possible, limits what you can learn from your sensory experience about what is actual, but not vice versa?”

24

david barnett

“I am.” “And what if the best explanation of all your sensory experience, past and present, is that your seemings about what is and isn’t possible are inaccurate?” “Then I’d be in a bit of a pickle.” “Then we’re back to where we started, Cletus. Thinking alone may not be enough to answer the question of what you are. Thinking alone may deliver plenty of seemings about what’s possible. But, first of all, what’s possible, from the perspective of that armchair, regarding your nature, may be too broad to bear in any meaningful way on what sort of thing you are and what sorts of changes you could survive – so you may need to call on your sensory experience to narrow down the possibilities. Second, your armchair seemings about what’s possible may ultimately be undermined by past, present and future sensory experience. So before you draw any hard conclusions, you’d better get out of that ragtag chair and check in with your five senses.” “But if we agree to pay heed to our sensory experience, then we ought to do a whole lot of poking around in the world – something I ain’t up for today.” “Me neither.” “So let’s have a gin-and-thinkin’ day today, and see where it gets us. Tomorrow we’ll start checking our results against as much sensory experience as we can gather up.” “Deal. Today we’ll stick with the following question: if things are as they seem to you, from a perspective in which you’ve set aside all knowledge that depends on the past or present accuracy of your sensory experience, regarding what is and isn’t possible, then what sort of thing might you be and what sorts of changes could you survive?” “Bingo. I’ll start by saying that I’m not just any old thinking thing; I’m a conscious being, the sort that has sensory experience, including the taste of sweet gin. I’m that kind of a guy.” “Not so fast, Cletus. You agreed to set aside the deliverances of your sensory experience. So on what grounds do you know that you have sensory experience?” “I agreed to not trust the accuracy of my sensory experience. I didn’t agree to ignore the fact that I’m having sensory experience. That much I know from the chair.” “Fine. I’ll grant that you’re a thinking, conscious, being. Now, let’s return to the idea of a rapid, complete, body swap – Freaky Friday style. What’s your feeling on it?”

Chitchat on personal identity

25

“Seems possible. Seems that a conscious being – a being of my stripe – could have one living body, with one set of organs, one moment, and a numerically distinct living body, with an entirely different set of organs, the next. So I’ll go with yes, it’s possible for me to survive such a swap. I’m that kind of a guy.” “Slow down, Cletus. It seems to you that you’re a conscious being, right?” “Check.” “And it seems to you that a conscious being could survive a rapid, complete, body swap?” “Check.” “Could conscious beings come in more than one stripe?” “I reckon so.” “Could one stripe – say, an embodied angel – be capable of surviving such a swap, while another – say, a conscious brain – be incapable?” “I reckon so.” “Have you established, from that chair, which stripe you are?” “I reckon not.” “Then you better slow down. Granted that your armchair seemings are accurate, all you know so far is that it’s possible for there to be a sort of conscious being who could survive a rapid, complete, body swap. You don’t know that you are that sort of being.” “Point taken, Jitney.” “Let’s turn next to brain transplants. Imagine there are two conscious beings, Elrod and Cooter. Each has a relatively normal human body. Elrod has two green eyes. Cooter has one brown eye and one blue. The doctors knock Elrod and Cooter clean unconscious, scoop their brains out, and swap each one back into the other’s skull. Elrod’s original brain is now part of a body with mixed eyes; Cooter’s original brain is now part of a body with green eyes. Cooter wakes up and looks in a mirror. What color eyes does Cooter have?” “Green.” “Why do you say that?” “In my experience, people tend not to stray far from their brains.” “You mean in your sensory experience?” “I do.” “So, based on the deliverances of your sensory experiences, you would say that it’s very likely that Cooter goes with his brain?” “Yep.” “And, for present purposes, you have agreed to set aside the deliverances of your sensory experiences?”

26

david barnett

“Yep.” “What’s wrong with you, Cletus?” “I’m drunk.” “Get it together. I’ll forget your first answer and let you try again. For all I’ve said, might Cooter be the sort of thing that could wake up with green eyes?” “Yep.” “And might Cooter be the sort of thing that could wake up with mixed eyes?” “Yep. As I said already, it’s possible for there to be a conscious being that could survive a rapid, complete, body swap. Surely, then, it’s possible for there to be a conscious being that could survive a rapid, partial, body swap – a mere brain swap. For all you’ve said, Cooter might be that kind of a guy.” “That’s better, Cletus. Now let’s talk about the celebrated brain chop for a moment. What are your feelings on it?” “Tell me more.” “Earlene is a conscious being with a relatively normal red-headed human body. Earlene’s not feelin’ herself today. Doctors say she’s got a stormy brain, so they chop it in two, right down the middle. One half they put into a new blond-headed body; the other into a new brunette-headed body. Someone wakes up with blond hair; someone wakes up with brunette hair. Which one’s Earlene?” “Which one acts like Earlene?” “Both.” “Which one claims to remember things that happened to Earlene?” “Both.” “And what about from the inside? Which one seems to remember things that happened to Earlene? And which one feels like Earlene?” “Both.” “Then it’s impossible to know.” “Agreed.” “So I’ll go with neither: Earlene got split into oblivion.” “Where’d that come from, Cletus? Suppose I’m about to flip a fair coin. Will it land heads or tails?” “Impossible to know.” “And by your lights, it follows that it won’t land heads or tails. But of course it will.” “Good point, Jitney. Let’s back up. All I know is that, before the chop, Earlene was a conscious being with a relatively normal human body. I don’t know her exact relation to that body. Could be she was that body. Could be she was its brain. Could be she was merely a prisoner of that body and its

Chitchat on personal identity

27

brain, something like an angel with a human body. If she was merely a prisoner, then, after the chop, she could get the blond’s brain; she could get the brunette’s brain; she could get some other brain; she could be liberated from the material world and sent to Heaven; she could go entirely out of existence. For all you’ve said, there’s some sort of conscious being for which any of these outcomes is possible.” “Sounds right to me. From these chairs, we are in no position to say that Earlene has the blond’s brain; we’re in no position to say that she has the brunette’s brain; and we’re in no position to say that she has neither.” “The sweet taste of progress.” “Progress? So far it seems that just about anything goes when it comes to mixing and matching conscious beings with bodies and their parts – at least if we leave open the proper type of conscious being that’s up for mixing and matching. How this counts as progress toward figuring out what sort of thing you are, and what sorts of changes you could survive, I don’t know.” “Feels like progress to me, Jitney. Let’s not lose our momentum. What’s next?” “Teletransportation. You got a feeling?” “Star Trek style?” “Close. Suppose that Rubyjane is a conscious being with a normal human body. She steps into the entry module of a teletransporter. The module surveys the complete physical state of Rubyjane’s body, destroys the body, and sends the information at the speed of light to an exit module, where a perfect physical duplicate of the body is constructed out of new material. Out walks a girl. Who is it?” “The girl looks just like Rubyjane from the outside, and she feels just like Rubyjane from the inside?” “Right.” “Is teletransportation a normal happening where Rubyjane’s from?” “Let’s suppose so.” “Then I suspect her folk would say the girl who walks out is Rubyjane. And I suspect the girl who walks out would say she’s Rubyjane.” “I suspect so.” “And I suspect you suspect I’ll conclude that the girl is Rubyjane. But I’m reformed, Jitney. I’m not about to make that mistake. For all you’ve said, Rubyjane could be something like an angel with a human body and brain. In which case, when her original body is destroyed, any of a number of things is possible: she could pop out of existence; she could continue to exist in a disembodied state; or she could take on a new body, including the body

28

david barnett

that just stepped out of the exit module. So it’s possible that Rubyjane is the girl who walks out of the exit module, and it’s possible that she’s not.” “Seems right to me.” “Could be this gin, but I’m pretty sure I taste another wave of progress.” “Speak for yourself, Cletus. I’m about to call it quits. From these armchairs, it seems that just about anything is possible. There could be a sort of conscious being who could survive a rapid, complete, body swap; a sort who could survive a rapid, partial, body swap; a sort who could survive a brain chop; and a sort who could survive teletransportation. But all this is consistent with there being a sort of conscious being who could not survive any of these events. And we’re no closer to knowing which sort you are, so we’re no closer to answering your question.” “Knowin’ what we can’t know is knowin’ something, Jitney. We know that, from the armchair, we can’t know who goes where in a brain swap, a brain chop, or a teletransportation, at least not without knowin’ more about the sorts of beings involved in such antics. And we know that we can’t, from the chair, know that rapid, complete, body swaps are beyond my repertoire. That’s a lot of knowin’, in my books.” “Whatever tickles your fancy. Let’s shift gears and talk about mental changes. What’s your take on complete amnesia?” “Do I know you?” “Let’s say that Trixie-Lynn has a normal human body. She swings a baseball bat at a basketball. The bat bounces back into her face, knocking her clean out. Someone wakes up dizzy, with a purple knot between her eyes. She doesn’t know where she is or how she got there. She has no memory of anything that ever happened to Trixie-Lynn. Who is she?” “If I had to bet, I’d say she’s Trixie-Lynn. But this bet would be based, once more, on the fact that folks don’t usually wander far from their brains – a fact I know by my past sensory experience, and I’m not about to make that mistake again. So all bets are off. From the chair, given only what you’ve said, all I can say is that the purple-headed gal might be Trixie-Lynn, and she might be someone else – some lucky lady who happened into Trixie-Lynn’s former body.” “And what if I told you that this gal acts nothing like Trixie-Lynn acted before the accident? This gal is polite and generous. Trixie-Lynn was rude and selfish. This gal is witty, creative, lively, and engaging. Trixie-Lynn was dull, boring, and aloof. This gal excels at math and baseball. Trixie-Lynn couldn’t add two and two, and didn’t know the difference between a baseball and basketball. In fact, Trixie-Lynn’s friends and family say that, ever since the accident, Trixie-Lynn is a totally different person.”

Chitchat on personal identity

29

“Are they drunk?” “No.” “And yet they think that Trixie-Lynn was one person before the accident and a different person after the accident?” “Be charitable, Cletus. They think that Trixie-Lynn is, and always has been, identical to one and only one person: namely, herself. However, they believe that, qualitatively speaking, Trixie-Lynn is radically different since the accident: her memories, personality and abilities are nothing like what they were before the accident. This is upsetting to her friends and family, all of whom formed close relationships with Trixie-Lynn based largely on their appreciation of her personality before the accident.” “You mean their appreciation of who Trixie-Lynn was?” “Loosely speaking – or, rather, qualitatively speaking – yes: I mean their appreciation of the family of mental characteristics by which they – and Trixie-Lynn herself – distinguished Trixie-Lynn from other people in meaningful ways, and on the basis of which they formed intimate relationships with Trixie-Lynn.” “For all practical purposes, Trixie-Lynn is dead to them.” “That’s right. All that they valued in their connections with Trixie-Lynn is gone. And all that Trixie-Lynn valued in her connections to herself at earlier times is gone. For all intents and purposes, Trixie-Lynn is dead.” “God bless.” “God bless, Cletus.” “She didn’t have to die.” “What?” “Sorry, I meant to say that she didn’t have to go changing like that.” “Sounds like you agree with her friends and family that, strictly speaking, Trixie-Lynn survived the accident with her original body, and that it’s not the case that a numerically distinct conscious being has taken over her body. Numerically speaking, the person after the accident is the same as before. Qualitatively, the person is very different. Is that your view?” “I suppose that would be my view if I were a member of the scenario who witnessed the events, or if I were considering the scenario with the aid of my past sensory experience. But I am no such person. To repeat: from this chair, given all you’ve said, it seems possible that the purple-headed gal is Trixie-Lynn, and it seems possible that she’s not. Without the aid of sensory experience, the facts of the story don’t tip the scale one way or the other.” “Then consider a different story, Cletus. This one starts like the last: Trixie-Lynn whops herself unconscious with a bat, and thereafter the person with Trixie-Lynn’s original body has a very different personality from the one

30

david barnett

Trixie-Lynn had prior to the whopping. What’s more, at the time of the whopping, Trixie-Lynn’s mother is across town, in a wicker rocker, on her front porch, in her pajamas, snoring like a walrus.” “Doesn’t change my view.” “There’s more. Some kids are playing rock-ball.” “Like baseball with rocks?” “Yep. One smacks an egg-sized rock straight between the mother’s eyes. The snoring stops; the rocking stops; the dreaming stops; the breathing stops – everything stops. A couple of minutes pass. Then, just as TrixieLynn’s body comes back to life, so too does the mother’s. Only, mentally speaking, the person with the mother’s body doesn’t resemble the mother prior to the event. Her personality is totally different. And yet it’s distinctively recognizable to her friends and family. Guess what they recognize it as?” “No!” “Yep. Trixie-Lynn’s personality, prior to her accident.” “Freaky.” “The gal with the mother’s body is now rude, selfish, dull, boring, aloof, mathematically inept and unable to distinguish a baseball from a basketball. Did I mention that the gal with Trixie-Lynn’s body has the personality the mother had prior to the accident? And we’re not even to the freaky part yet.” “Let’s get our freak on!” “Alright, Cletus: the gal with the mother’s body appears to remember all sorts of events that Trixie-Lynn experienced prior to her accident. And these appearances ain’t but a stone’s throw from the truth.” “No!” “They’re as accurate as anyone’s.” “And the gal in Trixie-Lynn’s body? Does she seem to remember what the mother experienced prior to her accident?” “Yep. Last thing she seems to remember from before the accident is falling asleep in a rocking chair on a porch.” “Who do these gals think they are?” “Trixie-Lynn and her mother. That’s who. They think they’ve swapped bodies, Freaky Friday style, only their apparent swap happened on a Sunday, not a Friday.” “Freaky Sunday. And what do their friends and family think?” “Most are body swappists. But a small group of dissenters maintain that body swapping is impossible and unscientific.” “How do they explain the fact that each gal seems to remember things that really did happen to the person with the other body prior to the accidents?”

Chitchat on personal identity

31

“Coincidence. Remarkable coincidence.” “So they deny that these are genuine memories?” “Yep. They say that each gal appears to remember experiencing all sorts of events that she never experienced. By remarkable coincidence, for each apparent memory, the other gal did in fact experience an event quite like the one apparently remembered. Trixie-Lynn seems to remember falling asleep in a rocking chair just before the accident, even though she never experienced this. By remarkable coincidence, her mother did experience an event of just this sort.” “Seems these dissenters have some explaining to do.” “They do. But so do the body swappists. How in the world – the natural world – do two people just swap bodies like that?” “Maybe the people are like angels – non-physical souls – with human bodies. Then there wouldn’t be much preventing such a swap, except maybe some angel-physical laws of nature.” “Or maybe they are like restaurants. Restaurants can swap buildings without a hitch. One day Denny’s is here and Taco Bell is there; the next Taco Bell is here and Denny’s is there. No big deal. No souls. No mystery. Everything is natural. Except the meat.” “Mmm . . . meat . . .” “Focus, Cletus. Are you a swappist on this one?” “If I were a member of the scenario who witnessed the events, or if I were considering the scenario with the aid of my past sensory experience, then there’s a good chance I’d be a swappist, as the swappin’ hypothesis would explain a lot. On the other hand, I’d be open to a remarkable coincidence hypothesis, with a no-angels clause, as it sure would simplify things. I might even want to hear more about your people are like restaurants idea, since it might make for a simple story too.” “But you’re not a member of the scenario, and you’re not considering the scenario with the aid of your past sensory experience. You’re a chair-bound, gin-soaked, amateur thinker. So what’s it gonna be?” “Same as usual: from this chair, given all you’ve said, it seems possible that the purple-headed gal is Trixie-Lynn; it seems possible that she’s Trixie-Lynn’s mother; and it seems possible that she’s neither. The same goes for the gal with the older body: could be the mother; could be Trixie-Lynn; could be neither.” “Love this progress.” “Feels good, doesn’t it?” “Sarcasm, Cletus. You’re no closer to knowing what you are and what sorts of changes you could survive than when we started. The most you

32

david barnett

know, even granting the accuracy of your armchair seemings, is that there could be a conscious being who could survive a rapid, complete, body swap; one who could survive a rapid, partial, body swap; one who could survive a brain chop; one who could survive a teletransportation; and one who could survive a rapid, complete, personality change. But all this is consistent with there being a sort of conscious being who could not survive these sorts of events. And we’re no closer to knowing which sort you are, so we’re no closer to answering your question.” “Slow down, Jitney. What sort of conscious being could not survive a rapid, complete, personality change?” “Suppose Trixie-Lynn is nothing but a collection of her beliefs, sensations, memories and personality traits. If that’s all she is, then what sense is there to the idea of her surviving a rapid, complete, change in beliefs, sensations, memories and personality traits?” “But you said Trixie-Lynn was a conscious being.” “I did.” “How in the world could a collection of features be a conscious being? How could a collection of sensations itself experience the sensation of getting hit in the face with a baseball bat?” “Just playing Devil’s advocate, Cletus. I agree that the idea seems rather preposterous on its surface, but maybe it’s a possibility nonetheless, in which case, you might be that sort of a guy. And if you are that sort of a guy, then perhaps you’re not the sort who could survive a rapid, complete, personality change.” “I’m startin’ to see what you mean about the lack of real progress. We need to narrow the possibilities, Jitney.” “You ready to get out of that chair and see what you’re made of?” “No, ma’am.” “Then how are we going to narrow the possibilities?” “Let’s quit focusing on the positive. It’s overrated. Let’s focus on the negative. Rather than asking what’s possible, for some conscious being or other, let’s ask what’s impossible, for any conscious being.” “Sounds like six of one, half-dozen of the other, to me.” “What I mean, Jitney, is that we should stop assuming that, when it comes to conscious beings, anything goes: it’s possible for them to be human bodies, human brains, non-physical souls, restaurant-like things, or even collections of features. Maybe not anything goes. Maybe it’s impossible for them to be restaurant-like things, souls or collections of features.” “Instead of asking what’s possible regarding the changes a conscious being could survive, on the assumption that just about anything is possible

Chitchat on personal identity

33

regarding the nature of conscious beings, you want to back up and question that assumption, by asking what sorts of things no conscious being could be. Do I have it right, Cletus?” “Bingo. If it turns out to be impossible for any conscious being to be, say, a restaurant-like thing, that’s one less possibility for this conscious being.” “So what sorts of things could no conscious being be?” “Let’s start with a squad of cheerleaders. No conscious being could be one of those. I don’t need to know what stripe of conscious being I am to know that I’m no squad of cheerleaders. From this chair, by pure thinking, I know I’m not that kind of a guy.” “You’re a true visionary.” “So you agree that it’s impossible for a conscious being to be a squad of cheerleaders?” “I agree; I’m just not impressed.” “Let me tell you a story, Jitney. There’s a squad of cheerleaders. The end.” “Now I’m impressed.” “Given only what I’ve said, can you know, from your chair, without the aid of any past or present sensory experience, that you’re not that squad of cheerleaders?” “I can.” “But I haven’t even said what team they’re cheering for.” “Doesn’t matter, Cletus. I’m not their squad.” “I haven’t told you what color their uniforms are.” “Doesn’t matter.” “But there’s so much I haven’t told you, Jitney. Maybe you should hold your judgment until you find out whether they have human bodies.” “Doesn’t matter. I’m not their squad.” “You don’t know their number, size or location.” “Doesn’t matter, Cletus.” “For all I’ve said, they could number in the billions.” “Fine.” “They could be really small.” “Fine.” “They could live in your skull.” “Fine.” “They could have spooky-shaped bodies.” “Fine.” “They could be caught up with one another in a twisted, complex, web of relations . . .” “I don’t care, Cletus. I’m not their squad!”

34

david barnett

“. . . while they’re dead!” “Now that’s different. Why didn’t you ask me in the first place whether I might be a squad of billions of miniature, talented, zombie cheerleaders, living happily inside my skull?” “If they’re happy, they’re not dead. And if they’re dead, they’re not happy, Jitney.” “You’re missing the point. There’s no need for you to fill in any details of your original story. You’re wasting your breath. All I need to know is that there is a squad of cheerleaders. I don’t need any more details to know that I’m not that squad.” “That’s progress.” “How so?” “If you can know that you’re not that squad of cheerleaders without knowing the number, location, or skillset of the squad, without knowing whether its members are dead or alive, miniature or gigantic, and without knowing what their bodies are like, then surely the fact that they are cheerleaders is irrelevant to what you know.” “Seems right. Could be a squad of child soldiers, matchbox cars or dead ants for all I care. I’m no squad of anything, Cletus. I’m not that kind of a gal.” “Then you’re no squad of neurons.” “And I’m no squad of sensations, beliefs or personality traits.” “And you’re no squad of cells.” “Or atoms.” “Or living organs.” “Or particles.” “That’s a lot of possibilities out the door.” “I do taste progress, Cletus.” “If a brain is nothin’ but a squad of neurons, caught up with one another in a twisted, complex, web of relations, then I know from this chair that I’m no brain. And if a human body is nothin’ but a squad of organs or cells, caught up with one another in a twisted, complex, web of relations, then I know from this chair that I’m no human body.” “Slow down, Cletus. We agreed from the start that we might be brains or bodies. It certainly seemed possible. So why think the results of our cheerleadercentered investigation are any better off on this score than how things seemed to us originally?” “Maybe our original seemings were gotten from thinkin’ of brains as big lumps of whatnot, and forgetting the fact that lumps of whatnot are often nothin’ but squads of tiny whatnots. If we think of brains as squads of tiny

Chitchat on personal identity

35

whatnots, caught up with one another in a twisted, complex, web of relations, those original seemings seem a lot less vivid. Ask yourself, Jitney: does it really seem possible that are you a squad of other things?” “It does not, Cletus.” “Hence the progress. Originally, something seemed possible. On reflection, it no longer does.” “Of course, without paying heed to our sensory experience, we can’t know what human brains and bodies are made of, or whether they even exist. What we can know is that, should they exist, there’s a good sense in which they are nothin’ but squads of things, caught up with one another in a possibly twisted, possibly complex, web of relations.” “And we know we’re not squads.” “So we know we’re not brains or bodies.” “Or restaurant-like things – squads whose membership can change rapidly and radically.” “Still, Cletus, there’s a lot we don’t know we’re not.” “Like pure and partless angels. They ain’t squads of nothin’.” “That’s right. We might be partless angels.” “Simple souls.” “Well, some souls are quite fanciful, Cletus.” “I mean simple in composition, not quality.” “Like quarks. We might be quarks.” “What’s that, Jitney?” “A little bitty physical particle.” “Ain’t a squad of other particles?” “Nope. Pure and partless, simple and plain.” “You learned that in science?” “Yes sir.” “No sensory experience allowed.” “Look who’s drunk now. I retract. For all we know from these chairs, quarks are simple, and for all we know, they’re squads.” “So we might be simple, non-physical, beings – like partless angels – and we might be simple, physical, beings – like partless physical particles.” “Or so it seems from these chairs.” “Correct. Tomorrow we’ll help ourselves to as much sensory experience as we can gather up.” “And the next day too.” “And the next.” “And if the best explanation of all that experience is that things just ain’t as they seem, from these chairs, regarding what is and isn’t possible, then

36

david barnett

we’ll begin the hard job of negotiating between what seems actual by our senses and what seems possible by our intellects.” “But maybe what seems possible by one line of intellectual thinkin’ would seem impossible by another, and vice versa. Maybe if we thought a new line of thought, it would no longer seem impossible for a conscious being to be a squad.” “Maybe, Cletus. But unless you’ve got a good reason for thinking so, I don’t have a good reason for taking this maybe seriously.” “Think about it this way, Jitney. We’ve only thought about the question of whether it’s possible for a conscious being to be a squad by one method: asking ourselves whether we might be squads. Our feeling on this question is a firm no: no matter what sort of conscious being we turn out to be, we ain’t no squads. But it’s just a feeling. There’s no argument.” “Speaking for myself, it’s a very strong feeling.” “But maybe it’s worth hittin’ the question with a different method, from the chair, to see whether we get a different answer. Instead of asking ourselves directly whether we might be squads, let’s ask ourselves whether the sort of feature had by every conscious being is the sort of feature that could be had by a squad.” “What sort of feature is that?” “Experiential. Like feeling sad.” “I see where you’re going, Cletus. You like French fries?” “Mmm . . .” “You like McDonald’s French fries?” “Mmm . . .” “You like their taste, don’t you Cletus?” “Mmm . . . taste . . .” “That’s an example of an experiential feature, right?” “Mmm . . . experiential feature . . .” “Wake up, Cletus!” “Right, Jitney. The question is whether a squad is the sort of thing that could have that sort of feature. Could a squad experience the taste of French fries?” “How do squads get their features?” “From their members, I suppose.” “And how do squads get their features from their members?” “The old-fashioned way?” “Nope. Not by hard work. Squads don’t have to lift a finger to get their features from their members. They get them automatically.” “The way McDonald’s fries get their golden brown color from that boiling beef-flavored lard?”

Chitchat on personal identity

37

“Nope. That is automatic in a different sense. The boiling lard causes the fries to change color, without anyone intervening. The squad members’ features do not cause the squad to have its features. Rather, they constitute them. The location of a squad of cheerleaders is constituted by the location of its members.” “And its weight is constituted by the weight of its members.” “And its shape is constituted by the shapes and spatial relations of its members.” “If God wants to give a squad a feature, He can only do so by giving other things – its members – certain features.” “What about the feature of being named ‘Ed’? Couldn’t God name the squad ‘Ed’ without laying a finger on any of the squad’s members?” “Good point, Jitney. I’m talking about inner features, the ones that give the squad its inner glow.” “Let’s call those intrinsic features, Cletus.” “Fine. If God wants to give a squad an intrinsic feature, He can only do so by giving other things – its members – certain features.” “Agreed.” “But conscious experience is an intrinsic feature. And giving other things features is no way to give someone the gift of conscious experience.” “Unless doing so causes that person to have the experience.” “Bingo. If you experience the taste of McDonald’s French fries, you do not qualify as having this experience by virtue of the fact that other things in the world have various features. This fact might cause you to have the experience – it might bring about the separate event of your having the experience – but it doesn’t itself qualify you as having the experience. If you experience the taste of McDonald’s French fries, you qualify as having the experience directly.” “Agreed, Cletus. Other things being the way they are may cause me to have my experience, but it can’t constitute it. Other things having various features and standing in various relations is one thing . . .” “Your having an experience is another.” “Agreed.” “So this confirms our original feeling that no conscious being could be a squad of anything. Squads earn their intrinsic features only indirectly: there must exist a number of things, namely, the members of the squad, none of which is individually identical to the squad, whose features jointly constitute the intrinsic features of the squad.” “You sure do sound smart, Cletus.” “By contrast, conscious beings have at least some of their intrinsic features, namely, their experiential features, only directly: they have conscious

38

david barnett

experience, not by virtue of other things having various features, but simply by virtue of directly experiencing sensations, such as the taste of French fries.” “If God wants to give me the taste of French fries, He can do it directly.” “No need for Him to dilly dally around with other things. To be sure, His dilly-dallying around with other things could cause a distinct event, namely, your experiencing the taste of French fries. But it could not constitute it.” “Or so it seems from the chair.” “Hallelujah, Jitney.” “Let’s recap our day’s fruit.” “We agreed to set aside the deliverances of our sensory experience, past and present.” “But not the sensory experience itself.” “Correct. You then took me on a fantastical journey.” “I told you some stories, Cletus.” “Right. And, from my chair, by hard thinking alone, given only what you said, I asked myself what seemed possible and impossible.” “And what seemed possible?” “Seemed possible for there to be a conscious being who could survive a rapid, complete, body swap; one who could survive a rapid, partial, body swap; one who could survive a brain chop; one who could survive a teletransportation; and one who could survive a rapid, complete, personality change.” “I wasn’t impressed.” “Because you said this was consistent with there being a sort of conscious being who could not survive these sorts of events, and that we were no closer to knowing which sort of conscious being I was, so we were no closer to answering my original question: what sort of thing am I and what sorts of changes could I survive.” “At this point, nearly anything seemed possible regarding the nature of conscious beings. It seemed that a conscious being could come in just about any stripe.” “A human body.” “A brain.” “A soul.” “A restaurant-like thing.” “Even a collection of features.” “Those were some strange stripes, Cletus.” “Then I had me an idea.” “I didn’t think you had it in you, Cletus. I really didn’t.” “We’d think real hard on what sorts of thing no conscious being could be.”

Chitchat on personal identity

39

“A squad of cheerleaders.” “A squad of dead cheerleaders.” “A squad of tiny dead cheerleaders.” “Caught up in a twisted, complex, web of relations with one another.” “And with the outside world.” “A squad of molecules.” “A squad of cells.” “A squad of features.” “A squad of organs.” “A squad of anything.” “A squad-like thing, like a human body.” “Or a brain.” “Or a restaurant-like thing.” “Then we surveyed the damage. Which possibilities were left, Jitney?” “A non-physical simple soul, like a partless angel.” “Or a physical simple particle, like maybe a quark.” “Rednecks are simple.” “And snobs from Boulder.” “All conscious beings are simple.” “I’m simple, Jitney! I’m that kind of a guy.” “Then we confirmed this chair-bound feeling of simplicity with an argument.” “Squads get their intrinsic features indirectly, from the features of other things.” “Conscious beings get at least some of their intrinsic features directly.” “They experience their sensations directly.” “No other way to experience a sensation.” “So, I’m simple. But am I physical?” “Can’t know from the chair.” “I suppose not, Jitney.” “And what sorts of changes could you survive?” “If I’m not physical, it’s clear that pretty much anything goes: I could survive a rapid, complete, body swap; a rapid, partial, body swap; a brain chop; a teletransportation; and a rapid, complete, personality change. At least this is so if we set aside any physical to non-physical laws of nature.” “You’re an impressive guy, Cletus. And what if you are physical?” “Then I suppose we’d need to pause to think about what it would be to have a body in the first place.” “If you’re not physical, then you have a particular body just in case you’re connected with that body in the right sort of intimate way.”

40

david barnett

“It just feels like my body.” “Perhaps.” “It feels under my authority.” “Perhaps.” “And yet, I feel like its prisoner: damage to the body is damage to my mind, whether I like it or not.” “Perhaps, Cletus. It’s hard to say what exactly that intimate connection is.” “And if I’m a simple physical particle?” “Then, in some sense, you are your body; you’re a tiny, simple, physical, body of one.” “That’s doubletalk, Jitney. Even if I’m a simple physical particle – a physical body of one – I could still have a larger body – a body of many – in the sense we just spoke about: I could have that special intimate connection with a large body of matter – say, a human body.” “And would you be a component of that body?” “Maybe, maybe not. Either way, I could have that intimate connection.” “Seems you could, Cletus.” “And since I wouldn’t be identical to that larger body, I’d be capable of having such intimate connections with other bodies, too, one after another.” “You’re that kind of a guy, Cletus.” “Whatever my stripe – simple physical or simple non-physical – I could survive a rapid, complete, body swap; a rapid, partial, body swap; a brain chop; a teletransportation; and a rapid, complete, personality change.” “Or so it seems from the chair.” “So it does, Jitney.” “Tomorrow we’ll take our philosopher hats off and put our mosquito suits on.” “We’ll get our hands dirty.” “We’ll look under rocks.” “And scalps.” “And if we find that our sensations don’t jibe with today’s result . . .” “If their best explanation is that things ain’t what they seem from the chair . . .” “Then we’ll start negotiating between our sensory and our intellectual experience.” “And if those negotiations end with the verdict that conscious beings like you and me are in fact squads of tiny, dead, strangely bodied cheerleaders . . .” “Or squads of tiny, strangely bodied squads of even tinier more-strangely bodied things . . .” “Or even squads of anything . . .”

Chitchat on personal identity

41

“Then I’ll be darned, Cletus.” “And I’ll be lookin’ for a new theory of the sorts of changes I could survive.” “You may not need to look too far. If negotiations end with the verdict that you’re a human body, then you’ll know that you could survive whatever changes a human body could survive. And if they end with the verdict that you’re a human brain, then you’ll know that you could survive whatever changes a human brain could survive.” “But maybe negotiations will take a turn for the bizarre. Maybe respect for our sensory experience will lead us to conclude that we’re squads, and respect for our intellectual experience will lead us to maintain that it’s possible for us to survive rapid complete body swaps, rapid partial body swaps, brain chops and teletransportations.” “What sort of a squad could survive those things?” “Funny you ask, Jitney. A four-dimensional, choppy, space-time worm, for one.” “I don’t even need to know what that is to know that negotiations shouldn’t take that turn. Think about it, Cletus. If respect for our sensory experience leads us to conclude that we’re squads, then it ought to lead us to conclude that things ain’t how they seem from the chair, regarding what is and isn’t possible. And if they lead us there, then we ought not have any respect for our positive intellectual feelings on the possibilities of body swaps, brain chops and teletransportation.” “Sounds like you never learned the art of negotiation, Jitney. It’s all about give and take. If respect for our sensory experience leads us to conclude that we’re squads, then we ought not have any respect for those intellectual feelings that don’t fit with our being squads, such as the feeling that it’s impossible for any conscious being to be a squad of tiny, dead cheerleaders. But there’s no reason to throw the baby out with the bathwater. We could still respect those intellectual feelings that do fit with our being squads, such as the feeling that it’s possible for us to survive body swaps, brain chops and teletransportation.” “Sounds like you never learned the art of housekeeping, Cletus. If the baby is a perennial source of dirty water, there’s only one way to clean the water: throw out the baby.” “Sounds cruel.” “The baby is the conception that you have of yourself and all other conscious beings from the armchair, prior to testing it against your sensory experience. Capiche?” “Mmm . . . capiche . . .”

42

david barnett

“That conception is of a simple thing, something that is not composed of other things. That conception is what generated your feeling, in the first place, that no conscious being is a squad of anything, and, in the second place, that you’re the sort of thing that could survive a rapid complete body swap, a rapid partial body swap, a brain chop, teletransportation, or a radical, complete, personality change.” “Agreed.” “So any reason you have to doubt that conception is a reason to doubt those feelings generated by it. If you get out of that chair and find that your sensory experience provides you with a strong enough reason to reject the idea that you’re simple, then you’re obligated to stop listening to any of your feelings based on your prior acceptance of that idea, including your feelings that you could survive these changes. Capiche?” “Yes, please.” “If you end up throwing out the dirty bathwater – the idea that no conscious being is a squad of anything – then you’d better throw out the filthy baby that’s dirtying the water – namely, your conception of conscious beings as simple – and all the dirty water that baby is generating, including your feelings that you could survive those changes. How’s that for negotiation?” “Respect, Jitney. If we ever get to the point of negotiating away our armchair conception of ourselves as simple, we’ll toss out all the feelings generated by that conception. Which is not to say that we’ll positively reject the deliverances of those feelings, but only that we’ll not count those feelings as reasons to accept their deliverances.” “If it comes to that point, and the question arises which squad of things are we, we’ll address that question without giving the slightest voice to that dirty baby.” “I don’t trust that baby.” “So far, this is all hypothetical, Cletus. So far we’ve not found a semblance of a reason to turn our backs on the baby. We’re still in our chairs, setting aside sensory experience. It would take a heap of powerful and unexpected sensory experience to bring us to the negotiating table in the first place. And even then, there’d be a ways to go before throwing out the baby.” “It just doesn’t seem possible that we could get to that place, Jitney. How in the world could I be a squad?” “In some sense it seems possible; in some sense it doesn’t. Surely there’s a chance our armchair seemings are off the mark, just as there’s a chance you’ve been dreaming your entire life. In that sense, it seems possible that the baby is dirty and ought to go. In another sense, I agree, it doesn’t seem

Chitchat on personal identity

43

possible. That’s the sense in which it seems that no conscious being, of any stripe, could be identical to a squad of things.” “I could go for a squad of golden French fries right now.” “Sounds like something Kid would have said.” “I might be Kid, Jitney.” “You might be, Cletus.”1 1

For helpful comments on an earlier draft of this chapter, I am grateful to Chris Heathwood, Brad Monton, Michaela Markham McSweeney, Bob Pasnau, and Alex Skiles. I am also grateful to the participants of my Spring 2010 Graduate Seminar on Identity for helping me as I developed some central ideas for this chapter.

chapter 2

In search of the simple view Eric T. Olson

simple and complex views We tell our students that accounts of personal identity over time fall into two broad categories. First there are “complex” views according to which personal identity consists in something else: when there is a fact about our identity or non-identity, there is some other, deeper fact that underlies and is responsible for it. There are familiar debates about what these deeper facts are – whether they have to do entirely with brute physical continuity, for example, or with some sort of psychological continuity, perhaps with a physical constraint. The second category, the “simple” view, rejects all this, and denies that our persistence or identity over time consists in anything. A certain past or future being is or is not you, and that is all there is to be said. Nothing further makes it you or not you. The identity or non-identity of an earlier person and a later one is already as deep as it gets. This taxonomy is shown in Figure 1. Advocates of simple and complex views are supposed to disagree about something more fundamental than those who differ over what personal identity consists in, namely, about whether it consists in anything at all. And because the simple view just denies what complex views assert, the two categories must be both exclusive and exhaustive: any account of personal identity over time must be either simple or complex and not both. If this is right, it ought to be possible to say what the two sides disagree about. What proposition is it that friends of complex views accept and friends of the simple view deny? What do you have to believe in order to accept a complex view, and what belief (or lack of belief) characterizes the simple view? This question is surprisingly hard to answer. The trouble is not merely that there are hard cases – views of personal identity that resist classification as either simple or complex – but that no answer even gets all the easy cases right. And those proposals that best approximate the traditional boundary between simple and complex lack the interest and importance that the boundary is 44

In search of the simple view

45

Views of personal identity over time

Complex views

Psychologicalcontinuity views

The simple view

Brute physical continuity views

Fig. 1: A taxonomy of views about personal identity over time.

thought to have. If there is such a doctrine as the simple view, it is certainly not what we thought it was. The debate over personal identity has been systematically misdescribed. After explaining why I believe this, I will say something about how the debate might be better characterized.

preliminaries First some brief points of clarification. There are three assumptions commonly taken for granted in discussions of personal identity that I am not going to make, rendering some of my formulations unusual. One is that if the identity over time of any person consists in something, then necessarily the identity of every person consists in something. A second is that the identity of all people whose identity consists in anything must consist in the same thing. I want to leave open the possibility that our own identity consists in something but the identity of certain other rational, intelligent beings consists in something else, or in nothing at all. (It may be, for instance, that we are biological organisms that persist by virtue of some sort of brute physical continuity but there is also an immaterial, personal god for whom this does not hold.) For this reason I will speak of the identity of human people, not of people generally. Third, I will not assume that a person must always persist as a person. I leave open the possibility that each of us starts out as an unthinking embryo and may end up in a vegetative state, and that we do not count as people at those times. So I will speak of the identity or non-identity of a human person existing at one time to a thing – person or not – existing at another time.1 If these traditional 1

For further discussion of this point see Olson (1997b, pp. 22–7).

46

eric t. olson

assumptions turn out to be true, my formulations will be equivalent to the usual ones and the difference will do no harm. A terminological note: although I have been speaking of “the simple view,” there may be more than one view about personal identity that is not complex. In that case we can take “the simple view” to name the proposition that some simple view or other is true. Likewise for “the complex view.” I take the simple view and its rivals to be views about personal identity over time. There are other views about personal identity broadly speaking that have certain affinities with this one. For instance, Wiggins says that we cannot give a complete account of what it is to be a person: personhood is, at least to some extent, primitive and indefinable (Wiggins 1980, p. 171). Or someone might deny that there are any non-trivial conditions for our “transworld identity” – that is, for how we could have been. As these claims appear to be independent of the simple view of our identity over time (they are answers to different questions), I will say nothing about them. There are also claims associated with the simple/complex debate that really are about our identity over time, but not about whether it consists in something other than itself. One is that our identity over time must always be determinate: there could never be a being existing at another time that was neither definitely you nor definitely not you. Another is that the facts of personal identity over time are somehow up to us to decide rather than to discover. To my knowledge, all those said to be advocates of the simple view deny both that our identity can be indeterminate and that it can be up to us to decide – though some philosophers said to hold complex views say the same.2 As I have been unable to discover any important connection among these issues, I will set them aside.

grounding and criteria Complex views are supposed to imply that if a person existing at one time is (or is not) numerically identical with a being existing at another time, something must make this the case – something beyond the mere fact that they are (or are not) the same. There seem to be two things that this might amount to. One has to do with how our identity over time relates to other things: advocates of the complex view think it consists in or depends on or holds by virtue of something other 2

Noonan (2003, ch. 6) accepts that identity statements can have an indeterminate truth-value, but denies that there can be indeterminacy of identity itself. Chisholm (1976, pp. 108–11), Olson (1997a), and Merricks (2001b) argue against “identity voluntarism.”

In search of the simple view

47

than itself. Not that a future being’s having inherited my mental states in a certain way (for example) causes him to be me, in the way that stubbing my toe causes me pain. The dependence is logical or metaphysical. So: Whenever a human person existing at one time is identical (or non-identical) to a being existing at another time, this consists in something else.

Call this the grounding claim. Alternatively, personal identity might not consist in anything else: it might be “brute” or “primitive.” An intermediate position is also possible: that when a human person existing at one time is identical or non-identical to a being existing at another time, this sometimes consists in something else and sometimes does not. There is no consensus about whether this would count as simple or complex. The second thought is that according to complex views there is a criterion of identity for human people – not an evidential but a “constitutive” criterion, giving conditions individually necessary and jointly sufficient for a human person to persist from one time to another (“persistence conditions”). Consider this formula: Necessarily, if x is a human person at time t and y exists at another time t*, x = y if and only if . . .,

where “=” expresses numerical identity. The thought is that there is a true completion of the formula that is not trivial or otherwise degenerate. Call this criterialism. Again, intermediate positions are possible. One is that there is a criterion of identity for some kinds of people and none for others. Another is that there are certain conditions necessary for a person to persist, and perhaps even certain sufficient conditions, but no set of conditions individually necessary and jointly sufficient (Merricks 2001a, pp. 195–6).3 In fact every supposed advocate of the simple view that I know of accepts some necessary conditions for personal identity: for instance that a thing existing at another time can be me only if it is not then a stone. If there are any modal truths at all, one of them is surely that it is impossible for something to be a human person at one time and a piece of basalt at another time. (Or if one of us really could come to be made entirely of stone, none of us could become abstract like the number seven.) But this sort of thing does nothing to suggest any non-trivial sufficient condition for our identity over time, or any way of completing the formula. 3

Noonan (2003, p. 97) calls such views complex. If I am right to say that everyone accepts some such necessary conditions, this implies that no one holds a simple view.

48

eric t. olson

What is it for such a criterion to be non-trivial? Here are some that are trivial: Necessarily, if x is a human person at t and y exists at t*, x = y iff x = y. Necessarily, if x is a human person at t and y exists at t*, x = y iff an omniscient being would believe that x = y. Necessarily, if x is a human person at t and y exists at t*, x = y iff whatever is true of x is true of y and whatever is true of y is true of x. Necessarily, if x is a human person at t and y exists at t*, x = y iff x and y have the same individual essence (where a thing’s individual essence is a property that it has essentially and that no other thing could possibly have). The truth of these claims should not vindicate criterialism, and no friend of the simple view would deny them. They are not “proper” criteria of personal identity. Why not? Well, they are entirely uncontroversial. And they are true not because of the nature of personal identity, but because of the nature of omniscience or individual essences or identity in general. Analogous principles apply to all objects: the result of replacing the term “human person” with any other count noun would be equally true and uncontentious. So the criteria tell us nothing about personal identity as such. That seems to account for their triviality. They are also uninformative, in that we could not know whether the conditions were satisfied without already knowing whether identity holds. If Buggins suffers severe and irreversible brain damage of a certain sort, we could not discover whether the resulting being lying insensible on the hospital bed had Buggins’ individual essence and use this to work out whether it was Buggins. Of course, an informant might tell us whether it had Buggins’ individual essence; but the original source of the information about essences could not have obtained it before knowing who was who. What is more, the examples fail to support the connection we expect to find between a criterion of identity and an account of what grounds it. Obviously x cannot be y because x is y. Nor can it be the case because an omniscient being believes it. The dependence is the other way round: whether an omniscient being believes something depends on whether it is the case. Likewise, whether everything true of x is true of y depends on whether x is y, as one of the things true of x is whether it is identical to y. Similar remarks go for the other examples. These criteria cannot tell us what makes x and y identical or non identical because they appeal to conditions that depend on whether this is so (or, in the case of the first example, to the very fact to be accounted for).

In search of the simple view

49

There is more to be said about what makes a criterion of identity over time substantive and non-trivial. Rather than exploring this further, I propose that a proper criterion is one that completes this formula: Necessarily, if x is a human person at t and y exists at t*, x = y iff and because . . .

where the word “because” expresses the logical or metaphysical dependence of the grounding claim. With any luck, this will avoid triviality and uninformativeness. Criterialism, then, is the claim that some such criterion is true. How the grounding claim and criterialism are related is a nice question. Although criterialism as stated entails the grounding claim, the converse entailment is less clear. You might think that if our identity consists in something else, it must be possible to articulate that something else in the form of a proper criterion of identity. But I do not know how to argue for this. Perhaps our identity over time could consist in something that is not expressible in terms of necessary and sufficient conditions. (Or perhaps there could be infinitely many such conditions, or even an uncountable infinity. I leave open whether this would count as a criterion of identity.)

anti-criterialism Let us turn now to attempts to state the simple and complex views. The most obvious proposal is that the complex view is criterialism and the simple view is its negation, anti-criterialism (that there is no proper criterion of personal identity). Anti-criterialism is probably the best candidate for being the simple view. Yet it is such a strong claim that few if any supposed friends of the simple view would accept it. Consider this thesis: Necessarily, if x is a human person at t and y exists at t*, x = y iff and because the thing that is x’s soul at t is the thing that is y’s soul at t*,

where a soul is an immaterial thinking substance, and a soul belongs to a particular being at a time just if that being has its mental properties then by virtue of the soul’s having them. Call this Cartesianism. Some philosophers accept it, namely those who think that we have souls as parts but are not ourselves souls – that a living human person is composed of a soul and a body (Swinburne 1984, p. 21). In that case, they say, a person existing now is identical to a being existing at another time if and only if and because that being has the person’s soul then. (If people are their souls, by contrast, personal identity and soul identity are the same, and one cannot depend on the other, contrary to Cartesianism.)

50

eric t. olson

Cartesianism is far from trivial: it tells us something about human people in particular, and is not true of concrete objects in general. It is also informative, in that one need not know whether x is y before knowing whether x has y’s soul. There may of course be practical obstacles to finding out whether we have the same soul before finding out whether we have the same person, owing to the fact that we cannot observe souls. But these obstacles appear to be only contingent. So Cartesianism is a proper criterion of personal identity. It is a version of criterialism, and thus, on the current proposal, a complex view. Yet it is considered a paradigm case of a simple view (Swinburne himself calls it simple: 1984, p. 19). Cartesianism also entails the grounding claim: it implies that personal identity consists in something else, namely sameness of soul. (That is unsurprising, given that criterialism entails the grounding claim.) This shows that both criterialism and the grounding claim are too weak to capture what distinguishes complex from simple views – assuming, at least, that the philosophical community is right about where the boundary lies. Accepting anti-criterialism would mean rejecting all proper criteria of personal identity. That would appear to rule out our being wholly material things, as in that case there could hardly fail to be a criterion in terms of atoms or matter, even if not a specific or satisfying one – something like the atomic criterion of p. 59 (Zimmerman 1998).4 It would also rule out our being partly material and partly immaterial, as that would presumably imply Cartesianism. It looks as if anti-criterialists must say that we are wholly immaterial souls. They would also have to reject all non-trivial criteria for a soul existing at one time to be identical to a soul existing at another time. For instance, there could be no criterion in terms of causal relations, no matter how vague. Even the causal-dependence criterion of p. 58 would be false. It would have to be metaphysically possible that you are not the soul who read the previous sentence – that it ceased to exist just now and you came into being in its place – even though all the causal relations are exactly as they actually are (see Merricks 2001a, p. 196). If we asked why that soul perished, or why another appeared, there would be no answer. It would not be because of the intervention of some deity, or a local disturbance in the ectoplasmic field, or anything else. There would be possible situations that differed only in the facts about our identity over time. Well, someone could say that. But I do not know whether any real philosopher ever has. It would be quite a surprise if this were the only possible 4

Merricks (1998b) denies this, though I believe he would accept the atomic criterion.

In search of the simple view

51

simple view – if some versions of substance dualism were simple and others, including those most commonly held, were complex, and the difference turned on such arcane matters as whether we are souls as opposed to soul– body compounds and whether any sort of causal dependence is necessary and sufficient for a soul to persist. If that were the simple view, it would hardly be worth including in the undergraduate curriculum. Whether there are human souls is an important fact about our identity over time, but whether there is any criterion of identity for them is at best a footnote.

analyzability Can we do better? The simple view is often said to assert that personal identity is “unanalyzable,” and the question of what it consists in is sometimes put by asking for an analysis of the concept of personal identity.5 This suggests that the issue at stake is whether any proper criterion of personal identity could count as an analysis. Analytic criterialism, then, would be the view that some such criterion is analytic – that is, true by virtue of the concepts that figure in it, and the negation of which can be transformed into an explicit contradiction by a process of analysis. Analytic anti-criterialism would be the view that no such criterion is analytic. Might these be the complex and simple views? Swinburne’s main objection to complex views is in fact an objection to analytic criterialism (Swinburne 1984, pp. 18–20). It is a version of Moore’s “open-question argument.” Consider any criterion of the form: necessarily, if x is a human person at t and y exists at t*, x = y iff and because x bears R at t to y as it is at t*, where R might be some sort of psychological or brute physical continuity. Now imagine that some future being really does bear R to you. Given this fact, Swinburne says, “mere logic” – or logic together with the meanings of the relevant expressions – cannot determine whether that being would be you. Neither those who say yes nor those who say no would thereby contradict themselves or commit any other logical or conceptual error. It is, as Moore would say, an open question. Thus, Swinburne concludes, the proposed criterion cannot be an analytic truth. And this holds for any proper criterion. Whatever the merits of this argument may be, it does not rule out such a criterion’s being true, but at most its being analytic. So if this is an argument for the simple view, the simple view must be analytic anti-criterialism. 5

Noonan, for instance, says that according to the simple view “personal identity is an ultimate unanalyzable fact, which resists definition in other terms” (Noonan 2003, p. 95).

52

eric t. olson

And maybe Cartesianism, if true, would not be analytic. That is certainly what the open-question argument suggests: given that each of us is a compound of a soul and a body, is it not an open question whether a being who has my soul at some other time would have to be me? (Locke famously said no.) Suppose my soul had its mental contents “erased” but was otherwise unharmed, and those contents were then replicated in another soul. No amount of conceptual analysis can determine whether I should then be the person with my original soul, the person with the new soul and my replicated mental contents, or neither.6 But whereas ordinary anti-criterialism was too strict to be the simple view, analytic anti-criterialism is too lenient: it is consistent with many views considered paradigms of complexity. Consider this one: Necessarily, if x is a human person at t and y exists at t*, x = y iff and because x’s biological life at t = y’s biological life at t*,

where a biological life is a self-sustaining chemical event that always coincides with a living organism (Van Inwagen 1990b, pp. 83–90). If this were true, our identity over time would consist in something other than itself, namely in the continuation of a biological life – a sort of brute physical continuity (unless we are ourselves lives – but no philosopher I know of believes that). This is a complex view if anything is. But it can hardly be analytic. Reflecting on the relevant concepts will never enable us to work out that we persist just as long as our biological lives continue. If nothing else, analytic truths are knowable a priori, and we cannot know a priori that human people even have biological lives, never mind whether they must always go where their lives go. So the life criterion is perfectly consistent with analytic anti-criterialism (and immune to Swinburne’s open-question argument). Many other proposed criteria that are considered complex are no more analytic: Unger’s “physically based approach,” for example (Unger 1990, ch. 5).

advocates of analytic criterialism Would any view of personal identity count as complex if the simple view were analytic anti-criterialism? Does anyone think that what it takes for us to persist is analytic? This would apparently mean that the concept of a human person had the persistence conditions of human people built into it: that 6

In fact Swinburne takes Cartesianism to be analytic (personal communication). If the simple view is analytic anti-criterialism, that would make Cartesianism a complex view.

In search of the simple view

53

you could work out a complete set of necessary and sufficient conditions for a human person to persist just by analyzing the concept of a human person. You might suppose that it is not the concept of a person, but the concept of personal identity over time that has the criterion of personal identity built into it. But what is the concept of personal identity, if not the concept of identity applied to people? And the concept of identity (or of identity over time) by itself provides no substantive criterion of identity. So if the criterion of personal identity is analytic, it must derive from the concept of a person. I know of two views that may have this consequence.7 One is Shoemaker’s theory that mental properties, by their very nature, give their bearers causal powers that fix their persistence conditions (Shoemaker 1999). It is a familiar thought that mental states are by nature disposed to combine in certain ways with other mental states to produce certain effects: this is the core of the functionalist theory of mind. Shoemaker adds to this the further claim that a mental state must be disposed to combine in this way only with states of the same subject, and to produce the relevant effects in that subject alone. Your being hungry, for instance, must tend to cause you to eat if you believe there is food before you, unless you have some competing goal – you and only you. Thus, any being that was caused, in this way, to eat by your state of hunger must be you. It follows that some sort of psychological continuity suffices for you or any other psychological being to persist.8 If a person is by definition a sort of psychological being, and if Shoemaker’s theory of mental properties is analytic, then this account of personal identity will be analytic, and presumably a version of analytic criterialism. A view that pretty clearly makes the conditions of personal identity analytic is the ontology of temporal parts or “four-dimensionalism.” It implies that for every candidate for being the career or history of a human person, and every part of such a candidate, there is a conscious, intelligent being whose history it is. So there is now a being sitting here that is just like me, both physically and mentally, except that it came into being at midnight last night and will cease to exist at midnight tonight. Another such being came into existence when my biological life began, some sixteen days after I was conceived, and will cease to exist when my life ends. Yet another came into being at conception, and will persist after my death as a corpse until my remains become dust.

7

8

Another possible candidate is Parfitian “reductionism,” but I do not understand it well enough to know for sure. Shoemaker (1985) and Noonan (2003, pp. 97–100) are good discussions. Merricks (1998b, pp. 111–16) argues against analytic criterialism. Though whether psychological continuity would be necessary for a person to persist is less clear; see Olson (2002).

54

eric t. olson

Many more such beings share my current stage. But according to most fourdimensionalists, only one of these beings (vagueness aside) is a person. Being rational and self-conscious is therefore insufficient for personhood. In fact the vast majority of beings psychologically like ourselves are not people, but merely share temporal parts with a person. To say what it is to be a person, on this view, we have to say what it takes for a person to persist. Thus Lewis (1976), for example, says that a person is a maximal aggregate of psychologically interconnected person-stages. A personstage is a more or less momentary being with the mental properties that characterize people: rationality and self-consciousness, perhaps. Two personstages are psychologically connected when one “inherits” its psychological properties from the other in a special direct way: when one has a memory of an experience of the other’s, for instance. An aggregate of psychologically interconnected person-stages is a being composed of person-stages, each of which is psychologically connected with every other, and it is maximal when it is not a part of any other such aggregate. This gives the conditions necessary and sufficient for a person to persist. And because it is part of the definition of “person,” it is analytic if anything is. But this can hardly be the frontier between simple and complex, with views like Lewis’s and Shoemaker’s on the complex side and all the rest, including those of Unger and Van Inwagen, counting as simple.

empiricist theories Another issue sometimes discussed in connection with the simple/complex debate is whether personal identity consists in any of the conditions we use as evidence for judgments about it. Swinburne complains that some philosophers fail to distinguish between evidential and constitutive criteria: “their account of the logically necessary and sufficient conditions of personal identity [is] in terms of the evidence of observation and experience which would establish or oppose claims of personal identity” (Swinburne 1984, p. 3; for similar remarks see Chisholm 1976, pp. 111–13). He calls these empiricist theories of personal identity. Might complex views be empiricist theories? Is this the issue at stake in the simple/complex debate – whether there are conditions individually necessary and jointly sufficient for personal identity that we use as evidence for claims about who is who? Well, what do we use as evidence for these claims? What is the basis for my belief that the man emerging from the office next door is my colleague Nils, rather than another colleague or a

In search of the simple view

55

complete stranger? Well, he looks like Nils did when I last saw him. And he acts like Nils: he greets me rather than staring blankly, and responds as Nils has in the past when engaged in discussion about departmental matters. And I take it to be rare for someone to look and act so much like someone else as to fool his colleagues. Our primary evidence for claims about personal identity seems to be physical and behavioral similarity. If an empiricist theory is one according to which there is a criterion of personal identity in terms of our primary evidence for such claims, then empiricists are those who hold that our identity over time consists in physical and behavioral similarity. But no one thinks that. Even if we could specify the respect of similarity generously enough to allow for the fact that people’s appearance and behavior change radically between infancy and old age, no one thinks that mere similarity of any sort is sufficient for identity: there could be someone other than Nils whose appearance and behavior were as similar to his as you like. This cannot be what Swinburne meant by an empiricist theory. What did he mean? Well, we take physical similarity to be evidence for some sort of physical continuity – causal dependence of later physical states on earlier ones – and we take behavioral similarity to be evidence for psychological continuity, which is also a sort of causal dependence. And many philosophers think that physical or psychological continuity is not merely evidence of personal identity, but what it consists in. So perhaps empiricist views are those according to which there is some condition that we use as evidence – not necessarily as primary evidence – for judgments about personal identity, which figures in a proper criterion of personal identity. But this would make Cartesianism an empiricist theory. If it were true (and we knew it), physical and behavioral similarity would be evidence for sameness of soul: I should be warranted in believing that the man next door has Nils’ soul because he looks and acts like Nils. At any rate this is consistent with Cartesianism. (It would be bad news for Cartesians if it were not true: it would seem to deprive us of any evidence for claims about who is who.) Yet having the same soul would not only be evidence of personal identity, but what it consists in. You might suppose that Cartesianism is not an empiricist theory because the condition that it takes personal identity to consist in is not observable: which human being has which soul is not “evidence of observation and experience.” But psychological continuity is not observable either. Both have to be inferred (in others, if not oneself ) from behavioral similarity. So it looks as if Cartesianism is an empiricist theory if psychologicalcontinuity views are. This is not the right boundary either.

56

eric t. olson bruteness

What could make Cartesianism, but not physical- or psychological-continuity accounts, a simple view? Perhaps this: even if personal identity consists in identity of soul, identity of soul consists in nothing. Nor is there any criterion of identity for souls. Though personal identity may depend metaphysically on something else, that something else does not itself depend on anything further. So this might be the simple view: personal identity either consists in nothing (that was anti-criterialism, or near enough), or consists in the identity of something else which consists in nothing. Alternatively, if there is any proper criterion of identity over time for human people, it appeals to the identity of something for which there is no such criterion. (For present purposes we can ignore the difference between these two variants.) Call this proposal bruteness. But some Cartesians deny bruteness. Swinburne, for instance, used to say that souls, like all substances, persist by virtue of sameness of form and continuity of stuff, so that their identity is not brute (Swinburne 1984, p. 27; 1997, pp. 153–4).9 To make Cartesianism compatible with bruteness, you would have to combine it with bruteness about souls (a view briefly explored on p. 50). It would be pretty surprising if this were the frontier between simple and complex views. There are also views everyone would classify as complex that are consistent with bruteness. Someone who endorsed the life criterion on p. 52, but believed that identity of lives was brute, would accept bruteness. Or imagine a Humean who believed that people are, as the great man said, “nothing but bundles of impressions,” and who endorsed the following criterion: if x is a human person at t and y exists at t*, x = y iff and because y is composed of impressions at t* and a certain proportion of those impressions are also parts of x at t. Combined with the view that there is no criterion of identity for individual impressions, this would entail bruteness.

noonan’s proposal Harold Noonan (2011) has recently proposed a subtle refinement of bruteness. The simple view, he says, is that the only necessary conditions on personal identity belong to a certain special sort. A condition will belong to this sort if it follows from what it is to be a person (or a human person) at a particular moment. For instance, you cannot be a person at a particular 9

He has since revealed in personal communication that he no longer holds this view. But he does not say that he once held a complex view and now holds a simple view.

In search of the simple view

57

moment if you are made entirely of stone then. Supposing, as Noonan does, that whatever is a person at some time must be a person whenever it exists (which we can grant for the sake of argument), it follows that no past or future being can be you if it is made entirely of stone at that time. But as everyone accepts this, it ought to be consistent with the simple view. Noonan calls such conditions “synchronic.” Conditions that are trivial in something like the sense discussed on p. 48 are also compatible with the simple view. And so are conditions that are “identity-involving”: those that require the identity over time of something other than the person in question. For instance, Cartesianism implies that a past or future being can be you only if it then has the same soul as you have now, where a soul is not a person. This is compatible with the simple view. The simple view, then, is that the only necessary conditions for personal identity are either synchronic, trivial or identity-involving. Complex views assert further conditions. Like the previous proposal, Noonan’s implies that at least some versions of Cartesianism are complex, depending on what conditions they impose on the identity of souls. If the only conditions for soul identity are synchronic (if they follow from what is necessary for a thing to be a soul at an instant), trivial or identity-involving, then Cartesianism may be a simple view. But if there are other necessary conditions for a soul to persist – for instance, if it requires a soul’s later states to depend causally in some way on its earlier ones – then these will also be conditions for personal identity, and the Cartesian view will count as complex. Likewise, it appears to count some versions of the life criterion and the Humean view as simple. If I have understood it, Noonan’s proposal makes the simple view so strong that no one actually holds it. No philosopher I know of would accept that a person could be wholly material at one time and wholly immaterial at another. If you are now made up entirely of matter, then no being that is not made up even partly of matter at some other time can be you. Likewise, if you are wholly immaterial now – if you are a Cartesian soul, say – then you could not come to be made up entirely of matter. This is not a synchronic condition for personal identity: it does not follow from what it is to be a person, or even a human person, at a particular moment. From the mere proposition that x is a human person at t, it does not follow either that x is material at t or that x is immaterial at t, never mind that x is material or immaterial at any other time. So it seems, anyway. Nor is the condition trivial or identity-involving. If so, Noonan’s proposal implies that you hold a simple view only if you reject this condition: only if you think it is possible for one of us to change from being wholly material to being wholly immaterial or vice versa. But every supposed advocate of the simple view accepts the condition.

58

eric t. olson specific and unspecific

I can offer no better account of what makes a view simple or complex. As far as I can see, no principle divides views about personal identity in the right place, or has the importance that the simple/complex distinction is traditionally ascribed. Let me finish by trying to say something positive. Suppose the traditional taxonomy of simple and complex views really is hopelessly wrongheaded. What ought to replace it? How should views of personal identity be categorized? What are we to tell our students? Well, we can still divide views into those that appeal only to brute physical conditions, those that appeal to some sort of psychological continuity, and those that appeal to neither. This is an important three-way distinction. Someone might suggest that simple views are precisely those appealing neither to brute-physical nor to psychological conditions – thus making the simple/complex distinction irreducibly disjunctive. That would make most versions of Cartesianism simple – the exception being those that give psychological persistence conditions for souls (Foster 1991, pp. 240–61). But this would be to abandon all our original starting points. It would do away with the thought that complex views take our identity over time to consist in something while simple views deny this. There are views according to which personal identity over time consists in something other than brutephysical or -psychological continuity – or, for that matter, identity of soul. Kantians, for instance, might say that it consists in something noumenal that we can form no conception of. The disjunctive proposal would count these as simple views. But should they not count as complex? It would also give up the conviction that simple views differ from complex ones over something more fundamental than what personal identity consists in, instead putting their difference on a par with that between physical and psychological views. Here is something else we can tell our students. Criteria of personal identity differ enormously in how much information they provide. This one tells us rather little: Necessarily, if x is a human person at t and y exists at t*, x = y iff and because the total state of x at t depends causally in an appropriate way on the total state of y at t* or vice versa.

It says only that personal identity consists in some sort of causal dependence. That is not trivial. If it is true, it is because of the metaphysical nature of human people. It may not hold for all objects, or even all concrete objects. It is not true of shadows and sunbeams, if there are such things:

In search of the simple view

59

their earlier states do not cause their later ones. It has controversial implications: it rules out our surviving without any sort of causal dependence on our earlier states, contrary to some views about life after death (e.g. Baker 2005; Hick 1990, pp. 123–4). But beyond that it is useless. It says nothing about what sort of causal dependence our identity consists in. Think of cases where it is hard to know who is who, even if we know everything else. Suppose we ask whether reincarnation is possible, or what would happen to someone whose brain was transplanted into a different head. The causaldependence criterion is consistent with any answer to these questions. Or consider this “atomic criterion” (adapted from Zimmerman 1998): Necessarily, if x is a human person at t and y exists at t*, x = y iff and because the arrangement of x’s atoms at t and the arrangement of y’s atoms at t* both belong to an atomic history of the sort that necessarily any human person has, and which is necessarily the history of a human person,

where “the arrangement of x’s atoms” includes both the intrinsic nature of the atoms composing x and their spatio-temporal and causal relations to one another and to other atoms, and someone’s “atomic history” is a complete description of her atoms throughout her existence. This is not trivial either. If it is true, it is because of how it is with human people and not merely because of the nature of atomic histories and arrangements. It has controversial implications, ruling out our surviving without being composed of any atoms at all. But it tells us no more about what happens in most of the puzzle cases than the causal-dependence criterion does.10 These accounts are enormously unspecific about what it takes for us to persist. The life account, that a past or future being is you just if it then has your current biological life, is far more specific. It tells us who would be who in most of the puzzle cases. The same goes for familiar psychological-continuity views. Yet even these views are less specific than they might be. The precise sort of causal dependence that psychological continuity consists in is elusive. No matter how much detail an account provides, there will be possible cases where a being existing at another time has mental states that depend causally on yours now (or vice versa) in such a devious way as to leave it unclear whether he or she is psychologically continuous with you. And even if we could specify the sort of causal dependence, psychological continuity will be a matter of degree, and it is hard, if not impossible, to say precisely what degree 10

It may be that these are somehow not “proper” criteria of identity, and thus ought to be consistent with anti-criterialism, supporting the claim that anti-criterialism is the simple view. Those with this suspicion will want to think about what makes them improper.

60

eric t. olson

is needed. I have never seen a psychological-continuity account specific enough to tell us, in conjunction with the relevant underlying facts, in every case whether the resulting being would be you, or not you, or whether it would be indeterminate whether it was you. You might think this is only because no one has taken the trouble to write out a completely specific psychological-continuity view, or because no one knows enough to do it. More generally, it may be that if some less specific account of personal identity is true, it must be because it follows from some more specific account – perhaps even a maximally specific account that would entail, in conjunction with the underlying conditions, all the facts about personal identity over time. I have no idea whether this is the case. There is certainly reason for doubt. Shoemaker, who has thought about these matters as deeply as anyone, confesses that it seems impossible to specify the sort of “appropriateness” that figures in the causal-dependence criterion without already knowing the precise conditions under which human people persist (Shoemaker 1979, p. 337). You might think it is impossible only for finite minds: God must be able to do it. But who knows? It is worth noting that Merricks, who presumably accepts the atomic criterion, explicitly denies that there is any more specific criterion of personal identity (Merricks 1998b). So here is an important difference among criteria of personal identity: some give us quite specific information about what our identity over time consists in, while others say little. This bears some affinity to the traditional simple/complex distinction: those views called complex often appear more specific than those called simple. But the simple/complex distinction is not that between unspecific and specific views. For one thing, the difference between more and less specific is a matter of degree, while the simple/complex distinction is supposed to be absolute: we tell our students that some views of personal identity are simple and others are not, not that some are more simple than others. And the two distinctions divide the territory in different ways. The causaldependence criterion is normally taken to be a complex view: Shoemaker explicitly contrasts it with that of Butler, Reid and Chisholm that “the crosstemporal identity of genuine continuants [cannot] be said to consist in the holding of any relations other than the relation of identity itself” (1979, p. 323). Yet it is consistent with the absence of any more specific criterion. Nor is it clear whether the soul criterion (p. 49) is any less specific, even though it is supposed to be a simple view. Think of reincarnation again. Could someone born after my death be me? The causal-dependence criterion gives an ambivalent answer: only if that being’s state then depended causally in the appropriate way on mine before my death. But it says nothing about

In search of the simple view

61

what way is appropriate, or whether it could be appropriate in this case. The atomic criterion tells us even less. The soul criterion gives what sounds like a more specific answer: such a being would be me if and only if it had my immaterial soul. If you were a deity wanting to reincarnate human people, this (if true) would be far more useful information. Thus, some views traditionally taken to be complex appear, if anything, to be less specific than views considered simple. And the distinction between less and more specific may be not only a matter of degree, but context-relative. The soul criterion might be less specific than the causal-dependence criterion relative to some contexts and more specific relative to others – as in the case of our imaginary deity. But I do not think anyone ever took the simple/complex distinction to be context-relative. explanatory demands The distinction between specific and unspecific criteria of identity is closely related to a difference in expectations. Philosophers discussing personal identity often make explanatory demands. Suppose once again that we are considering the possibility of reincarnation. Here is an almost irresistible question: what could make it the case that some infant born after I am dead was me? Out of all those future newborns – many billions of them – what could make this one me, rather than some other one? To put it the other way round, what could make it the case that a certain infant born after my death was me rather than someone else – you, say, or Socrates, or someone who had never existed before? Or if reincarnation is impossible, we shall want to know what makes it impossible. What is it about the way someone born after my death would have to relate to me that would necessarily rule out its being me? Philosophers disagree about when such explanatory demands are legitimate: about what must have an explanation in matters of personal identity, and where explanation comes to an end and we face brute facts. They disagree about which questions must have answers. They also disagree about what sort of answers we can expect. At one extreme, someone might say that an infant born after my death could be me but nothing would have to make it me. It would not have to be me because it had my soul, or because it bore any psychological relation to me, or because its states then depended causally in a certain way on mine now, or because God willed that it be so. It could simply be me, and that would be that.11 At the other extreme, someone might say that such an infant could not possibly be me because its total state then 11

This may be the view of Mavrodes (1977).

62

eric t. olson

would not bear to my total state now the sort of causal dependence that is both necessary and sufficient for personal identity; and she might be able to describe this relation precisely. (Certain versions of the life criterion might be like this.) Nearly all the answers actually proposed – both those considered simple and those considered complex – fall between these extremes. When it comes to personal identity, some philosophers expect more facts to be explained than others do, and there is probably some correlation between this and whether one holds a view classified as complex or one classified as simple. But the correlation is far from perfect. Nor can I see any way of transforming this difference of expectations into a difference of doctrine. The simple view remains elusive.12 12

For discussions on the topic of this chapter I thank Andrew Jones, Harold Noonan, Richard Swinburne and the audience at the 2010 Obergurgl conference.

chapter 3

Personal identity, indeterminacy and obligation Ryan Wasserman

It is the foundation of all rights and obligations, and of all accountableness; and the notion of it is fixed and precise.

(Thomas Reid 1975 [1785])

personal identity and indeterminacy In Reasons and Persons (1984), Derek Parfit defends the complex view1 of personal identity, which he formulates as follows: The complex view: Facts about personal identity consist in other, impersonal facts.2

More specifically, Parfit defends the view that facts about personal identity reduce to facts about psychological continuity or connectedness (Parfit 1984, p. 216). He then goes on to argue that this view implies the possibility of indeterminacy: The Indeterminacy of Personal Identity: It is possible for questions about personal identity to lack determinate answers.

Parfit explains his reasoning as follows: We can describe cases where, between me now and some future person, the physical and psychological connections hold only to reduced degrees. If I imagine myself in such a case, I can always ask, “Am I about to die? Will the resulting person be me?” On the Reductionist View, in some cases there would be no answer. My question would be empty. The claim that I was about to die would be neither true nor false. (Parfit 1984, p. 214)

Many people find this conclusion incredible, and not in a good way. Chisholm, for example, writes that: 1 2

In Reasons and Persons, Parfit’s focus is on what he calls reductionism, but it is clear that he takes reductionism to be equivalent to the complex view. See, for example, Parfit (1982, pp. 227–8). Where “impersonal facts” are those which do not presuppose the existence of a continuing person.

63

64

ryan wasserman

When we use “the same person” in [the] strict way . . . although cases may well arise in which we have no way of deciding whether the person x is the same person as the person y, nevertheless the question “Is x the same person as y?” will have an answer and that answer will be either “yes” or “no.” If we know that x is a person and if we also know that y is a person, then it is not possible to imagine circumstances under which the question “Is x the same person as y?” is a borderline question – a question admitting only of a “yes and no” answer. (Chisholm 1970a, p. 171)

This way of thinking suggests the following argument against the complex view: if the complex view is correct, then personal identity can be indeterminate. But personal identity cannot be indeterminate. Hence, the complex view is wrong. This simple argument is a common one, having been suggested by Chisholm (1970a), Swinburne (1973−4), and many others.3 But the central premise of the argument is seldom defended: why think that personal identity cannot be indeterminate? 4 This omission is surely due to the inherent plausibility of the premise. When I imagine myself in some of the cases described by Parfit and I consider the question “Will I survive?” I do not know how to answer – but something deep inside of me says that there must be an answer. There must be an answer, which is just to say the Indeterminacy of Personal Identity must be false. The strength of this conviction is undeniable, but its source is unclear. The goal of this chapter is to identify the source of our anti-indeterminacy intuition and thus provide a better understanding of the standard argument against the complex view.5 indeterminacy and obligation Imagine a machine that can make ever so subtle changes to your body – including your brain – and is therefore capable of making equally small changes to your psychology. There are approximately fifty trillion cells in the human body, so let us imagine that the machine has a keypad that allows us to enter any number from zero to fifty trillion. If we enter “0” and you enter the machine, nothing will happen – you will emerge from the machine in almost exactly the same state that you entered. If we enter the number “1,” 3 4

5

Harold Noonan suggests that this line of thinking is “historically one of the central components of the Simple View” (1989, p. 226). Noonan (1989) bemoans the fact that this “question is seldom asked by the proponents of the Simple View” (see also Thomson (2008, pp. 173–4)). For some exceptions to the rule, see Williams (1970) and Hossack (2006). More carefully: the goal of this chapter is to identify at least one source of this conviction. Our opposition to indeterminacy could be overdetermined.

Personal identity, indeterminacy and obligation

65

however, the machine will destroy a randomly selected cell from your body and replace it with a new cell – or perhaps several cells – which qualitatively duplicates the corresponding part in the body of Albert Einstein (circa 1950). If we enter the number “2,” the machine will replace two cells in the same way. And so on. If we enter “50,000,000,000,000,” every cell of your body will be destroyed and replaced so that the person who emerges from the machine will be a qualitative duplicate of Einstein himself. Now, let us suppose that Frank is a convicted killer who has been sentenced to die. Suppose further that we place Frank into the matter-replacement machine and enter the number “20,000,000,000,000,” so that twenty trillion Frank cells are replaced with approximately twenty trillion Einstein cells. The man who emerges from the machine will be, so to speak, 60 percent Frank and 40 percent Einstein. Let us call this creature “Frank−Einstein” or, more simply, Frankenstein. Outwardly, Frankenstein looks a bit like Frank and a bit like Einstein. Inwardly, the transformation is similar. Frankenstein has many of the same beliefs, desires and psychological dispositions that Frank had. But he also has some of the psychological features of Einstein. And some of his psychological features are a “blend” of the two. Frank – before entering the machine – gave zero credence to the proposition that he had won a Nobel Prize; Frankenstein is not so sure. Frank had never been that fond of riding bicycles; Frankenstein is rather keen on the idea. Perhaps, most importantly, Frank clearly remembered killing a man in cold blood; Frankenstein seems to remember this . . . but it is all a bit hazy. This kind of case raises two questions – one metaphysical and one moral. The metaphysical question is this: is Frankenstein the same person as Frank? If the complex view is correct, then Frank has survived the procedure just in case there is sufficient bodily continuity or sufficient psychological continuity or sufficient continuity in both respects. This disjunction indicates that there are various versions of the complex view. One could focus on physical facts or psychological facts or both. One could focus on continuity or connectedness or both. And with respect to psychology, one could focus on memories or beliefs or desires or emotional states. The point is that there are many, many versions of the complex view. But every reasonable version will invoke continuity or connectedness in some way or other. Continuity and connectedness are a matter of similarity, and similarity comes in degrees. So, continuity and connectedness come in degrees as well. How continuous does a series need to be to be continuous? That is like the question: how tall do you have to be to be tall? Neither question has a determinate answer. Since there will be unsettled questions of continuity and connectedness, there will

66

ryan wasserman

be unsettled cases of personal identity. I will assume that the Frankenstein case is a case like this, so I will assume that – on the complex view – it is indeterminate whether or not Frank is Frankenstein. Now for the moral question: what is to be done with Frankenstein? Let us assume that Frank deserved to die for his crimes. Or, to put it in deontic terms, executing Frank would have been the right thing to do; letting him go would have been wrong. What about Frankenstein? If Frank is Frankenstein, then we are obligated to kill Frankenstein, for Frankenstein is a killer sentenced to die. If not, not. In other words, we are obligated to kill Frankenstein if and only if Frankenstein is Frank. And we are obligated to not kill Frankenstein just in case Frankenstein is not Frank. If it is indeterminate whether or not Frankenstein is Frank, it is indeterminate whether or not we ought to carry out the punishment. It is also indeterminate whether or not we ought not to carry out the punishment. More formally: O(Punish) ↔ Guilty ∇Guilty ∴ ∇O(Punish)

O(¬Punish) ↔ ¬Guilty ∇¬Guilty ∴ ∇O(¬Punish)6

The upshot is that the question “What is to be done with Frankenstein?” is one more “empty question.” But that is crazy! Or at least it seems crazy. How can a question of life and death be an empty question? To illustrate the oddity, imagine being juror in this case. You have to decide what is to be done with Frankenstein. And while you and the other jurors are debating, God himself comes down to join in the conversation. You now have the ultimate advisor before you – perfect in knowledge and wisdom. So you ask: “Should we sentence Frankenstein to death?” In response, God shrugs. “Should we let our prisoner go?” God shrugs again. “But God,” you persist, “What are we supposed to do?” To which God replies, “Don’t look at me!” I find that picture unsettling. It is one thing to say that we humans do not know what to do in a case like this. That is unsurprising. But if there is no fact of the matter, then there is nothing to know – not even for God. One can know all the facts without knowing what to do. And that does seem shocking. The same seems true when we turn from morality to rationality. To illustrate the point, let us quickly consider a second story involving the matter-replacement machine. 6

Where O is the obligatory operator and ∇ is the indeterminacy operator.

Personal identity, indeterminacy and obligation

67

Suppose you are about to be forced into the Einstein machine. The machine is set to “20,000,000,000,000.” You know what is about to happen. But before the machine is turned on, you are approached by a car insurance salesman. The salesman has it on good authority that the person who emerges from the machine – You* – will be in a car accident tomorrow. There will be $10,000 of damage done – damage that You* will be responsible for. However, the salesman offers to sell you a limited-term insurance policy. The policy will cost you $1,000 and it will cover all of the damages that will be incurred by You*. This is clearly a bad deal for the salesman. But is it a good deal for you? That is, are you rationally obligated to buy the insurance? This is not so clear. For the sake of simplicity, let us suppose that you care only about money – your money. If you are going to be the person in the accident tomorrow, then your expected utility of buying the insurance is $9,000. If you are not going to be in the accident, your expected utility is −$1,000. So, rationality dictates that you ought to buy the insurance if and only if you are going to be in the accident.7 And you ought not to buy the insurance just in case you are not going to be in the accident. If it is indeterminate whether or not you are You*, it is indeterminate whether or not you are going to be in the accident. So, it is indeterminate whether you are rationally obligated to buy the insurance or to reject the deal. Once again, this is a strange result. Once again, the image of God shrugging helps to emphasize the point. You have to decide whether or not to buy the insurance. God is there to serve as your financial advisor. He knows everything about the machine. He knows everything about the upcoming accident. He knows everything about your goals and utilities and your financial situation. All you want is a straight answer: “Should I buy the insurance?” And the best that God can do is . . . shrug. That seems wrong. That seems deeply wrong. Questions about what is to be done – whether rational or moral – must have answers. In other words, the following principle must be false: The Indeterminacy of Obligation: It is possible for questions about obligations to lack determinate answers.

If this principle follows from the Indeterminacy of Personal Identity, then that principle must be rejected as well. 7

Objection: you are either going to die when you step into the matter-replacement machine or not (though it may be indeterminate which). If you are going to live – and therefore be the person in the car accident – you should buy the insurance. Moreover, there is no benefit to saving money if you are about to die. Thus, there is no reason not to buy the insurance. Response: let us suppose that you will have the opportunity to spend the $1,000 on pleasure-producing items if you do not spend it on insurance.

68

ryan wasserman indeterminacy and moral dilemmas

I have suggested that our intuitive resistance to the Indeterminacy of Personal Identity flows from a more basic opposition to the Indeterminacy of Obligation. This is not a new idea. In his Essays on the Intellectual Powers of Man, Thomas Reid writes that: The identity . . . which we ascribe to bodies, whether natural or artificial, is not perfect identity; it is rather something which, for the convenience of speech, we call identity. It admits of great change of the subject, providing the change be gradual . . . But identity, when applied to persons, has no ambiguity, and admits not of degrees, or of more and less. It is the foundation of all rights and obligations, and of all accountableness; and the notion of it is fixed and precise. (2002 [1785], p. 113)

As I read Reid, he is suggesting an argument against the Indeterminacy of Personal Identity. This argument has an unstated premise – namely, that facts about rights, obligations and responsibility are, in his words, “fixed and precise.” So the argument, made explicit, is this: morality must be determinate, and morality is founded on personal identity, so personal identity must be determinate as well. This old argument suggests a new question: why think that morality cannot be indeterminate? Several answers to this question will be suggested.8 But first, a point of clarification: when Reid writes of morality, he mentions rights, obligations and accountableness (or responsibility). I have chosen to focus on obligations, and one might object to this decision. Suppose, for example, that one is attracted to satisficing consequentialism – the thesis that an action is morally permissible if and only if it is “good enough.”9 If the satisficing view is correct, then one is obligated not to perform an action if and only if it is not “good enough.” What is it for an act to be “good enough”? On the picture presented by the satisficer, all of one’s available actions can be ranked in terms of which are better than others (where there may be many ties in the ranking). The satisficer then says that an action is “good enough” just in case it appears “high enough” in the rankings. Presumably, the line between those options which are “high enough” and those which are not is unsettled, so it can be indeterminate whether or not one is obligated to not perform a certain action. This result is not obviously absurd. Compare: every human being can be ordered in terms of who is taller than whom (where there may be many ties in the rankings). Someone counts as “tall” (relative to a context) just in case he 8 9

See below. See also section “Objections to epistemicism” (p. 74) and Wasserman (2004b). On satisficing, see Slote (1985, pp. 45–7).

Personal identity, indeterminacy and obligation

69

or she appears “high enough” in the ranking. This threshold is determined by context but, presumably, most contexts fail to determine a precise point in the rankings.10 What would seem absurd, for the satisficer, would be indeterminacy in the rankings. That is, it would be absurd to think that there could be two available acts, A and B, where there is no fact of the matter as to whether A is better than, worse than, or equal to B. (Once again, the image of God shrugging helps to reinforce this feeling.) Given this, the arguments of the previous section can be easily adapted for the satisficer. If personal identity can be indeterminate, then it can be indeterminate whether one option is better than (or worse than or equal to) another. For example, it will be indeterminate whether it would be better to punish Frankenstein or to let him go. But that is impossible. So the Indeterminacy of Personal Identity is unacceptable. The lesson is that the Indeterminacy of Personal Identity will imply indeterminacy in many moral matters, and one might focus on different matters, depending on one’s moral theory. That being said, I will now set satisficing aside and continue to focus on the Indeterminacy of Obligation. Let us return to our question: why think that morality – and moral obligations in particular – cannot be indeterminate? One reason has to do with the possibility (or impossibility) of moral dilemmas. A moral dilemma is commonly defined as a situation in which one is obligated to perform each of two mutually inconsistent actions (not prima facie obligated, but obligated obligated). In a genuine moral dilemma, one is guaranteed to fail in one’s moral duty. A common example is the story of Sophie’s Choice, in which a Polish woman is arrested by the Nazis and sent to Auschwitz. On arrival, she is “honored” for not being Jewish by being given a choice: one of her children will be spared the gas chamber, but she has to choose which one. In the agony of indecision, as both children are being taken away, she chooses her son (who, she reasons, is older and stronger, and therefore more likely to survive). Tragically, she loses track of her son and never learns his fate. The suggestion is that, in this case, Sophie is obligated to save her son and obligated to save her daughter, despite the fact that she cannot do both. Of course, many people reject this description of the case, since they deny the possibility of genuine moral dilemmas. There are at least three reasons for thinking that such cases are impossible. First, moral dilemmas would violate the “ought-implies-can” principle, at least if we accept the standard distribution principle: 10

For more on gradable predicates and comparatives, see Bolinger (1972).

70

ryan wasserman

Distribution: (OA & OB) ↔ O(A & B)

Second, the existence of moral dilemmas is inconsistent with two other standard principles of deontic logic: Principle of Deontic Consistency: OA → ¬ O¬ A Principle of Deontic Entailment: (A → B) → (OA → OB)

The third reason for rejecting the possibility of moral dilemmas begins with the thought that morality must be action-guiding – and not just actionguiding, but uniquely action-guiding. A set of rules that tells you to do both A and B (where these are incompatible options) tells you too much and thus does not tell you enough: it does not tell you whether you should do A rather than B or B rather than A. And that is exactly what one wants to know in a case like Sophie’s Choice. This third reason for rejecting the possibility of moral dilemmas is also a reason for rejecting the possibility of moral indeterminacy – at least the kind of indeterminacy that concerns Parfit. According to Parfit, there is no fact of the matter about what ought to be done in the Frankenstein case. It is not obligatory to kill Frankenstein, nor is it obligatory to let him go. Neither action is permissible; neither is impermissible. In short, we have a failure of moral guidance. In fact, one might go on to argue that the Frankenstein case is itself a moral dilemma – at least for those who accept indeterminacy. Let us suppose that, generally speaking, the punishment ought to fit the crime. In particular, let us suppose that (all else being equal) it would be wrong to determinately punish someone for a crime he has not determinately committed. Also, all else being equal, it would be wrong to determinately absolve someone who is not determinately innocent. More formally: Principle 1: (¬ Δ Guilty) → (O¬Δ Punish) Principle 2: (¬Δ¬ Guilty) → (O¬Δ¬ Punish)11

Frankenstein is not determinately guilty and not determinately not guilty. Given the two principles, we are obligated to not determinately punish him and to not determinately not punish him. In other words, we are obligated to make it the case that it is indeterminate whether or not Frankenstein is punished. But that might not be within our power. Perhaps we could throw Frankenstein back into the matter-replacement machine. In that case, justice would be served since it would be indeterminate whether or not we carried out the death sentence. But if we do not have something like 11

Where Δ is the determinacy operator.

Personal identity, indeterminacy and obligation

71

this available to us, we are stuck. We must either determinately punish or determinately not punish. But we are obligated to refrain from both. Hence, we have a moral dilemma. indeterminacy and epistemicism Having identified the source of our anti-indeterminacy conviction, we can now return to the original argument against the complex view. Given the preceding, we can formulate the argument as follows: The Indeterminacy Argument (1) If the complex view is true, then the Indeterminacy of Personal Identity is true. (2) If the Indeterminacy of Personal Identity is true, then the Indeterminacy of Obligation is true. (3) The Indeterminacy of Obligation is false. So, (4) The complex view is false. We will consider two objections to this argument, the first of which is the epistemicist response. Epistemicism is a view on vagueness often associated with Timothy Williamson (1994).12 In a slogan, the view says that vagueness is ignorance. Williamson believes, for example, that there is a precise line between the bald and the non-bald – we just do not know where that line is. A borderline case of baldness, then, is either a case of unknowable baldness or a case of unknowable non-baldness. Slightly more carefully, Williamson identifies vagueness with a specific kind of ignorance: ignorance due to semantic plasticity.13 I will illustrate the idea of plasticity and then explain its connection to ignorance. Let us focus on the predicate “bald.” The extension of this predicate is fixed by its use in accordance with the semantic laws that link use and meaning. Of course, there are many candidate properties to serve as the semantic value of “bald” – there is the property of having fewer than 1,000 hairs on one’s scalp, the property of having fewer than 1,001 hairs on one’s scalp, and so on.14 Let us suppose, for the sake of argument, that the overall pattern of use determines that “bald” expresses the first property on our list, so that someone is bald just in case he has fewer 12 13 14

For an alternative version of epistemicism, see Sorenson (2001). I owe this phrase to Hawthorne (2006). These are all “candidates” in the sense that none of them is obviously not the semantic value of “bald” – they are all in the running, so far as we can tell.

72

ryan wasserman

than 1,000 hairs on his head. That is how things are, but things could have been slightly different – a few more people could have been a little less inclined to apply the term “bald” to individuals with 1,000 hairs. If usage had differed enough, the semantic laws would have drawn slightly different lines. So, it very easily could have been the case that “bald” picked out the property of having fewer than 999 hairs on one’s scalp. It is in this sense that the predicate “bald” is semantically plastic – it could have very easily been “bent” toward a different semantic value. Now, suppose I believe a particular man to be bald, just on the basis of observation. And suppose that the man in question has exactly 999 hairs on his head. In that case, my belief will be true. But, Williamson says, my belief will not count as knowledge. His explanation begins with the thesis that knowledge requires safety. If someone forms a true belief, but could just as easily have formed a false belief, then he is not being safe – epistemically speaking – and thus does not know. For example: suppose that I look at a big tree and guess that it has 1,872 leaves on it. That might turn out to be true, but we would not say that I knew the precise number of leaves on the tree. That is because my vision is not sufficiently sensitive to small variations. So, I would have made the same estimate, even if there had been one leaf more or one leaf less. So, I could have easily been mistaken. So, I do not believe safely. So, I do not know. Williamson says that it is the same thing when it comes to my belief that Mr. 999 hairs is bald. In this case, my belief-forming mechanism is not sufficiently sensitive to small changes in use. I would still have believed that Mr. 999 is bald – I would still have assented to the sentence “Mr. 999 is bald” – even if the dispositions of other language-speakers had been slightly different. However, if use had been different, I would have manifested a false belief by assenting to the sentence “Mr. 999 is bald.” Hence, I could easily have been mistaken. Hence, I do not know.15 This is how semantic plasticity leads to ignorance in borderline cases. And that is the kind of ignorance that is characteristic of vagueness. Now that I have explained epistemicism, I can explain the epistemicist response. When it comes to the Indeterminacy Argument, the crucial question for the epistemicist is: how are we to understand talk of “indeterminacy”? Some philosophers (in at least some contexts) treat “indeterminate” as synonymous with “indefinite” and thus understand talk of “indeterminacy” in terms of vagueness (see, for example, Evans 1978 and Wright 1987, section 5). Others 15

For a more complete presentation of this argument, see Williamson (1994, ch. 8).

Personal identity, indeterminacy and obligation

73

treat “indeterminate” as synonymous with “metaphysically unsettled,” so to say that something is “indeterminate” is to say that there is simply no fact of the matter (or, perhaps, no determinate fact of the matter) (see, for example, Field 2003 and Barnett 2009). Since there are two readings of “indeterminacy,” there are two ways of understanding the Indeterminacy Argument. On each interpretation, the epistemicist will deny a different premise. Let us take the second reading first, since this is what Parfit means when he says that personal identity admits of indeterminacy (Parfit 1984, pp. 213–15). And let us focus on the first premise of the Indeterminacy Argument. On the current reading, the Indeterminacy of Personal Identity amounts to the following claim: it is possible for personal identity to be metaphysically unsettled. Thus it is possible for there to be “yes or no” questions about personal identity where we cannot answer “yes” or “no,” since there is no fact of the matter either way. The epistemicist will deny that this conclusion follows from the complex view. In other words, she will deny the first premise of the Indeterminacy Argument. It is not difficult to see why. Parfit’s combined spectrum presents us with a traditional sorites argument. If the matter-replacement machine replaces one of Frank’s cells, he clearly survives. If we replace all of Frank’s cells, he clearly does not survive. In between, there are many borderline cases – like the one where we replace 40 percent of Frank’s cells. In that case, it is unclear whether or not Frank survives. But, for the epistemicist, this lack of clarity is simply the result of our epistemic limitations. There is still a fact of the matter as to whether or not Frank survives, so our question in this case will not be empty. That is just to say that there is no indeterminacy, in Parfit’s sense of the word. What about the other sense? Let us suppose that “indeterminate” simply means “vague.” And let us turn our attention to the third premise of the Indeterminacy Argument. On this reading, the Indeterminacy of Obligation amounts to the following claim: it is possible for questions about obligations to lack non-vague answers. For the epistemicist, this is true just in case it is possible for there to be moral questions that we cannot know the answer to (where this ignorance is due to semantic plasticity). I think the epistemicist will be happy to endorse that claim and, thus, to deny premise (3). The crucial point is that, when we are thinking like epistemicists, the Indeterminacy of Obligations does not seem that bad. That is because indeterminacy, on this reading, is consistent with there being some fact of the matter. Go back to the car insurance case and the thought experiment involving God as your financial advisor. You are about to get into the

74

ryan wasserman

matter-replacement machine and you must decide whether or not to buy the insurance for tomorrow’s crash. Since this is a borderline case, you do not know whether you are going to be the person in the crash. But God does! God knows where the line is between sufficient and insufficient continuity, just as he knows where the line is between the bald and nonbald individuals. The good news is that God is not going to shrug. He is going to give you an answer. Thus, there is no embarrassment in embracing the Indeterminacy of Obligation.

objections to epistemicism I have suggested that the epistemicist has a simple reply to the Indeterminacy Argument: If “indeterminate” means “metaphysically unsettled,” then premise (1) is false. If “indeterminate” means “vague,” then premise (3) is false. Either way, the argument is unsound. There are at least three reasons why one might reject this response. First, one might think that epistemicism is simply crazy.16 It is insane to think that one hair could make a difference between being bald and being not, or that one penny could make a difference between being rich and being not, or that one cell could make a difference between being the same person and being someone else. I agree that epistemicism is crazy, but that does not mean that it is false. Here, it is important to remind ourselves that the sorites paradox is a genuine paradox – the argument is intuitively valid, the premises are intuitively true, and the conclusion is intuitively false. Every solution will thus be counterintuitive (at least initially), in which case the oddity of epistemicism need not count against it.17 Second, one might have a special reason to dislike epistemicism when it comes to the case of personal identity. Recall that (Williamson’s version of) epistemicism requires many candidates. In the case of “bald,” there was the property of having fewer than 999 hairs, the property of having fewer than 1,000 hairs, and so on. The predicate “bald” expresses one of these properties, but it would have expressed a different candidate if use had been a bit different. This raises the question: what are the many candidates in the case of personal identity? 16 17

Field, for example, describes Williamson’s view as “beyond belief” (2003, p. 458). Of course one might, on reflection, judge epistemicism to be more counterintuitive than its competitors, and that might be a mark against the view. My point is that the view should not be rejected just on the grounds that it is counterintuitive.

75

Personal identity, indeterminacy and obligation A+B A

B

PS1

PS2

t1

t2

PS3

PS4

t3

t4

Space

Time

Fig. 2: Different sums of person-stages.

Let us focus on the proper name “Frank.” It is clear that this name refers to the person before the operation, but it is vague whether or not it refers to Frankenstein – the person who exists after the operation. If it is vague, then, for Williamson, there must be multiple candidates. The name “Frank” refers to one of these objects, but it would have referred to a different object if use had been a bit different. What are the objects in question? The most natural suggestion is: different sums of instantaneous temporal parts. To illustrate the idea, let t1 be the first moment of Frank’s existence and let t2 be the last moment before the matter-replacement machine is activated. Let t3 be the moment at which the machine is activated18 and let t4 be the final moment of Frankenstein’s existence (see Figure 2). According to the doctrine of temporal parts, persistence through time is like extension across space. Just as an extended line is made up of smaller line segments (and, ultimately, point-sized parts), a persisting object is made up of shorter-lived objects (and, ultimately, instantaneous temporal parts).19 Thus, the temporal parts theorist will say that, in the case described, there is a distinct person-stage – PS1, PS2, etc. – for each time in the story. There is 18 19

To simplify matters, let us suppose that the machine works instantaneously, so that the moment of activation is also the moment at which the machine completes its matter-replacement work. For a more careful characterization of the temporal parts view, see Sider (2001, ch. 3) and Wasserman (2004a).

76

ryan wasserman

also the sum of all the pre-operation person-stages – A – and the sum of all the post-operation stages – B. Finally, there is the sum of all the personstages from t1 through t4 – A⊕B. Now, suppose for the sake of argument that there is sufficient continuity between PS2 and PS3, so that the person after the operation is the same person that we had before. In that case, the name “Frank” will refer to the sum A⊕B. However, if use had been slightly different, then the predicate “is sufficiently continuous with” would not have applied to PS2 and PS3, in which case the predicate “is the same person as” would not have applied to the pre-op person and the post-op person. In that scenario, the name “Frank” would refer to A. This would be yet another example of semantic plasticity.20 This gives rise to ignorance and that, according to the epistemicist, is what gives us vagueness. Of course, to offer this explanation, the epistemicist must first accept the existence of temporal parts, as well as multiple cross-time fusions of those parts. Many philosophers do not accept the existence of such objects, so they must reject the epistemicist response. Here is a third and final worry for the epistemicist. Suppose, once again, that there is sufficient continuity between PS2 and PS3 (Figure 2). We are assuming this is a borderline case, which means that there is just barely enough continuity. But still, as things stand, Frankenstein – the person after the operation – is the same person as Frank – the man from before the operation. So, Frankenstein can be held morally responsible for Frank’s crimes. So it is permissible for us – and perhaps even obligatory for us – to carry out the relevant punishment. But here is the problem: if we had used certain words – words like “person” and “similar” and “same” – just a little bit differently, then that same sentence would have expressed a falsehood. The sentence “We ought to carry out the punishment” is true, but could easily have been false. The worry is not that that sentence could have easily been false because the non-linguistic facts could have easily been different. That is true, but unobjectionable. Rather, the worry is that the sentence in question could have easily been made false by our dispositions to use certain words in certain ways. To clarify: I am not claiming that we have the power to determine moral matters with our words. The objection is rather this: if one is an epistemicist on moral matters, for example, one requires many candidate meanings for the relevant moral terms. Moreover, the moral terms must be semantically plastic with respect to those candidates – for example, one must say that 20

For more on the relation between vagueness, personal identity and temporal parts, see Sider (2001, pp. 120–39) and Sider (2000).

Personal identity, indeterminacy and obligation

77

“is morally obligatory” has one of the candidate semantic values but could very easily have had one of the others. And that seems wrong. It seems as if our moral terms should be rigid in a way that many non-moral predicates are not. I suggest that we understand rigidity, in this context, in terms of reference magnetism. Start with David Lewis’s theory of meaning, according to which semantic content is determined by fit and eligibility (Lewis 1983, 1984; see also Sider 2001 pp. xxi−xxiv and, especially, Sider 2012). Fit is a function of use – this is our contribution to content determination. Eligibility, on the other hand, is a contribution from the world. On Lewis’s picture, properties are ranked in terms of naturalness. Naturalness, like fit, comes in degrees: some properties are perfectly natural (being negatively charged, perhaps), some are less than perfectly natural (like being blue) and some (like being grue) are highly gerrymandered and unnatural. The more natural a property, the more eligible it is to serve as the semantic value of a predicate.21 Thus, Lewis claims that the content of a term is whatever does the best overall job of matching use and naturalness.22 On this “best candidate” theory, a natural property acts as a “reference magnet.” To illustrate: take the predicate “gold,” which expresses a natural property (not perfectly natural, but pretty natural). We could have differed slightly in our use of that term – we could have, for example, been somewhat more disposed to call fool’s gold “gold.” In that case, some other candidate semantic value – like the property of being gold or fool’s gold – might have been a better fit with use. But the actual semantic value – being gold – would still have fitted fairly well, and it would have done a much better job with respect to naturalness. Hence, it would still have done the best overall job. Hence, “gold” would still have meant gold, despite the difference in use. It is a very different story when it comes to non-natural properties. Take, for example, the case of “bald.” Once again, there are many candidate meanings for this predicate – there is the property of having fewer than 1,000 hairs on one’s scalp, the property of having fewer than 1,001 hairs on one’s scalp, and so on. All of these properties are equally natural, so the world does no work – the semantic content of “bald” is determined solely by use, which is why that predicate is so pliable. 21 22

The issue of what qualifies a property as perfectly natural is a contentious one; for further discussion see Lewis (1986, pp. 59–69) and Sider (1996). It is no part of the theory that exactly one candidate will score highest in both fit and eligibility. There may be ties. There may be cases where one property scores highest in fit and another scores highest in eligibility. What the theory says is that the reference relation picks out the property with the highest overall score.

78

ryan wasserman

We can now put the final worry as follows: on the epistemicist picture, moral predicates are more like “bald” than “gold.” And I think that a moral realist – a serious moral realist – should resist this conclusion. For the serious moral realist, moral terms pick out metaphysically important properties.23 The property of being morally obligatory, for example, is more like the property of being gold, than being bald. So, the semantics of moral terms should be more like the semantics of natural kind terms. So, moral terms are not semantically plastic. So, the epistemicist response must be denied. indeterminacy and subjectivism I will close by considering a second response to the Indeterminacy Argument. Let us begin with the following case, due to Frances Howard-Snyder: Agnes’s brakes fail. Should she continue straight into the busy intersection or should she swerve into a field? Add to the story what Agnes does not and cannot know, that the following two counterfactuals are true. If she were to enter the intersection, other cars would swerve and narrowly miss her car, no one would be hurt and she would stop on the slight hill across the intersection. If she were to swerve into the field, she would hit a man hidden asleep in the grass and kill him. (2005, p. 265)

The question is: should Agnes swerve? Some say “yes.” Some say “no.” And some say “yes and no.” In one sense, Agnes ought to swerve, since that would maximize expected utility. (It is the best option from her point of view.) In another sense, Agnes ought to go straight, since that would maximize actual utility. (It is the best option from the fully informed point of view.) Those who give the third response distinguish between the subjective “ought” and the objective “ought.” If “ought” is ambiguous in this way, that will have significant implications for our understanding of the Indeterminacy Argument. Consider again the Indeterminacy of Obligation. One is obligated to do what one ought to do, so if “ought” is ambiguous, we have two different readings of that principle. There is: The Indeterminacy of Objective Obligation: It is possible for questions about objective obligations to lack determinate answers.

And: The Indeterminacy of Subjective Obligation: It is possible for questions about subjective obligations to lack determinate answers. 23

This way of thinking about moral realism is hinted at in Wasserman (2004).

Personal identity, indeterminacy and obligation

79

Now return to the car insurance case. Remember, you have the option of buying car insurance for $1,000. The policy will cover the $10,000 of damages that will result when You* gets into an accident tomorrow. But it is indeterminate whether or not you are You* and thus indeterminate whether or not you will be the person in tomorrow’s car accident. Let us suppose that we are working with a non-epistemic reading of indeterminacy, so there really is no fact of the matter. Our question is: are you rationally obligated to purchase the car insurance? Or should you turn the salesman down? Objectively speaking, you ought to purchase it if and only if you will be in the car accident and you ought to decline it if and only if you will not be in the accident. Since there is no fact of the matter about whether or not you will be in the accident, there is no fact of the matter about what is to be done, objectively speaking. But that does not mean that there is no fact of the matter about what is to be done, subjectively speaking. In the subjective sense of “ought,” you ought to maximize expected utility. In order to calculate expected utility, you must first assign credences and utilities to the relevant outcomes. And this brings up a very interesting question: what credence should you give to P when P is indeterminate? There are various ways of thinking about this question. One could treat “indeterminate” as synonymous with “indefinite” and then think about the question as an epistemicist. But the question is most gripping when one employs a semantic or metaphysical reading of “indeterminate”: how should you distribute your credence across P and ¬P when there is no fact of the matter as to whether P or ¬P? I do not have a general answer to this question. But I will make a few specific suggestions about a few specific cases. First, imagine that your car has been placed in a car crusher and that you cannot free it until tomorrow. Unfortunately, the machine has just been turned on. Lucky for you, the machine happens to be an indeterministic system – a signal has been sent from the start button, but there is a 40 percent chance it will fizzle out before the crushing mechanism is engaged. (Of course, there is also a 60 percent chance that the signal will not fizzle, in which case your car will soon be crushed.) More good news: you happen to have an insurance salesman with you. He offers you a limited-term policy for $1,000 which will replace your $10,000 car in the event that it is crushed. Should you buy the insurance? Intuitively, the answer is yes. Of course you should buy the insurance! (At least if your utilities go with dollars and money is the only relevant consideration.) That is the intuitive reaction. Here is the theoretical support. First, we have the Principal Principle: subjective credence ought to match known

80

ryan wasserman Table 1 Crushed

Buy Do not buy

.6 × −$1,000 .6 × −$10,000

Not-crushed + +

.4 × −$1,000 .4 × $0

Expected utility = =

−$1,000 −$6,000

objective chance.24 So, in this case, you should assign .6 to the proposition that your car will be crushed and .4 to the proposition that it will not. Second, we have decision theory: once credences and utilities are set, the question of what you ought to do is just a matter of number crunching (see Table 1). When I say, with the decision theorist, that you ought to buy the insurance, I am using the subjective “ought.” Buying the insurance is what makes sense, given what you know. Objectively speaking, there is no fact of the matter, since what you ought to do depends on what the machine is going to do, and there is no fact of the matter about that. Now go back to the original car insurance case, where you are being put into a machine and where 40 percent of the molecules in your body are going to be swapped for Einstein cells. I suggest that these two cases are relevantly similar, so you should approach the cases in the same way. In both cases, it is indeterminate whether or not you are going to be stuck with a crushed car. In one case, that is because it is indeterminate whether or not the car will be crushed. In the other case, it is because it is indeterminate whether or not it will be you with the crushed car. But still, it is indeterminate in both cases. So, you should approach the cases in the same way. In particular, you should assign the same credences (.6 that you will survive; .4 that you will not) and use the same decision theory matrix. Now go back to the case of Frank and Frankenstein. Once again, it is indeterminate whether or not Frankenstein is the same person as Frank. And, once again, we are operating with a non-epistemic reading of indeterminacy. Our question is: ought we to punish Frankenstein? Or are we obligated to let him go? Objectively speaking, we ought to punish Frankenstein just in case he is guilty and we ought to let him go just in case he is not guilty. There is no fact of the matter as to whether or not he is the same person as Frank, so it is indeterminate whether or not he is guilty. So it is indeterminate what is to be done, objectively speaking. 24

For a more careful statement (and defense) of the Principal Principle, see Lewis (1980).

Personal identity, indeterminacy and obligation

81

Subjectively speaking, matters are more complicated. Note, first, it is difficult to assign utilities in this case. It is bad to punish the innocent and it is bad to let the guilty go unpunished. Many will say that the first is worse than the second, but it difficult to put dollar amounts on these outcomes. It is also difficult to assign credences in this case.25 You might think that, since 60 percent of Frank’s cells remain, we should be .6 on the proposition that Frank is Frankenstein (and .4 on the proposition that he is not). But that is too simplistic. Credences should not map cleanly onto cell replacement, or even onto degrees of continuity – after all, a .99 degree of continuity is presumably sufficient for identity over time, and determinately so. Despite these complications, there may well be some fact of the matter about what is to be done, subjectively speaking. For my own part, I am approximately .5 on the claim that Frank is Frankenstein. And I think it is quite a bit worse to punish an innocent person than to let a guilty party go. So I say, let him go. You might disagree with that verdict, but the main point is that there could still be a fact about what you should do, subjectively speaking, even if there is no objective fact in this case. To tie all of this back to the Indeterminacy Argument: I have suggested that there are two ways of understanding the Indeterminacy of Obligation, so premises (2) and (3) are both ambiguous. On the subjective reading, (3) is true, but (2) is false. As I have just argued, the Indeterminacy of Personal Identity is consistent with there being some determinate fact about what ought to be done, from the subject’s perspective. On the objective reading, (2) is true, but (3) is false. To deny (3) is to accept the Indeterminacy of Objective Obligation, but that seems like an acceptable consequence. Crucially, objective indeterminacy does not entail divine shrugging, for God can still offer advice on what He would do, were He in your shoes. That is just to say that there might still be some fact about what is to be done, subjectively speaking. If personal identity is genuinely indeterminate, then that is all we can ask for. Perhaps that is all we need.26 25 26

This second worry also applies in the previous car accident case. I thank the audience at the 2010 conference on personal identity in Obergurgl, Austria. Special thanks to the conference organizers at the University of Innsbruck. I also thank Berry Crawford, Dan Howard-Snyder, Frances Howard-Snyder, Hud Hudson, Ned Markosian and Dennis Whitcomb – all of whom read this chapter and provided me with helpful comments.

chapter 4

Personal identity and its perplexities Harold W. Noonan

introduction What is the problem of personal identity? What is the distinction between the simple and complex views of personal identity? What distinguishes the sortal concept of a person from other sortal concepts? How can we account for indeterminacy in personal identity? First, I characterize the problem of personal identity in a way that conforms to Lewis’s dictum (1986) that there are no problems about identity. Second, I build on this to draw the distinction between simple and complex views of personal identity. Third, I argue that the concept of a person is indexical and appeal to this contention to defend the neo-Lockean psychological continuity account against the animalist’s “too many minds” objection. Fourth, I argue that we cannot account for indeterminacy in personal identity in terms of indeterminacy in identity and that complex theorists should accept that it requires a multiplicity of thinkers.

the problem of personal identity Lewis denies that there are problems about identity: “We do state . . . genuine problems in terms of identity. But we needn’t state them so. Therefore they are not problems about identity” (Lewis 1986). If what Lewis says is correct, the traditional problem of personal identity should be reformulated as a question that is not about identity. In what follows I offer such a reformulation. Forget about personal identity for a moment and think about another familiar puzzle case: that of the statue and the clay. Statues can survive changes pieces of clay cannot and vice versa. Hence, a statue (Goliath) and the piece of clay from which it is made (Lumpl) may be actually at all times coincident but possibly divergent. Our concept of a statue plausibly implies 82

Personal identity and its perplexities

83

that no statue can survive radical reshaping. Our concept of a piece of clay plausibly implies that any piece of clay must survive radical reshaping in which all its matter is preserved in one coherent whole. These propositions specify persistence conditions for statues and pieces of clay and, as they illustrate, these persistence conditions are of two types. The proposition that no statue can undergo radical reshaping can be expressed as: Necessarily, if x is a statue then if the matter that constitutes x at t is radically reshaped at t, then x ceases to exist.

This specifies a “passing-away” condition for statues (this terminology comes from Penelope Mackie, personal communication). The proposition that any piece of clay must survive radical reshaping in which all its matter is preserved in one coherent whole can be expressed as: Necessarily, if x is a piece of clay then if the matter that constitutes x at t is radically reshaped at t but preserved in one coherent mass, x survives.

This specifies a “preservation” condition for pieces of clay. Sortal concepts for persisting things are governed by such conditions because they constrain the histories of the things they apply to, and such constraints can always be expressed in the form: Necessarily, if x is an S then if x exists at t and t* then Rxtt*,

or in the form: Necessarily, if x is an S then if Rxtt* x exists at t and t*.

The “passing-away” condition for statues is entailed by a principle of the first form (stating that a statue cannot have radically different shapes at different times) and the “preservation condition” for pieces of clay is entailed by a principle of the second form (stating that if the matter composing a piece of clay is in one coherent mass at both of two times, whatever shape it is in, the piece of clay exists at both times). These principles state necessary conditions for being a thing of sort S. What distinguishes sortal concepts under which persisting things fall from non-sortal concepts under which persisting things fall (even ones that necessarily apply to a thing at any time it exists, like permanent bachelor) is that they are governed by such de dicto modal principles (Mackie also argues that sortals provide (de dicto) passing-away conditions whereas concepts like permanent bachelor do not, personal communication).

84

harold w. noonan

Questions about personal identity over time can similarly be rephrased as questions about necessary conditions of membership of the kind person. One question we can ask, without mention of identity, is: (Q1) What conditions R satisfy the following schema: (P1) Necessarily, if x is a person then if x exists at t and t*, Rxtt*?

Another is: (Q2) What conditions R satisfy the following schema: (P2) Necessarily, if x is a person then if Rxtt* x exists at t and t*?

Solutions to the problem of diachronic personal identity will take the form: “all and only the following conditions satisfy (P1): . . . and all and only the following conditions satisfy (P2): . . .” Of course, in answering these questions we only specify some of the necessary conditions of personhood. One other type of necessary condition can be thought of as a synchronic constraint, capturable in the form: Necessarily, if x is a person then if x exists at t, Fxt,

where F represents an expression for a non-historical property, i.e. one entailing nothing about the past or future. In addition the general problem of the criterion of identity for persons can be formulated by asking for a specification of the relation R satisfying the condition: (1) Necessarily, if x is a person and y is a person then (x = y iff Rxy).

(1) is equivalent to the conjunction of: (1a) Necessarily, if x is a person then Rxx

and (1b) Necessarily, if x is a person then if y is a person and Rxy then (x = y).

Together these imply that necessarily any person is (identical with) the person R-related to it. They thus give a sufficient condition for identity with any person by giving a necessary condition of personhood (to say that any person is the person R-related to it is to say something of the form “any person is . . .” which is the form of a statement of a necessary condition). Thus all intelligible questions about personal identity are equivalent to questions about the necessary conditions for being a person.

Personal identity and its perplexities

85

the distinction between the simple and the complex view of diachronic personal identity I now offer an account of the distinction between the simple and the complex views of personal identity which conforms to the Lewisean dictum.1 Traditional defenders of the complex view are Locke and Hume, defenders of the simple view Butler and Reid. We also think of Chisholm and Swinburne as defenders of the simple view and Shoemaker, Parfit, and Lewis as defenders of the complex view. But how is the distinction to be characterized? One difference is that defenders of the simple view emphasize the distinction between diachronic personal identity and the identity of other objects; they insist that in the case of the other familiar things that figure in philosophical enquiries about identity – ships, plants and so on – the correct view is the complex one. We therefore need an account of the distinction which allows us to speak generally of “the complex/simple view of the diachronic identity of things of sort S” where S is a sortal term. Recall two of the types of constraint on personhood just distinguished. There are synchronic constraints capturable in the form: Necessarily, if x is a person, then if x exists at t, Fxt,

where F represents a term for a non-historical property. And there are diachronic constraints capturable in the form: Necessarily, if x is a person, then if x exists at t and t*, Rxtt*.

At first pass, the complex view is that there are diachronic constraints of this form not logically equivalent to or logically implied by the constraint “x exists at t and t*” or “x is a person and x exists at t and t*” and not entailed by the totality of synchronic constraints. If we call a diachronic constraint on personhood logically equivalent to or logically implied by one of the two just mentioned “trivial,” and call a diachronic constraint entailed by the totality of synchronic constraints “redundant,” we can put this by saying that the complex view is the view that there are non-trivial, non-redundant diachronic constraints on personhood. The simple view is that there are none. This fits well with the classification of the complex theorists listed above. 1

Parfit (1984) proposes that the simple view of personal identity is the view that personal identity does not consist in any “further fact.” Arguably this trivializes it: for any x, the fact that x = x does not consist in any further fact, so since for any x and y, if x = y the fact that x = y is the fact that x = x, for any x and y, if x = y the fact that x = y consists in no further fact.

86

harold w. noonan

For Locke, for example, personal identity consists in the relation of coconsciousness. So if a person x exists at times t and t*, x will be co-conscious at t with x at t*, a constraint which is non-trivial and non-redundant in the sense explained. Again, the familiar modern-day psychological continuity accounts satisfy this definition of a complex view, as do the competing nonpsychological, or bodily, accounts. But this characterization of the complex view is inadequate. One version of the simple view is that persons have bodily parts so they are not souls, but personal identity is constituted by identity of soul (Swinburne and Shoemaker 1984). This implies a non-trivial, nonredundant diachronic constraint on personhood: if x is a person who exists at t and t* then x has at t the same soul that x has at t*. This is non-trivial, since it is not logically equivalent to “x (is a person and) exists at t and t*” (even if it is metaphysically equivalent in some sense). It is non-redundant since even if a person needs a soul whenever he exists he may have different souls at different times. Nevertheless, the view that personal identity is constituted by soul identity (if nothing more can be said about the latter) must be classified as a version of the simple view of diachronic personal identity. To accommodate this point call a diachronic constraint on personhood, Rxtt*, “identity-involving” if its satisfaction requires that something which is not a person exists at times t and t*. That x has at t and t* the same soul is an identity-involving diachronic constraint. We may now say that the complex view is that there are non-trivial, non-redundant, non-identityinvolving diachronic constraints on personhood. The simple view is that the only non-trivial, non-redundant diachronic constraints on personhood are identity-involving.2 This characterization of the distinction accords with the widespread idea that the simple view is that the diachronic identity of persons cannot consist in the holding of any relations other than the relation of identity itself. But I have to consider an objection. Swinburne suggests an account of personal identity according to which personal identity is constituted by soul identity and the criterion of identity of souls, as (Aristotelian) substances, is 2

More precisely, the simple view is that the concept of a person is the concept of a sort of persisting object not governed by non-trivial, non-redundant, non-identity-involving diachronic constraints. The complex view is that the concept of a person is the concept of a sort of persisting object which is governed by such constraints. A third view is that the concept of a person is not a sortal concept at all. The proponent of the simple view needs to explain what distinguishes the concept of a person from non-sortal concepts (so does the proponent of the complex view, but for him it is easy).

Personal identity and its perplexities

87

identity of form and continuity of immaterial stuff (Swinburne 1984, p. 27). On this proposal there is a non-trivial, non-redundant, non-identity-involving diachronic constraint on personhood: if person x exists at t and t* the soul which x possesses at t must have continuity of immaterial stuff with the soul which x possesses at t*. This is not an identity-involving constraint, since continuity of stuff does not require identity of stuff, and by itself, without identity of form, does not entail identity of substance, i.e. soul. I think the thing to say is that this “Aristotelian” account is, indeed, contrary to Swinburne’s intention, a version of the complex view. Why not? It is a dualist view. But this does not rule out complexity. Locke’s account of personal identity is dualist: it explains personal identity as consisting in coconsciousness of immaterial thinking things that “think in” persons. But it is paradigmatically complex. A feature of the “Aristotelian” account not possessed by Locke’s is that it entails that if a person exists at two times, some other entity – his soul – which determines his identity, exists at those times. But this is also a feature of the brain account of personal identity according to which, although persons are not brains, a person persists just in case his brain persists (whether or not it carries psychological continuity with it), which is paradigmatically a version of the complex view (brains are paradigms of material objects whose persistence does not require persistence of matter but requires, but is not entailed by, material continuity). Swinburne’s “Aristotelian” theory, according to which souls can undergo a complete replacement in their psychologies, but, though only parts of persons, determine personal identity, is a sort of dualist version of the brain account. I conclude that we can define the complex view of personal identity as the view that there are non-trivial, non-redundant, non-identity-involving diachronic constraints on personhood. So defined it comes in two varieties. According to the first, if a person exists at times t and t* there exist at those times some entities distinct from the person, the obtaining of a certain relation between which suffices for the person’s existence at both times. Locke’s view is of this variety. If a person exists at t and t* then there are thinking substances S and S existing at t and t*, such that S at t is co-conscious with S* at t*. The Lewisean psychological continuity account has the same structure: if a person exists at t and t* there are person-stages existing at t and t* which are psychologically continuous (R-related). Complex accounts of diachronic personal identity of this variety are “two-level”; they ground the identity of persons in a relation (other than identity) between non-persons. According to the second variety of the complex view, the existence of a person at two times does not require the

88

harold w. noonan

existence at those two times of any entities distinct from the person (which are not adjectival on, or states of, the person) related in a way that suffices for the existence of a person at both times. Shoemaker’s account of personal identity is of this kind, as also, perhaps, is Van Inwagen’s.3 Since complex accounts of this second type are possible, denying that personal identity is grounded in a relation between other things does not entail acceptance of the simple view. I finish this section with a comment on the link between the simple view and the possibility of indeterminacy in diachronic personal identity. Standard puzzle cases of diachronic personal identity are often described as cases where, though everything else is clear, diachronic personal identity is unclear. If the Brown/Brownson case (Swinburne and Shoemaker 1984), in which Brown’s brain is transplanted into Robinson’s body with consequent transference of psychological traits, is thought of as such, for example, then the situation must be one in which it is perfectly determinate that there is exactly one person where Brown is before the transplant, perfectly determinate that there is exactly one person where Brownson is after the transplant, but indeterminate whether Brownson is Brown, hence indeterminate whether there is someone, exactly one person, where Brown is before the transplant and where Brownson is afterwards. According to either variety of the complex view, as I have explained it, this indeterminacy can be due to indeterminacy whether a non-trivial, nonredundant, non-identity-involving diachronic constraint on personhood, Rxtt*, is satisfied, where t and t* are the two times pre- and post-transplant. But according to the simple view there are no non-trivial, non-redundant, non-identity-involving diachronic constraints on personhood. So, given that everything but the fact of diachronic identity is clear, the simple theorist can account for the indeterminacy only by saying that it is indeterminate whether some identity-involving constraint on personal identity, Rxtt*, is satisfied, e.g. that it is indeterminate whether the soul x possesses at t is the soul x possesses at t*. In this case there cannot be any indeterminacy in the terms “the soul x possesses at t ” and “the soul x possesses at t*,” for everything is clear except the fact of diachronic personal identity. Indeterminacy in diachronic 3

Shoemaker’s neo-Lockean psychological continuity account of personal identity is neutral between materialism and dualism. So the existence of a person at two times does not require, on this account, the existence at the times of brains or material particles. Shoemaker is also opposed to the fourdimensionalist Lewisean position. Finally, in allowing the possibility of dualism he is not allowing the possibility of a Lockean position according to which as well as persons there are thinking substances which “do the thinking for” persons (Shoemaker 1985). Van Inwagen (1990b) thinks that human persons are human animals and that in the material world there are at most living organisms and simples. But he does not claim that the existence of organisms requires the existence of simples.

Personal identity and its perplexities

89

personal identity according to the simple view as characterized above, then, must be accounted for in terms of indeterminacy in identity itself, in a way that offends against Evansian (Evans 1978) sensibilities.4

the indexicality of the concept of a person So far nothing I have said about the concept of a person requires it to be fundamentally different from any other sortal concept. But it is different, because it is indexical. We can formulate the problem of personal identity over time by asking: “What are the constraints on the history of a person?” Or each of us can formulate it for himself by asking, “What are the constraints on my history?” The latter, indexical, formulation is the basic one. Our interest in personal identity is, of course, fundamentally an interest in our own identity (the problem of personal identity is more illuminatingly titled “the problem of the self’s identity”). Thus “person” as it occurs in standard formulations of the problem just means “object of first-person reference,” i.e. that which each of us calls “I.”5 Thus a gap exists between the concept of a person and the concept of a thinker of first-person thoughts. That a thinker of first-person thoughts is an object of first-person reference is not trivially analytic. The neo-Lockean psychological continuity theorists (Lewis, Parfit, Shoemaker et al.) can exploit this gap to respond to the “thinking animal” or “too many thinkers” objection. The problem of the thinking animal (Olson 1997b), briefly, is that it seems indisputable that all normal healthy adult human beings are thinkers. But so, by definition, are persons. However, according to the psychological continuity theorist, persons are not human beings (they differ in their persistence conditions). So the psychological continuity theory entails the existence of too many thinkers. Moreover, it creates an impossible epistemic problem: how do I know that I am the person and not the coincident human animal 4

5

Evans denies that statements of identity can be indeterminate when both singular terms are determinate in reference. This is what we think possible in the case of statements about friendship, for example. But Evans takes to be ruled out what correspond to two prima facie possibilities in the case of friendship: that “friend of” is indeterminate in reference or that it is epistemically indeterminate in the sense that though it has a particular reference it does so unknowably because it could easily have had another. What remains is just the possibility that the relation of identity (the relation of friendship) itself is vague. Thus Evans sees himself as arguing against the possible ontic indeterminacy of identity. Note that the move from the indexical to the non-indexical formulation embodies the substantial assumption that there is some kind of thing for which persistence conditions can be given to which all objects of first-person reference belong.

90

harold w. noonan

thinking falsely that he is a person? Finally, if human animals, in addition to persons, are thinkers, they must be persons after all, since their thoughts have whatever complexity and sophistication any definition of “person” could require – they have the same thoughts, after all, as the persons with whom, on the psychological continuity account, they “cohabit.” So the neoLockean’s attempt to identify the persistence conditions for persons collapses into incoherence, since he must acknowledge different kinds of person with different persistence conditions. I think a defender of the neo-Lockean position can respond to the first charge against it – that it entails the existence of too many thinkers – by accepting that his theory entails that in a case in which the line of psychological continuity goes one way and the line of biological continuity goes another, there are two temporarily coincident non-identical entities present – the person and the human animal – each of which is, say, thinking that it will rain. But this is no more problematic, he can say, than accepting that there are two coincident non-identical entities present, each of which weighs 140 pounds, is flesh-colored, is breathing and is walking,6 which is in turn no more problematic than accepting that the temporarily coincident statue and piece of clay are two non-identical entities each of which weighs five pounds, is brown and is being carried across the room. The same kind of situation can exist as far down the evolutionary tree as there are interesting psychological differences between members of the same species. Thus, if in the brain swap story Brown and Robinson are two dogs, there are two temporarily coincident non-identical entities where Brown is before transplant, each of which weighs 50 pounds, has a thick black coat, is breathing and walking and thinking about a bone. The “too many thinkers” objection is no more persuasive than a “too many walkers” objection. To the charges that the neo-Lockean psychological theory creates in the specific case of persons – objects of first-person reference – an impossible epistemic problem, and collapses into incoherence since it entails the existence of different kinds of person with different persistence conditions, the answer must be more complex. I claim: (a) that the most sensible thing for the psychological continuity theorist to do in response to these charges is to distinguish the concepts of an object of first-person reference and a thinker of first-person thoughts and to say that when an animal coincident with but distinct from a person is a 6

So if the psychological continuity theorist is confronted with the challenge of the thinking animal or the problem of too many thinkers, he is also confronted with the challenge of the animate person or the problem of too many animals.

Personal identity and its perplexities

91

thinker of a first-person thought, not he, but the coincident person, is the object of the thought and (b) that this is not a reductio of the psychological continuity theory. My argument is as follows: (1) All persons and only persons are objects of their own first-person reference. I take this to be trivially analytic, given that the concept of a person is an indexical concept. (Sometimes, of course, we pretend that something which is not a person is an object of first-person reference: I put a sign on my door saying “I am unlocked, please come in and wait.”) (2) All persons are psychological continuers. This is a formulation of the standard neo-Lockean view. (3) Some normal healthy adult human animals are not psychological continuers. This is undisputed by neo-Lockeans. (4) All normal healthy adult human animals are thinkers of true firstperson thoughts. This is what the animalist7 urges on the neo-Lockean as an evident truth. Some neo-Lockeans, notably Shoemaker (2007), deny it. I explain below why I disagree with Shoemaker. From (1)–(4) follows: (5) Some normal healthy adult human animals have first-person thoughts which are not about themselves but about psychological continuers with which they are not identical. I claim that (5) is what the neo-Lockean, the defender of (2), must say or deny a trivial analytic truth, (1), or a plain empirical fact, (3), or an evident truth, (4). But is this not a reductio of neo-Lockeanism? I argue not. Given the triviality of (1) the substantive neo-Lockean claim (2) becomes the claim that objects of first-person reference are psychological continuers. It follows (given empirical facts and evident truth) that there are more thinkers 7

Animalists say that we are animals, understood as entities with purely biological persistence conditions. Much of the plausibility of animalism derives from its apparent consonance with the common-sense thoughts that we are animals and that animals are the subjects of psychological attributes. But it is not obvious that animals have purely biological persistence conditions. When the brain or cerebrum of my dog, Brown, is transplanted, I think that my dog goes with his brain or cerebrum and acquires a new body. I do not deny that persons are animals in the same sense that dogs are. But neither are biological animals in the sense of entities with purely biological persistence conditions. Nevertheless, I think that entities with purely biological persistence conditions are also the subjects of psychological attributes and hence the psychological continuity theorist must accept that there are many coincident thinkers.

92

harold w. noonan

than common sense acknowledges, some of which are not persons. But it is hardly news that neo-Lockeanism leads to this unless, like Shoemaker’s version, it denies what seems evidently true elsewhere. Locke himself distinguished between thinking substances and persons,8 and so acknowledged the existence of thinkers distinct from persons that were not objects of first-person reference. And, of course, the four-dimensional version of the psychological continuity account, defended by Lewis, endorses a multiplicity of thinkers that are not objects of first-person reference (Lewis 1979, 1981) and entails that some thinkers of first-person thoughts are not their objects. But perhaps what is wrong with my response to the problem of too many thinkers is that it creates a mystery. How can a normal healthy adult human being, capable of sophisticated first-person thought, not be capable of referring to himself in the first-person in such thought, if the psychologically indistinguishable person is? How can the human animal lack this capacity which its psychologically indistinguishable twin has? But to ask this is to misunderstand the proposal. The neo-Lockean claims that it can be established by conceptual analysis that the following is a de dicto necessary truth: All objects of first-person reference are psychological continuers. Or: Only psychological continuers are objects of first-person thought.

Consequently, if A is a human animal it is a de dicto necessary truth that: If A is not a psychological continuer A’s first-person thoughts are not thoughts about A.

It does not follow, given that A is not a psychological continuer, that A is necessarily or essentially something whose first-person thoughts are not about itself – no de re necessity follows. This is so since the neo-Lockean story does not entail that A could not have been a psychological continuer. To say that A is not a psychological continuer is to say something about A’s history, and, consistently with neo-Lockeanism, A could have had a different history. The sense in which the neo-Lockean account entails that A cannot think “I”-thoughts about itself is just that, qua something which is not a psychological continuer, its “I”-thoughts cannot be about itself, just as qua someone who is not Prime Minister, I cannot be correctly referred to by the Queen as “my first Minister.” 8

Locke did not have to confront the thinking animal problem since he did not ascribe thoughts to men.

Personal identity and its perplexities

93

Discomfort may remain. According to the neo-Lockean, something that is not a psychological continuer cannot be an object of its own first-person thoughts, no matter how sophisticated its thoughts.9 It cannot think that it itself is F, though it can think of itself that it is F (as Perry’s shopper (1979) thinks of himself that he is making a mess, but does not think that he himself is making a mess).10 But why should this be so? However, the explanation is simple: to be a psychological continuer is to have a certain kind of history. It is like being a past and future Prime Minister. But whether something is a thinker now and the level of sophistication of its present thoughts cannot depend upon its indefinitely long-distant future or its indefinitely longdistant past. So no matter how sophisticated a creature’s present thoughts, they cannot insure that it is a psychological continuer. Consequently, given the neo-Lockean thesis (2), they cannot insure that it is an object of its own first-person thoughts. As a psychological continuer theorist I think that human animals think “I”-thoughts whose referents are coincident but non-identical persons because I deny that whether something is a thinker now, and the level of sophistication of its present thoughts, can depend upon its long-distant past or future. On this point I disagree with Shoemaker. Shoemaker thinks that (biological) human animals are not thinkers at all, given the difference in their histories, despite having, in the sense the animalist emphasizes, “what it takes.”11 This is perfectly coherent. Many properties are had by things at times in virtue (partly) of what happens at other times. This holds of the property being a widow. It also holds of the property being a statue. Whether something is a statue depends on its originating in some maker’s intentions, so the matter now constituting a statue cannot count as a statue just because it coincides with one. It is equally implausible that the matter momentarily 9 10

11

This is also trivially true on Shoemaker’s account, since on that account it is not possible for something that is not a psychological continuer to think (first-person thoughts) at all. X thinks that he himself is F in the relevant sense if he thinks a first-person thought and in doing so thinks about himself. Perry’s shopper thinks about himself but not in the first-person way; the sophisticated animal thinks in the first-person way but not about itself. Geach (1976) suggests that we explain “X thinks that he himself is F” in a Fregean manner as “for some way of thinking α, α = ego(X) and X thinks that α is F.” Here ego(X) stands for the way of thinking X employs when he thinks in the first-person way. In the case of the sophisticated animal the way of thinking it employs when it thinks in the first-person way is not a mode of presentation of itself. Similarly, when a Lewisean person-stage employs a first-person way of thinking in thinking about the future, its thought is not a thought about itself but about the perduring person of which it is a part. If we imagine hypothetical beings that are otherwise identical but differ from psychological continuers only in their futures, e.g. because they come into existence fully formed, Shoemaker would still deny that they were thinkers, because of their differing futures. In fact, he would deny that human beings permanently coincident with persons were thinking things (just as he would deny that Goliath was Lumpl).

94

harold w. noonan

coinciding with a person or an animal thereby qualifies as a person, an animal or a thinker. It is thus perfectly consistent to say both that I, a thinking being, currently coincide with a human animal distinct from me, and that that animal is not a thinker, because like the matter which currently constitutes both of us, it is not thinking my thoughts or any thoughts. This is Shoemaker’s position. But the difficulty is this. I and the animal with which I currently coincide have been and will be coincident for many years. So if I am thinking it is raining now and the coincident human animal is not, why is this so? The difference must either be ungrounded or grounded in some far future difference between us (but there need be none) or grounded in some long-past difference between us (e.g. such as the fact that I was never a fetus and the human animal was) or grounded in some difference in unactualized potentialities. But neither the thought that the difference between a thinker and a non-thinker may be ungrounded, nor the thought that that difference can be grounded merely in a difference in unactualized potentialities seems acceptable. And though properties can be possessed at times in virtue of what happens at other times, such long-past differences as are present in this case seem irrelevant. Of course, there are properties closely related to the property of thinking it is raining that I now possess and the human animal does not if it is not a person: for example, the property being a person thinking it is raining. And it may be suggested that the property we in fact attribute with the predicate “is now thinking it is raining” is a property of the form is a person who now has property P. But although this is a possible response to the thinking animal objection to the psychological approach it implausibly requires that far-distant past or future differences are relevant to whether something now satisfies the predicate “is a thinker.” Although it is absurd to say that the matter now momentarily coinciding with me is thereby thinking it is raining, it seems far from absurd to say that the (biological) human animal with which, on Shoemaker’s view, I have been and will be so long coincident, is now thinking it is raining. Surely it “has what it takes,” in terms of present equipment, and sufficiently appropriate past and future? Thinking, like breathing and walking, takes some time, which is why it is absurd to describe the momentarily coincident parcel of matter as doing any of these things. Whether something qualifies as breathing or walking or thinking at a time depends on its properties in an extended period including that time.12 But the period does not have to be years long. So Lewisean day-long

12

A defender of Sider’s stage ontology (2001) would rather say “depends on properties possessed at other times by its temporal counterparts.”

Personal identity and its perplexities

95

person-stages can count as walking and thinking (Lewis 1981, p. 76),13 as can the animal which is coincident with me for so much longer than a single day. So I accept that human (and other) animals are thinking things, accept the multiplicity of thinkers and resist the animalist’s attempted reductio of the psychological continuity account at another point.

the complex view and indeterminacy All complex theorists accept the possibility of indeterminacy in personal identity. It is part of the complex view that personal identity is not radically different from the identity of other familiar types of thing that figure in philosophical puzzle cases about identity – ships, statues, plants and so on. But the possibility of indeterminacy in these cases is undeniable. Moreover, to deny the possibility of indeterminacy in personal identity the complex theorist must make a large difference – the difference between life and death – depend on a trivially significant one, a fraction more or less brainmatter, one more or less insignificant memory, etc. I want to argue, first, that complex theorists cannot regard the source of this indeterminacy as indeterminacy in identity itself – even if identity can be indeterminate. In fact no one can (and so, unsurprisingly, the simple theorist, who must do so, must deny the possibility of indeterminacy in personal identity). There are two familiar types of case in which according to complex theorists vagueness is possible for persons. First, a person’s spatio-temporal boundaries may be vague. In this respect people are like clouds or mountains or (other) animals. Second, it may be vague whether a person existing at one time is a person existing at another. If the brain transplant is botched it may be indeterminate whether Brown, the person in room 100 before the operation, is Brownson, the person in room 101 after the operation. In this case it is clear that there is exactly one person at a certain time with a certain property, clear that there is exactly one person at a certain time with another property, but indeterminate whether the person with the first property at the first time is the person with the second property at the second time. It is again uncontroversial that we can construct puzzle cases about the identity of entities other than persons with this structure. Familiar examples 13

Lewis adds “a stage cannot do everything that a person can do, for it cannot do those things a person does over a longish interval.”

96

harold w. noonan

in the literature are Shoemaker’s case of the four-centuries-old bridge of Santa Trinita in Florence, Parfit’s example of the reconstituted club and Shoemaker’s spatial example of Alpha Hall and Beta Hall and the linking walkway (Shoemaker and Swinburne 1984). Now in all these cases we can formulate statements involving the concept of personal identity which are vague. It will be vague in cases of the first sort whether the person in place P1 at time t1 is identical with anyone alive at another time t2. And in cases of the second type it will be vague whether the person in place P1 at time t1 is identical with the person in place P2 at time t2. But there will also be indeterminate statements not involving the concept of personal identity. In both cases it will be indeterminate whether some person is at place P1 at time t1 and at place P2 at time t2 (though in the second case it will also be determinate that there is some person (exactly one) in place P1 at time t1 and determinate that there is some person (exactly one) in place P2 at time t2). Other statements not involving personal identity will similarly be indeterminate, e.g. whether someone is fat at one time and thin at another. These statements can be trivially reformulated to bring in the concept of identity. But it does not follow that their indeterminacy can be explained by appeal to the indeterminacy of identity. Suppose it is indeterminate whether the cleverest boy in the school is tall. Now this statement is logically equivalent to one in which the concept of identity is employed – “the cleverest boy in the school is identical with someone tall” – but indeterminacy in identity can be no part of the explanation of its indeterminacy. It may be indeterminate because “the cleverest boy in the school” is indeterminate in reference between a tall referent (who, say, excels at math) and a short referent (who excels at English). Or it may be that, in accordance with the epistemic view, “the cleverest boy in the school” refers to a single (tall) person but unknowably so because there is another (short) candidate which it would have referred to if our linguistic practices had been marginally different (if we had counted ability at math as slightly less important) but we cannot know that our linguistic practices are not thus marginally different. Or it may be that “the cleverest boy in the school” refers to a single vague object with a rather large region of spatial (vertical) indeterminacy.14 Or it may be that “the cleverest boy in the school” determinately and knowably refers to a perfectly 14

To make sense of this last possibility consider this situation. A mouse has a partially severed tail, which is no longer clearly part of it. Some philosophers suppose that there is a single vague object which is determinately the reference of “the mouse” but is indeterminate in spatial extent (so it is indeterminate whether it is 5 inches or 12 inches long).

Personal identity and its perplexities

97

precise object, a boy who is manifestly intellectually superior to every other boy in the school but is of middling to tallish height and that the predicate “is tall” is vague so that the statement is indeterminate. What is clear is that the explanation of the indeterminacy of the statement that the cleverest boy in the school is tall cannot be that the relational expression “is identical with” or the concept of identity or the relation of identity is in some way vague (in a way that it might plausibly be thought that either the relational expression “is a friend of” or the concept of friendship or the relation of friendship is in some way vague), since this relational expression does not occur in the statement. If a statement is indeterminate some expression occurring in it must be the source of the indeterminacy (by what it refers to being referentially or epistemically indeterminate or by determinately referring to an entity which is in itself indeterminate in its nature); it can be neither sufficient nor necessary that some expression not occurring in it is characterized by indeterminacy.15 It is the same with the statements given above, not involving any expression for identity, which are indeterminate in the puzzle cases of personal identity. In these cases the crucial concept involved is that of a person (we can suppose that the application of all the other concepts is clear).16 Once again the source of the statement’s indeterminacy in truth-value may be indeterminacy in reference (indeterminacy in the reference of “person”), epistemic indeterminacy (ignorance about the reference of “person”) or ontic indeterminacy (the existence of vague objects which are indeterminate examples of personhood). But its source cannot be any indeterminacy in identity since the statements do not contain any expression for identity. That this is so, of course, follows from the thesis argued above, that questions about personal identity are not questions about identity but about the conditions of personhood. If so, indeterminacy in their answers cannot be due to indeterminacy in identity – but I have here argued for this consequence independently.

15

16

There are counter-examples to this claim, but ones clearly irrelevant to our concerns. If my shirt is borderline blue “My shirt is in the extension of ‘blue’” is indeterminate but contains no vague expression (it contains the quotation name of the word “blue” not the word “blue”). Alternatively, the indeterminacy of “some person exists at t and t*” may be located in the vagueness of the relational expression “exists at,” which may be thought to be vague in the way “friend of” is vague. But in the botched brain transplant it is also determinately true that there is exactly one person who exists at t and exactly one person who exists at t*. However, vagueness in “exists at” may be appealed to to explain vagueness in explicit statements of spatio-temporal boundaries – the idea being that one’s extent is vague in the way that the extent of one’s circle of friends is vague. I put this implausible idea aside in what follows.

98

harold w. noonan

What I wish to argue now is that whatever account of indeterminacy he offers any complex theorist must accept the existence of a multiplicity of coincident or nearly coincident thinkers in cases of indeterminacy in personal identity or allow that tiny differences in spatial extent or far-off future or past differences between some pairs of temporarily (nearly) coincident entities can disqualify one of them from being a current thinker. Let us think about the first kind of vagueness noted above, vagueness in spatio-temporal boundaries. If persons have vague boundaries then on any account of their vagueness that locates it in indeterminacy of reference this is a matter of the existence of a multiplicity of candidates, largely coincident in spatio-temporal extent, between which our singular terms are indeterminate in reference.17 On an epistemic story there must likewise be a multiplicity of such candidates in nearly the same place and time. These will all now have what it takes to be thinkers unless it is allowed that the tiny differences in their spatial extents or far-off past or future temporal boundaries matter to whether they count as thinkers now. To deny the multiplicity of thinkers it seems that the complex theorist who does not want to allow this must insist that there is in fact in any such case only one person candidate present – an object with ontically indeterminate boundaries. But can this view be maintained? Let us call the (putatively unique) person candidate present in such a case “T.” The view is that “T” determinately refers to a single vague object, a person, an object whose boundaries are either or both spatially and temporally fuzzy. There may indeed be many spatio-temporally overlapping precise objects in the vicinity of T, but none of them is T, since they are all determinately not persons, but at most person-constituters. Persons are vague objects and there is only one such object where T is. An alternative view, still in conformity with the existence of vague objects, is: “T” is a singular term of indeterminate reference. There is a multitude of spatio-temporally imprecise, massively overlapping vague objects between which its reference is indeterminate (none is determinately a person, since they are all determinately distinct from one another and there is determinately only one person where T is). Only vague objects can qualify as persons: hence any precise objects in the vicinity – if any – are at most personconstituters, not persons. But this is not enough to secure “T” a determinate reference, since there are many massively overlapping vague objects roughly 17

In accordance with the previous footnote I assume that it is not the relational expression “exists at” which is indeterminate in reference. I leave it as an exercise for the reader to see what effect discarding this assumption has on the following arguments.

Personal identity and its perplexities

99

where T is and hence “T” is indeterminate in reference between them. On this alternative the answer to the question whether “T” or T is vague is: both. Is the view that there must be a unique vague object in what we think of as the vicinity of T defensible? It is hard to see how the possibility of massively overlapping vague objects, like the possibility of massively overlapping precise objects, can be denied. If persons and other animals are vague objects, conjoined twins provide an example. Consider also this variant on the familiar example of Tibbles and Tib, the “tail complement” of Tibbles, i.e. that object which is all of Tibbles without his tail.18 Suppose Tibbles has been in a scrap, and both his ears are partially severed so it is indeterminate whether they are parts of Tibbles any more. According to the vague object theorist, Tibbles is a vague object. So is Tib. So Tib and Tibbles are two massively overlapping vague objects. Consider now the left-ear complement of Tibbles and the right-ear complement of Tibbles. These are also vague objects (it is indeterminate whether the right ear is part of the left-ear complement) and they are distinct from Tibbles and each other since parts that are determinately not parts of, for example, the left-ear complement are indeterminately parts of Tibbles and the right-ear complement. So they are massively overlapping vague objects, differing only with respect to the region of indeterminacy. (Examples of temporally overlapping vague objects can be similarly imagined.) Returning to the case of T, we can argue as follows: if there is a vague object where T is, there could be an exactly similar vague object elsewhere. If so, there could instead be a vague object elsewhere very, but not exactly, similar, differing perhaps only in the region of its indeterminacy (as the leftear complement of Tibbles differs from the right-ear complement of Tibbles). But if there could be such a very similar vague object elsewhere there could be a vague object exactly similar to it where T is, with a location massively overlapping the location of T. Moreover, suppose, with the first, that is, the unique vague object, view, that there is only one vague object in the vicinity of T, a person to which “T” determinately refers. There is a counterfactual community using singular and general terms similar to our terms “T” and “person” – let these be “T*” and “person*” – which they apply very largely as we do “T” and “person” but with some slight differences. Some few particles we say are indeterminately parts of T they say are determinately parts or not parts of T* and/or some few particles we say are determinately parts/not parts of T they say are indeterminately parts of T*. If the many vague objects view is correct we can charitably interpret their 18

Some philosophers deny the existence of Tib and indeed tail. For them the example of Tibbles will not convince. The other arguments to be given do not depend upon its acceptance.

100

harold w. noonan

term “T*” as referring to something other than our term “T” and do not have to ascribe mistaken views to them. But if the unique vague object view is correct, their term “T*” is either without reference or refers to a precise object or is coreferential with our term “T,” in which case if we are speaking the truth, they are making false statements about T. But their disagreement with us may be faultless; there need be no cognitive or perceptual failing from which they, or we, suffer. The same is true of the innumerable other counterfactual linguistic communities we can imagine differing from each other and from us as we differ from the first counterfactual community just imagined. Only one of these communities can be right about the boundaries of T if the unique vague object view is correct. So if this view is correct it is overwhelmingly likely that we are just wrong about the boundaries of T (according to this view there is just one vague object in the vicinity, which is the determinate reference of our term “T,” an object about the boundaries of which there is a matter of fact, but this view does not require that our opinions about these boundaries are correct). But T can be any vague object which is a thinker and according to the vague object view persons are typically such vague objects, so it implausibly implies that all the time we are very probably making incorrect ascriptions of (indeterminate) temporal and spatial boundaries to persons. Consider now the second kind of vagueness noted above − vagueness whether a person existing at one time, Brown, is the person existing at another time, Brownson. In this case it is indeterminate whether there is (exactly) one person in place P1 at time t1 and at place P2 at time t2, though it is determinately true that there is exactly one person in P1 at t1 and determinately true that there is exactly one person in P2 at t2. This is similar to the way in which, in Shoemaker’s case of Alpha Hall and Beta Hall, it is indeterminate whether there is (exactly) one building in which Smith, who is located in Alpha Hall, and Jones, who is located in Beta Hall, are both lecturing, though it is definitely true that Smith is located in just one building and Jones is located in just one building. On any account according to which there are multiple person candidates present (one existing both at t1 and at t2, one existing at t1 and not existing or only indeterminately existing at t2 and one existing at t2 and not existing or only indeterminately existing at t1) it is not true that there is only one thinker existing at t1 (t2) unless far-off future or past differences between a pair of temporarily coincident entities can disqualify one of the pair from being a current thinker. However, because of the temporal symmetry of the situation (comparable to the spatial symmetry of the Alpha Hall/Beta Hall example) one vague object cannot suffice to explain the indeterminacy. It is definitely true that a person exists at t1 and definitely true that a person exists at t2. So to postulate a single vague object which is

Personal identity and its perplexities

101

indeterminately a person is inconsistent with the description given of the situation as one in which it is definitely true that a person exists at t1 and definitely true that a person exists at t2. And if there is a single vague object which is determinately a person but which is such that it is indeterminate whether it exists at the earlier time and indeterminate whether it exists at the later time it is false that, as specified, it is definitely true that a person exists at the earlier time and definitely true that a person exists at the later time. So to accept this kind of indeterminacy in personal identity the complex theorist must accept that in this kind of case there are many thinkers present at t1 and at t2 or else allow that far-off future or past differences between some pair of coincident entities can disqualify one of any pair from being a current thinker. If, as I have argued above, this is implausible, the complex theorist, who has to allow this kind of indeterminacy in personal identity, must accept that any such case involves a multiplicity of thinkers.

part ii

Arguments for and against simplicity

chapter 5

How to determine which is the true theory of personal identity Richard Swinburne

the problem The simple view of diachronic personal identity holds that personal identity is not constituted by continuities of mental or physical properties or of the physical stuff (that is, the bodily matter) of which they are made, but is a separate feature of the world from any of the former, although of course it is compatible with personal identity being caused by such continuities. On the simple view, as I shall understand it, a person P2 at t2 can be the same person as a person P1, at an earlier time t1, whatever the physical or mental properties and whatever the body possessed by each person. P2 may not at t2 remember1 anything done or experienced by P1 at t1 or earlier, and may have an entirely different character and a totally different body (including brain) from P1 at t1. The main arguments in favor of the simple view consist in adducing thought experiments in which persons undergo radical changes of mental life and bodily constitution, and in which – it is claimed – it is “possible” that they continue to exist; from which it follows that continuities of the kind mentioned are not necessary for personal identity. I begin with one example (of very many which have been put forward) to indicate the role of thought experiments in supporting the simple view. Suppose I have a severe brain disease affecting the right brain hemisphere. The only way to keep my body functioning is to replace this hemisphere. So the doctors remove my current right hemisphere and replace it by a right hemisphere taken from a clone of me. The new right hemisphere, let us suppose, contains the brain correlates of (that is the neurons, the states of which are the immediate causes of ) similar but slightly different memories and character traits from mine. The resulting person would presumably to 1

Ordinary language sometimes assumes that only true beliefs are correctly called “memories.” Thus it assumes that if I am correctly said to have a “memory” that I did so-and-so, I really did so-and-so. I shall not follow that usage here, but shall understand by “memory” what on that usage would only be an apparent memory: it seeming to the subject that he “remembered” so-and-so.

105

106

richard swinburne

some extent behave like me and remember having done what I did and also to some extent behave like my clone and remember having done what my clone did (when what I did was different from what my clone did). Now suppose that the disease spreads to the left hemisphere, and that too – two years later – is replaced by a left hemisphere taken from a different clone of me, again containing the brain correlates of similar but slightly different memories and character from mine. Then my body would be directed by a brain made of totally different matter and sustaining rather different memories and character from those I had two years previously. Yet presumably to some extent, but to a lesser extent than after the first operation, the resulting person would still behave like me and remember having done what I did. But would the resulting person be me? That person would be a person largely continuous with the earlier me two years ago, apart from having had two large brain operations. One might think that the continuity of many mental and physical properties over this period has the consequence that the same person continues to exist. Yet the resulting person would have none of the brain matter and only some of the memories and character which were previously mine. I suggest that it is totally unobvious whether in this situation the resulting person would or would not be me. Yet the question “Would the resulting person be me?” is logically equivalent to the question “Would I survive the operations?” and so have the (pleasurable or unpleasant) experiences of the resulting person. And surely no one about to have a serious operation can think that the question of whether he will “survive” a brain operation is simply a question requiring an arbitrary decision about which of two senses we should give to the word “survive.” We (or at least most of us) seem to understand the alternatives as mutually incompatible factual alternatives – that I survive, or that I do not survive – in one clear and natural sense of “survive.” Yet it is totally unobvious what is the answer. If you think that – one way or other – the answer is obvious, it is easy to alter the thought experiment in such a way that the answer is no longer obvious. If you think it is obvious that the continuity of at least half the brain matter over each of the operations two years apart insures that I continue to exist, suppose the second operation to be performed after only one year or six months. If you think it obvious that when half my brain matter is removed in one operation I no longer exist, suppose a series of operations replacing only a tenth of the matter each time. In such a situation, which I call an ambiguous situation, it does seem possible that I have survived (i.e. continued to exist), and possible that I have not survived; and yet that we do not know (and have no further means of finding out) whether I have or have not survived. If what seems possible is

The true theory of personal identity

107

indeed possible, my survival does not require any particular degree of strong physical and mental continuities2 which make it obvious that I do survive. It then follows that the difference between the situations of different degrees of continuity consists in the strength of the evidence that I continue to exist. Under normal conditions of very strong continuity of body (and in particular of the brain, the physical sustainer of mental life), memory (of what happened to a person with that body) and character, it is enormously probable that I continue to exist; it becomes less and less probable until we reach the ambiguous situation where it is as probable as not that I continue to exist. Why it is enormously probable that under those normal conditions I continue to exist is first because it is a fundamental epistemological principle that (apparent) memory beliefs are probably true (in the absence of counterevidence), and my personal memories (that is memories “from the inside” about what I did and experienced) concern the actions and experiences of the person who had a brain strongly continuous with my present brain. Unless memories as such (in the absence of counter-evidence) are probably true (and so do not require to be rendered probable by evidence of some other kind in order to be probably true), we would know very little about the world. For we depend on memory for all the knowledge which we believe that we have acquired from others about history and geography, etc.; and while my memory of these things may coincide with yours, at any time I depend on my own memory of what others have told me for my belief that our memories do coincide, and so the personal memory of each of us must as such be probably true if we are to have virtually any knowledge at all. And the second reason why it is enormously probable that under those normal conditions I continue to exist is that the simplest, and so most probable, hypothesis supported by the strong continuity of memory and character sustained by the same brain is that these are mental properties belonging to the same person. It would be less simple, and so less probably true, to suppose that the memory and character strongly continuous with the previous memory and character sustained by a brain having strong continuity with the previous brain are the memory and character of a different person. So being the same person does not entail strong continuity of brain, character and memory; but the latter is good evidence of the former. This is the simple view. 2

By brain, memory or character being “strongly continuous” with a previous brain, memory or character, I understand that there exist between them both what Derek Parfit (1984, ch. 10) calls strong “connectedness” (that is strong similarity) and what he calls strong “continuity” (that is overlapping chains of strong connectedness), the continuity of memory and character being causally sustained by the strong continuity of the brain.

108

richard swinburne

Some philosophers hold that personal identity, like the identity of artifacts, can be a matter of degree. On this view a later person can be only partly identical to some earlier person, and so the result of such operations as I have described might be that the resultant person was only partly me. I do not myself think that it is logically possible that some person be partly me. But even if this were a possible result of the operations, it does not seem to be a necessary truth that the operations will have this result, because the history of all the physical bits and all the mental properties associated with them seems compatible with the subsequent person not being partly me. It still seems possible that, just as the resulting person is fully me if I have both heart and liver replaced, so after the half-brain transplants the resulting person is still fully me; and it is also possible that it is not me at all. Yet if we include the subsequent person being partly me as a possible result of the operations, we would now be ignorant about which out of three (rather than two) possible results of the operations had in fact occurred. If what “seems possible” is possible, that I survive the operations not merely in part but wholly (or alternatively not at all), partial survival is compatible with the simple view. The alternative to the simple view, the complex view, claims that personal identity is constituted (not merely evidenced) by a certain particular degree of continuity of physical and mental properties and of the matter which forms a person’s body. The main arguments in favor of this view are that the paradigm examples of personal identity are all ones in which there is continuity of these kinds, and that the simple view leads to contradictions. There are innumerable varieties of the complex view according to which degrees of continuity ensure the identity of a later person with an earlier person. One variety is the view that the concept of personal identity has no application outside normal situations of strong physical and mental continuities. Another variety of the complex view holds that necessarily (not merely possibly, as in the version of the simple view) personal identity is a matter of degree. The weaker the continuities, the lesser the degree to which the later person is the same as the earlier person. Again there is an issue about this variety of the complex view, as about the similar variety of the simple view, as to whether the notion of partial identity of persons makes any sense. logical possibility Which of these views is correct depends on what is possible, and so I come to a central topic of this chapter: what is it for some situation to be “possible,” and how can we know whether it is or not? We can have no discussable

The true theory of personal identity

109

knowledge of possibilities (or necessities or impossibilities) which cannot be expressed in sentences, and so I will discuss the question of which situations are “possible” (or whatever) by discussing which sentences describing situations describe “possible” (or whatever) situations, and I shall call such sentences “possible” (or whatever) sentences. In talking about possibility in this kind of context, we are talking about the widest or weakest kind of possibility there can be, which I shall follow most contemporary philosophers in calling “metaphysical possibility.” Some state of affairs may not be practically or physically possible, but it may still be metaphysically possible. Hence metaphysical necessity and metaphysical impossibility are the narrowest or strongest kinds of necessity or impossibility. Among metaphysical possibilities, etc., are ones discoverable a priori: that is, discoverable by reflection on what is involved in the claim made by the sentence. I shall call these logical possibilities, etc. No sentence could be more strongly impossible than a selfcontradictory sentence. It claims both that something is so and also that it is not so (and is normally expressed by a sentence of the form “s and not-s”). For such a sentence could only be true if that something is so, and the sentence asserts that it is not so. But any sentence which entails a self-contradiction would be as strongly impossible as a self-contradiction. For similar reasons no sentence could be more strongly necessary than one whose negation entails a self-contradiction. Such necessities and impossibilities are logical ones, since they are discoverable by deriving the relevant contradiction by means of the rules of the language. I see no reason to suppose that there are any other a priori impossibilities as strong as those which entail a contradiction, or any other a priori necessities as strong as those whose negation entails a contradiction. To begin with the case of impossibilities: what is asserted could only be a priori impossible if the impossibility is detectable merely by understanding what is involved in what is actually said. To be impossible a sentence must have the form of a declarative sentence in which the component words already have a sense in the language. It will be a subject−predicate sentence, an existential generalization, or some other one of many recognized forms of declarative sentences. It will – to put the point loosely – assert something about some substance or property or event or whatever that it has or does not have some property or relation to some other substance, property, etc.; or that there are or are not certain substances, properties or whatever. Words have a sense insofar as it is clear what are the criteria for an object, property or whatever picked out by the word being that object, property or whatever. They therefore delimit a boundary to the sort of object or property it can be or the sort of properties

110

richard swinburne

it can have. Hence it will be inconsistent to affirm that an object picked out by some word is of a kind ruled out by the very criteria for being that object. And the form of any sentence will exclude some alternative; and so it will be inconsistent to affirm the sentence together with that alternative. If a sentence is not inconsistent in these ways (or does not entail one that is), we would have no reason to suppose that that sentence is a priori impossible. A similar argument shows that we would have no reason to believe that any sentence whose negation is not inconsistent in these ways is a priori necessary. So I shall assume that all logically impossible sentences entail a contradiction, and all logically necessary sentences are such that their negations entail a contradiction;3 and so assume that all declarative sentences which do not entail a contradiction are logically possible. But what determines the rules of a language, and so the truth conditions of sentences, and so what entails what? Sentences of a language mean what most of its speakers (or some group of expert speakers) mean by them. Each of us learns the meanings of certain sentences by being shown many paradigm observable conditions under which those sentences are regarded as true or false, and by being told of other sentences to which a speaker is regarded as committed by uttering those sentences, and other sentences which are such that someone who utters them is regarded as committed to the former sentences. We learn the meaning of a word by being taught the difference to the meaning of a sentence made by that word playing a certain role in the sentence. By being taught the meanings of individual words and of sentences of various forms, we may then come to an understanding of the meaning of a sentence in which those words are arranged in a certain way, even if we have not been shown observable conditions under which the sentence is regarded as true or as false. Showing a language learner “observable conditions” may involve pointing to them or describing them in terms already introduced. We need to observe many different paradigm examples of observable conditions under which a sentence containing a certain word in various roles is regarded as true or false, and of the commitments speakers who use sentences containing that word in various roles are regarded as having; and this allows us to acquire an understanding of the conditions under which some new sentence containing that word would be regarded as true or false. We extrapolate, that is, from a stock of supposedly paradigm examples (of observable conditions and relations of commitment) to an 3

Robert Adams (1987, pp. 213–14) argues (in effect) that there are logical necessities whose negations do not entail a contradiction. In Swinburne (2010, pp. 318–19) I argue that the example of a sentence by which he seeks to show this does not support his view.

The true theory of personal identity

111

understanding that the sentence would be regarded as true (or false, as the case may be) under conditions sufficiently similar in certain respects to most of the paradigm examples. This process normally leads, within limits of vagueness and minor idiosyncrasies of use, to words (and longer expressions) and sentence forms, and so sentences having a “correct” use. It leads, that is, to public agreement about what in general are the circumstances in which a given sentence would be true and the circumstances under which it would be false; and so to the commitments of sentences to other sentences. We may call a rule for what one is objectively committed to by a sentence a rule of mini-entailment. s1 mini-entails s2 if and only if anyone who asserts s1 is thereby (in virtue of the rules for the correct use of language) committed to s2. s1 entails sn if they can be joined by a chain of mini-entailments, such that s1 mini-entails some s2, s2 mini-entails some s3, and so on until we reach a sentence which mini-entails sn. Given this agreement, we are then in theory in a position to determine the logical necessity, possibility or impossibility of sentences. To show some sentence s to be logically impossible we need to find an agreed chain of mini-entailments from s to a contradiction; and to show s to be logically necessary we need to find such a chain from not-s to a contradiction. Getting agreement that such a chain has been found is, however, often a difficult matter. An opponent of the claim that s entails a contradiction may challenge some suggested link in the chain − say the suggestion that p mini-entails q, by claiming that q is not something to which anyone is committed when using p in the correct sense. This disagreement may be overcome if the proponent of the claim that s entails a contradiction can get his opponent to recognize some r such that p mini-entails r and r mini-entails q. Or the disagreement may be bypassed if the proponent can find a different chain of mini-entailments from s to a contradiction which an opponent will recognize as such. A sentence is logically possible if it does not entail a contradiction. Of course any logically necessary sentence is logically possible. But to show some other sentence to be logically possible (and so logically contingent) may be an even more difficult matter than to show a sentence to be logically impossible or necessary. Sometimes it is very obvious that some sentence does not entail a contradiction, and so is logically possible. A true sentence entails no contradiction, and if it is obvious that some sentence (e.g. “my desk is brown”) is true, then it is obvious that it is logically possible. Sometimes too it is very obvious that some sentence, which may be false, entails no contradiction (e.g. “my desk is red”). And more generally it is sometimes very obvious that some description of a world very different from

112

richard swinburne

our world entails no contradiction. To show some disputed sentence s to be logically possible requires showing that it is entailed by a sentence r which disputants agree to be logically possible without this needing to be shown by argument. The argument then consists in showing that r entails s by a sequence of agreed mini-entailments. For if r does not entail a contradiction, neither does any sentence entailed by r. For example, someone may try to show that “there are two spaces” − a space being a system of places each of which is at some distance in some direction from each other place of the system and from no other places – is logically possible, by describing in detail a situation under which it would be true.4 That is, they claim that the latter description entails that there are two spaces; and that since the latter description is logically possible, “there are two spaces” is also logically possible. However, the use of these procedures to determine logical possibility presupposes that it is clear what are the truth and falsity conditions of sentences, and which sentences mini-entail other sentences. But the language-learning process, which normally produces very similar understandings of meanings in members of a language group, sometimes produces somewhat different understandings of these conditions and entailments in different sub-groups. This may occur because different learners learn meanings from somewhat different paradigm examples; and when this is recognized, language users can acknowledge that the same word or sentence has more than one meaning. But it may also occur even when both sub-groups acknowledge the same paradigm examples of observable conditions and commitments. And then it sometimes happens that one of two sub-groups objects that the sense in which some word (or longer expression) derived from the same paradigm examples (of observable conditions and mini-entailments) used by the other sub-group is not a real or legitimate sense of that word, in that its use in that sense entails contradictions. Or one of the sub-groups may object that the sense in which the word (or longer expression) is used by the other sub-group is not the sense implicit in some of the paradigm examples. It is objections of these two kinds that produce the disputes about the meaning of “personal identity.” Most of us have been taught the meaning of the expression “is the same person as the person who” or its more natural equivalent “is the person who” by many observable paradigm examples of the same kind (e.g. “this is the same person as the person you saw last week,” “you are the person who had a headache only thirty seconds ago”) and many similar paradigm 4

For an argument of this form in favor of two spaces being a logical possibility, see Quinton (1962).

The true theory of personal identity

113

mini-entailments (e.g. “A is the same person as B” and “B is the same person as C” mini-entail “A is the same person as C”). We will all recognize most of these observable paradigm examples of “same persons” as examples of persons with strongly continuous bodies, memories and character. So some philosophers provide an analysis of the meaning of “P2 is the same person as P1” in terms of P2 having a body, memory and character strongly continuous with those of P1. That is, they advocate the complex theory of personal identity as a conceptual truth, in my sense a logically necessary truth. But others of us (including myself) think that it is not the normal sense of “P2 is the same person as P1”; and that in the normal sense this expression designates a continuing identity of a different kind which normally underlies the strong physical and mental continuities but is not constituted by them and can occur without them. That is, we claim, the simple theory of personal identity is a conceptual truth, in my sense a logically necessary truth. The only way to resolve this disagreement is by persistent continuing use of the methods described earlier. Advocates of the complex theory as a conceptual truth try to get us to recognize the logical impossibility of a personal identity independent of strong mental and physical continuities. They do this by trying to show that some sentence using the expression “is the same person as” in a different sense from theirs entails a contradiction which would not arise if the expression were used in their sense. For example, they may claim that “Socrates is the same person as the mayor of Queenborough, but has none of the same brain, memory or character as the mayor,” together with what they may claim to be a necessary truth “no one should be punished for any act which they cannot remember doing” entails “both {the mayor should be punished for any immoral acts of Socrates} and not-{the mayor should be punished for any immoral acts of Socrates}.”5 If they can get us to recognize this in one case, then they may get us to recognize that other sentences where the expression “is the same person as” is used in a sense other than their sense will have the same consequences, and so to see that any such sense of the expression is not a legitimate one. We who claim that the simple theory is a conceptual truth are of course likely to deny, with respect to the example just discussed, either that the first conjunct of the purported entailment (“the mayor should be punished for any immoral acts of Socrates”) is indeed entailed, or that “no one should be punished for any act which they cannot remember doing” is a necessary truth. 5

This example is of course a formalized version of Locke’s argument for the necessity of same memory (which he calls “same consciousness”) for personal identity (Locke 1975 [1690], ii, xxvii, 19).

114

richard swinburne

We advocates of the second sense are happy to acknowledge that the sense in which our opponents understand “is the same person as” − that is, as something like “has a body, memory and character strongly continuous with” − is a perfectly legitimate sense of an expression, but claim that it is not the normal sense of “is the same person as.” This is because – contrary to the claims of the advocates of the complex theory – it is not compatible with the sense implicit in some of the paradigm examples of personal identity by which the expression has been introduced into language. Some of these examples concern our opponents’ own identity. They must recognize that they themselves often have streams of overlapping experiences. For example, the second half of the experience of a pain during a “specious present” may overlap with the first half of an experience of some noise; this noise may continue for a short while (during several overlapping periods of specious present), and overlap with a certain tactual experience, and so on. It is a paradigm example of personal identity that two overlapping conscious events are experiences of the same person, from which it follows that any stream of such events are also experiences of the same person.6 Then they must recognize some very recent past experiences which they remember so vividly that it is obvious that they occurred (e.g. “you are the person who had a headache thirty seconds ago”); and it is obvious that – as Reid (2002 [1785], iii.4) put it – “my memory testifies not only that [a certain past action] was done, but that it was done by me who now remembers it.” We thus point out to any opponent that some of the paradigm examples of personal identity that he must recognize are ones in which he has a direct awareness of personal identity; and what the awareness is of is not continuity of body, memory and character, but something which can only be described as an awareness of himself as a continuing subject of experience. Once we have focused on the paradigm examples of our opponent’s own personal identity over time, which give rise to the understanding of himself as a continuing subject of experience, we must get our opponent to recognize that, as with any experiences, what he is aware of (the continuity of his mental life) could occur without his subsequently remembering it. And so, more generally, we must get him to see that this continuity could occur without any of his criteria of personal identity being in any way satisfied. To do this we need to describe some situation in much detail by a sentence (normally consisting of a long conjunction, such as a thought experiment in which someone is described as surviving events of a kind 6

It was John Foster who drew attention to the phenomenon of a stream of overlapping experiences as the foundation of our understanding of personal identity. (See Foster 1979, 176: “it is in the identity of a stream that we primarily discern the identity of a subject.” See also Foster 1991, 246–50.)

The true theory of personal identity

115

described at the beginning of this chapter) which our opponents are prepared to recognize as logically possible; and then find a route of mini-entailments from it to a sentence that claims that some later person is the same person as an earlier person without our opponents’ criteria being satisfied. Given that this is logically possible, our opponents’ theory which claims that it is not logically impossible must be false. My feeling about how this debate goes these days is that we are getting our message across. If it is admitted that it is physically possible, and so a fortiori logically possible, that a series of operations such as those described at the beginning of the chapter could occur and that it is logically possible that the person before the operations could subsequently have the experiences of the person after the operations, it then follows that being the same person as a previous person does not entail having the same brain or strongly continuous memory and character. Our opponents may, however, insist on some residual physical continuity, e.g. that the replacement of brain matter does not occur all at once. But someone softened up by physically possible stories of the kind described at the beginning of the chapter may then begin to acknowledge the logical possibility of a person acquiring a new body all at once without gradual replacement of parts; and so come to acknowledge that it is logically possible that a person could be the same person as a person at a later time without there being any continuity of body (including brain), memory or character between them. So, our opponent should recognize the second sense as the normal sense of personal identity. The same arguments that will show that there is no contradiction in an unnamed person continuing to exist and have experiences under these circumstances, are unaffected by whom one supposes the person to be. So we may conclude that it is logically possible for me or any other human to survive total replacement of body, memory and character. Logical possibility is the kind of metaphysical possibility which can shown to be such a priori. But to determine whether it is metaphysically possible for me or any other human to continue to exist without any continuity of body, memory or character, we need to show that this is also a posteriori metaphysically possible. That, someone may claim, will depend on what sort of persons we humans are − that is, what is the essence of a human person − and that that is not something to be determined a priori. a posteriori metaphysical possibility I will begin my discussion of a posteriori modal claims in terms of what it is for a sentence to be a posteriori metaphysically necessary; the application to

116

richard swinburne

a posteriori metaphysical impossibility and possibility will then become apparent. A posteriori metaphysical necessity is supposed to be a necessity as hard as logical necessity, yet discernible only a posteriori. No different type of necessity could be as hard as logical necessity, and so a posteriori necessity must be in some way reducible to logical necessity. And, as far as I can see from all the plausible examples of a posteriori metaphysically necessary sentences that have been adduced, the way “a posteriori” comes into it is that we need to make empirical enquiry to determine more adequately what is the substance or property or whatever about which the claim being made by the sentence is being made. When we have determined that, if the claim is metaphysically necessary, the necessity of the claim will be detectable a priori. Sentences pick out the substances, properties or whatever with which they are concerned either by “rigid designators” (as defined by Kripke 1981, p. 48) − that is, expressions which (given that their meaning remains the same) always refer to the same substance, property or whatever, however different the world might be from how it is − or by non-rigid designators which may pick out different substances or whatever if the world is different. “Green,” for example, is a rigid designator because it always refers to the color green, whereas “Amanda’s favorite color” is a non-rigid designator because it would refer to a different color if Amanda had different color preferences from her present ones. I will call a rigid designator φ an “informative designator” if we can (when favorably positioned, faculties in working order, and not subject to illusion) recognize when something is φ and when it is not, merely in virtue of knowing what the word φ means. I would not understand the word “green” unless (when the stated conditions are satisfied) I could recognize when an object is green and when it is not. When our referring expressions are informative designators (or can be defined in terms of informative designators), we know the necessary and sufficient conditions for the things referred to to be what they are; and so, I shall say, we know the essence of what is being designated. When all the designators in a sentence are informative (or can be defined in terms of informative designators), it is a pure a priori exercise to determine whether the sentence is logically necessary or whatever. When we know what we are talking about, mere thought can show what that involves. Mere thought, for example, I suggest, can show that “all trilateral figures are triangular” (a “trilateral” figure is a closed surface bounded by three straight lines; a triangle is a closed surface bounded by straight lines and having three interior angles) or “no surface can be both green and red all over” are logically necessary; and we can now see that this is because “surface,” “straight,” “green,” “red,” etc. are all informative designators, and so

The true theory of personal identity

117

understanding the sentences involves knowing the essences of what is being referred to and so comprehending fully what the sentences are claiming. But there are rigid designators which refer to substances or whatever, such that a speaker can understand to what they are referring on some occasions when the thing exhibits certain non-essential features, without knowing the essence of what is referred to. Clearly one will not understand what some rigid designator means (what its role is in the language) unless one knows how to use it on some occasions, and understands the kind of thing to which it is used to refer and so the kind of criteria by which to distinguish one thing of that kind from other things of that kind; but language users may not be able in practice to use these criteria to determine to which thing of that kind it is referring. But in that case one would not be able to recognize that thing on occasions other than the ones on which it exhibits those non-essential features (its “stereotype”). I shall call such designators for which the criteria of “informative designators” are not satisfied (and cannot be defined in terms of such designators), “uninformative designators.”7 A sentence is then a posteriori metaphysically necessary if it would be logically necessary when we substitute informative designators (or expressions definable in terms of these) for its uninformative designators. Thus – to use the example discussed by Putnam (1975) – the word “water” as used in the eighteenth century was an “uninformative designator.” This is because although people used “water” as a designator of a stuff, and so knew that to be the same stuff something would have to have the same chemical essence, they picked out a volume of stuff as water in virtue of its superficial contingent properties (being liquid, in our rivers and seas, etc.), yet – in ignorance of what that chemical essence was − they would not be able to recognize it on occasions when it did not have those superficial contingent properties. So they were unable to say whether or not sometimes stuff found elsewhere than in our rivers and seas was water or not. When people discovered that chemical essence ( H2O), they could then recognize whether stuff not in our rivers and seas was water or not. Hence, since the claim being made about water is a claim about H2O, we can substitute “H2O” for “water” in “water is H2O” (as used in the eighteenth century) and the sentence then reduces to a logically necessary truth. (I assume here that “H2O” can be defined in terms of informative designators: that is, in 7

Similar distinctions to my distinction between “informative” and “uninformative” designators are those made by Chalmers (1996) between expressions with “primary intensions” and ones with “secondary intensions,” and by Bealer (1996) between “semantically stable” and “semantically unstable” expressions.

118

richard swinburne

terms of such expressions as “mass,” “volume,” “smaller by 10–1 than” and so on.) Somewhat similar is the sentence used by Kripke (1981, p. 100) to illustrate a posteriori metaphysical necessity, “Everest is Gaurisanker,” where – Kripke supposes – “Everest” was used by early explorers to designate a mountain having a certain shape when seen in the distance from Tibet and “Gaurisanker” was used to designate a mountain having a certain shape when seen in the distance from Nepal. The explorers understood the sentence “Everest is Gaurisanker” because they understood what it would be like for the two mountains to be the same – it would consist in their being made of the same chunk of rock. The rock of which each mountain was made constituted its essence. But they did not know whether the two mountains were the same (whether they had the same essence), and it required empirical investigation to discover that they did. Once they knew what each mountain essentially was, they knew that the claim being made by the sentence was necessarily true with a necessity as hard as that of “all squares have four sides.” Thus – in my terminology – they could know that there is an informative designator of the form “mountain made of such and such rock” which can be substituted for both “Everest” and “Gaurisanker,” so that the sentence has the form of an identity sentence “a is a” and so is logically necessary. The application of my account of a posteriori metaphysical necessity to a posteriori metaphysical impossibility and possibility should now be evident. The metaphysical modal status of a sentence is its logical status when informative designators are substituted for uninformative ones. Even if we cannot find out what is the essence of some substance or whatever, our understanding of how to use the designator may give us enough knowledge of the kind of essence involved to enable us to see or deduce the modal status of some sentence using it. For example, even if we do not know the essence of water, we can see that “water is the number 42” is impossible. But only if all designators in a sentence are informative (or can be analyzed in terms of informative designators) is it guaranteed that mere a priori reasoning can determine its logical status. Many of the words by which we pick out properties are informative designators (“green,” “square,” “has a length of one meter,” etc.). And many words by which we pick out properties, which are not themselves informative designators, can be analyzed in terms of them – e.g. “has a length of 10–18 meters” can be defined in terms of the informatively designated property “has a length of one meter” and eighteen applications of the informatively designated relation of “being shorter by one-tenth than.” However it seems that there are at present some substances which we can only pick out by

The true theory of personal identity

119

uninformative designators. For example, it is an unresolved issue (French 2006) whether some fundamental particles, such as quarks and electrons, are the particles they are merely in virtue of their properties (such as mass and charge, and causal relations to other particles), or whether they are what they are partly in virtue of the particular matter of which they are made. For this reason we do not know what are the necessary and sufficient conditions for some fundamental particles to be the ones they are (that is, we do not know their essence) and have to pick them out by uninformative designators. Now what sort of designator is “I,” or “Richard Swinburne,” as used by me? These seem to be informative designators. If I know how to use these words, I cannot be mistaken about whether or not they apply to a certain person – given that I am favorably positioned (e.g. his body is my body), with faculties in working order, and not subject to illusion. And when I am considering applying these words to a person in virtue of his having some conscious event, these conditions will be maximally satisfied and no mistake is possible. I am, in Shoemaker’s (1994, p. 82) phrase, “immune to error though misidentification.” I cannot know how to use the word “I,” recognize that someone is having some conscious event (e.g. a pain) and still wonder whether it is I who am having that event or someone else, in the way that an early explorer could know how to use the word “Everest,” and yet wonder whether the mountain at which he is looking from Nepal is Everest. My knowledge of how to use “I,” like my knowledge of how to use “green” and “square,” means that, in the sense analyzed above, I know the essence of what I am talking about when I use the words. Hence if “I will exist tomorrow with a new brain” or “I will exist without any memory of my previous existence” are logically possible, they are also metaphysically possible. I claim therefore that the considerations which should lead people to conclude that such sentences express logical possibilities should also lead them to conclude that they express metaphysical possibilities. And since I can know this merely in virtue of knowing to what my use of the word “I” refers, other people can know the same about themselves. Each of us, we may properly conclude, can continue to exist without any continuity of brain, memory or character. Of course I can still misremember what I did in the past, and indeed misremember how I used the word “I” in the past. But this kind of problem arises with every claim whatsoever about the past. “Green” is an informative designator of a property, but I may still misremember which things were green and even what I meant by “green” in the past. The difference between informative and uninformative designators is that (when my faculties are in working order, I am favorably positioned and not subject to illusion) I can

120

richard swinburne

recognize which objects are correctly picked out at the present time by informative designators, but not necessarily when they are picked out by uninformative designators (in the absence of further information). And so I know what a claim made now about the past or future amounts to when it is made by informative designators (but not when it is made by uninformative designators) whether or not I have any reason for supposing it to be true. For to claim that some informatively designated object a will exist or have an informatively designated property φ tomorrow is just to claim that something which I can understand (a existing or being φ today) will hold tomorrow. It follows that I can understand what it is for me to exist tomorrow or yesterday and to have such and such experiences. Not so with Everest or water, when these things can be picked out only by uninformative designators. I do not know what would constitute a past or future substance being water or Everest if I am merely in the position of the “water” user in the eighteenth century, or the early explorers using “Everest” in the way described. I conclude that given that each of us can come to see that it is logically possible that they can survive without any continuity of brain, body or character, in the crucial sense in which they subsequently have the experiences of the surviving persons, and so come to see that this is metaphysically possible and so come to see that the simple theory of personal identity is true. the human soul I stated the simple view as the view that personal identity is not constituted by continuities of mental or physical properties or of the physical stuff (that is, matter) of which persons are made, but is a separate feature of the world from any of the former. But this leaves open the possibility that personal identity might be constituted by a non-physical part, a “soul.” Substance dualism holds that each human on earth consists of two parts, a body and a soul – the soul being the essential part, and the body a contingent part. On this theory, while any physical stuff of which the body is composed, and any particular physical and mental properties, are not necessary for the continued existence of a person, the continued existence of his soul is necessary. I suggest that this further step is forced upon us if we admit the logical possibility of a certain thought experiment, and the high plausibility of a certain principle about what it is for any substance to continue to exist. The thought experiment is this. Suppose that there exists instead of our actual world W1, a world W2 which is exactly the same as W1 except in that instead of a certain person S1 who lives a certain life in W1 there is a person S2 who has the same body and the same mental and physical properties

The true theory of personal identity

121

throughout his life as S1, but is not S1. And surely our world could be different in the sole respect that the person who lived my life was not me. For it is not entailed by the full description of the world in its physical aspects and in respect of which mental properties are associated with which bodies that the person who has any particular body and mental and physical properties should be me. We can see this if we imagine that before this world exists we are shown a film of what is going to happen in it; and that the film has some device for showing us what will be the mental lives of the people in the world. Each of us would still not know whether we were going to live one of the lives in this world. And if so, which one. So because W2 can be seen, I suggest, when we reflect on it, to be logically possible; and because – as before – persons can be picked out by themselves by the informative designator “I,” W2 is a metaphysically possible world. But a substance at a particular time is the substance it is in virtue of its parts, what they are made of and their properties, and the relations the parts have to each other. For example, if there is a substance composed of certain fundamental particles with certain properties (including certain relations to past particles) related to each other by certain causal and spatio-temporal relations, there could not be instead of it a different substance composed of all the same particles with the same properties and relations to each other. Andre Gallois (1998, p. 251) has called the view that there could be another such substance “strong haecceitism.” He writes: Strong haecceitism seems to me incredible. Consider a car on a parking lot. It is not at all incredible to suppose that a qualitative duplicative of the car in question might have existed even if there is no qualitative difference at any place or time as a result. It is incredible to suppose that throughout history all of the atoms that actually exist might have been configurated at each time in exactly the way they are actually configurated without the car on the parking lot existing.

I suggest that it follows from our very understanding of what a substance is that what Gallois describes as “incredible” is false; and in particular it follows from our understanding of what a person is that two persons could not be different if all their parts and all their properties were the same. (This does not commit me to the identity of indiscernibles, which holds that two substances are the same if they have all the same properties (in the sense of universals). It allows that two substances may be different even if they have all the same properties, so long as their parts are different.) It follows that in the thought experiment described above, S1 and S2 must have different parts; and since all their physical parts are the same, the difference must arise from each having different non-physical parts, that is

122

richard swinburne

different souls. My earlier thought experiments suggested that a person can continue to exist with a different body and different mental properties. The thought experiment just described in this section suggests that a person needs a particular soul in order to continue to exist. It is, however, compatible with substance dualism to hold that to exist a person needs a body, but not any particular body – though I believe that further thought experiments show that a person can exist without a body and so without any physical properties at all. So the only essential properties necessary for a person to exist are the essential properties of any soul, which – I suggest – are simply the one property of having (in some sense) a capacity to be conscious.8 8

In my book The Evolution of the Soul (Swinburne 1997, pp. 153–4, and pp. 327–32), I followed St. Bonaventure in analyzing a soul as a form (a collection of essential properties) instantiated in some mental stuff, soul-stuff. But it now seems to me that any stuff must be capable of being divided into smaller chunks of the same stuff; and given my view that humans (and so their souls) cannot be divided, the soul cannot be made of any stuff. It is an “immaterial particular.”

chapter 6

Against simplicity Sydney Shoemaker

i Proponents of simple views of personal identity often contrast personal identity with the identity of other sorts of things. Sometimes, as in the case of Bishop Butler and Thomas Reid, this takes the form of a view which takes identity of persons to be identity in the “strict and philosophical sense” and construes the identity we ascribe to other sorts of things, especially artifacts, as identity in only a “loose and popular sense,” something which, as Reid put it, we call identity “for conveniency of speech.” Whether or not the identity ascribed to other sorts of things is held to be “fictitious,” not really identity, what is thought to distinguish it from personal identity is its being “analyzable,” its consisting in facts closely related to the facts that serve as our evidence for ascribing it. There are held to be “constitutive criteria” for the identity of ships and trees, and it is denied that there are such criteria for the identity of persons. The tendency to regard personal identity as special in this way often – but not always – goes with acceptance of some form of mind–body dualism. But if the rejection of constitutive criteria for the identity of persons is the mark of the simple view about personal identity, then on one understanding of constitutive criteria there are plenty of philosophers who hold a simple view about the identity of ships, trees, rivers, etc. For there are many that reject “reductionism” about the identity of such things. There are, of course, different understandings of what it is for an account of the identity of so-and-sos to be “reductive,” and what it is for the identity of so-and-sos to be “analyzable.” But it is worth asking whether there are any grounds for denying that the identity of persons is not analyzable that are not equally grounds for denying that the identity of other sorts of things is not analyzable. I will discuss this later. But first I want to present a case in favor of a complex view about the identity of things other than persons – I will focus on the identity of trees – and then argue that the case applies equally to persons. 123

124

sydney shoemaker ii

When a thing persists over an interval of time there is a series of events and property instances that constitutes the career of the thing during that interval. I will speak of this simply as a series of property instances, since the events in the series can be thought of as causally connected series of property instances. But the causal connections in the series are of course not limited to those holding among property instances that make up individual events; it will be true generally that what property instances occur later in the series will be causally due, in part, to what property instances occurred in it earlier. A special case of this will be the retention of a property over time; my wedding ring’s having a certain shape at a time is due in large part to its having had that shape earlier. But there are also cases in which there is systematic change in which later states of a thing grow out of earlier states of it and are causally and counterfactually dependent on them – as when a tree gradually changes its shape and size in playing out its biological nature. What a career is really a series of are not individual property instances but sets of simultaneous property instances. These sets will consist of instances belonging to a single thing – instances standing to one another in a synchronic unity relation. These can be called “thing-stages.” If the career is the career of a tree, each property instance will be a state of affairs consisting of a tree’s having a certain property at a time, and each tree-stage will be a state of affairs consisting of a single tree having a certain multiplicity of properties at a time. The question of whether a series of tree-stages is the career of a tree is the question of whether it is one and the same tree that is involved in these different states of affairs. There is of course a question as to what the synchronic unity of a thing-stage consists in – e.g. what it is for two simultaneous tree-states to be states of one and the same tree. But my concern here is with diachronic unity. Perhaps some proponents of “perdurance” views of persistence over time would identify persisting things with their careers. But most of us would not. And one does not have to hold that view in order to hold that the existence of a career entails the existence of a thing whose career it is. That is a claim on which proponents of the simple view and proponents of the complex view can agree. Where they cannot agree is on what it takes for a series of thing-stages to be a career. On a complex view, what makes a series of thing-stages a career is the relation between the properties instantiated in it at different times, including the nature of the causal relations between different property instantiations in the series. These make the series a career; and in so doing they constitute the

Against simplicity

125

persistence of the object over the interval occupied by the career. This has the consequence that if a series of thing-stages duplicates the series of property instances that makes up the career of some thing of kind K, in the sense that in the duplicate series the same properties are instantiated in the same temporal order and the property instances are causally related in exactly the way their counterparts are in the career, then the duplicate series is itself a career of something of kind K, and there is a thing of kind K to which the property instances in the series belong.1 Notice that such an account of the persistence of things through time makes free use of the concepts of the kinds of things said to persist. For its thing-stages are states of affairs involving such things – e.g. a tree-stage is a state of affairs consisting in a tree having certain properties at a time. One can say if one likes that tree-stages and tree-property instances “get their identity” from the trees they involve. But it would be a confusion to say that this makes the account of the identity of trees across time circular. The notion of a tree enters into the characterization of the series of tree-stages, but the notion of sameness of trees over time does not – nowhere in the characterization is it said that it is the same tree that is involved in the different tree-stages. A similar charge of circularity will be addressed later on. A proponent of a simple view about the trans-identity of things of a certain kind can certainly agree that in general the occurrence of a series of thingstages that duplicates in this way a series that makes up a career of a thing of that kind will itself make up the career of a thing of that kind. But can he hold that the existence of such a series constitutes the career of a thing of that kind? If he does, he will be holding that the existence of such a series constitutes the persistence over time of a thing of that kind. But that is what he is supposed to be denying in holding a simple view about the persistence of things of that kind. It would seem that he must hold that it can at best be only a contingent truth, not a necessary one, that duplicating the series that is the career of a persisting thing of a kind gives us the career of a thing of that kind. Of course, if we are allowed to include in the specification of the series that makes up a career that the different thing-stages in the series stand to one another in an unanalyzable relation of diachronic unity, a relation that 1

This is true only with a qualification that does not affect the point I am making. To use a well-known example of Donald Davidson’s, we can suppose that a highly unlikely consequence of lightning striking a swamp might be the coming into being of a creature, a “swamp man,” that is a physical duplicate of an actual man. We could also have swamp tigers and swamp trees. On some views of species individuation these swamp creatures would not belong to the biological kinds their real world physical duplicates belong to – although arguably they would persist through time in the way their duplicates do. We can allow for this by modifying the claim in the text so that it concerns cases in which the duplicate series occurs in a setting, including a pre-history, like that of the series it duplicates.

126

sydney shoemaker

necessarily holds just in case the relata belong to the very same thing, then the advocate of the simple view can happily allow that duplicating the series making up a career will itself make up a career. On the simple view, this relation of diachronic unity will have to be something over and above the causal relations that hold among the property instances. And apparently it will be a relation that could be there whether or not the causal relations hold. The proponent of the simple view is apparently committed to holding that a series that duplicates the career of a persisting thing with respect to what property instances it contains and how these are ordered and causally related could fail to be the career of a persisting thing of that sort, namely through the members failing to stand in the unanalyzable diachronic unity relation to one another. And he also seems committed to holding that a series of property instances whose members are causally unrelated could nevertheless be the career of a persisting object, because its members stand to one another in the unanalyzable diachronic unity relation despite their lack of causal connectedness. These are, to put it mildly, highly counterintuitive consequences. According to the first, what we can establish empirically about a series of thing-stages is insufficient to establish that it is the career of a persisting object. I observe over a period of time what I take to be a tree growing in my yard, and what accounts for what I observe is the impact on my sensory system of a series of instantiations of tree properties whose members stand in precisely the sorts of causal relations that obtain among the property instances in the career of a growing tree.2 But on the simple view, apparently, it is compatible with what I seem to observe in such a case that the series of property instances is not the career of a tree. What more could I observe, or otherwise establish, that would show me that there really was a single tree persisting throughout that interval? It is hard to see how on this view we could have any justified beliefs about the persistence of trees through time. According to the second consequence of the simple view, there could be a series of tree-property instantiations that constitute the career of a tree even though there are no causal relations among its members. So there are no relations of counterfactual dependence among its members – it is not true of our tree that if a branch had been lopped off yesterday it would have had fewer branches today. Of course, its having the branches it has today should 2

I am not saying that what I observe is a series of tree-stages. What my perceptual states represent is a tree persisting over time, not a succession of tree-stages. But my having perceptual states that represent this is due to my sensory system registering the impact on it of there occurring, within my field of view, a succession of tree-stages, i.e. a succession of tree-property instances.

Against simplicity

127

have a cause. But it would seem that on the simple view it is not required that this cause should lie at least in part in its past career. Moreover, if the diachronic unity relation among property instances cannot consist partly in causal relations, it is hard to see how it could consist partly in spatio-temporal relations. And then it is not clear why a series that we would take to be the sum of parts of the careers of different trees – a part of a different tree career for each day of the week – could not be the career of a single tree. It may be suggested that these embarrassing consequences vanish when we see that the properties instantiated at different times in a career include sortal properties, like the property of being an oak tree. The instantiation of such a property at a time and place, it will be suggested, imposes constraints on what can be true at temporally and spatially adjacent times and places. A tree cannot exist for just an instant; if it exists at a time then it must exist throughout an interval that includes that time. The existence of a tree at a place at a time requires its existence at nearby places at nearby times. And different trees cannot be at the same place at the same time. So if there is a temporally and spatially continuous series of instantiations of the property of being a tree, there must be a single tree that is the subject of all of these instantiations. Temporally proximate instantiations of the property must belong to the same tree, since otherwise we would have different trees occupying the same place at the same time. And by the transitivity of identity, this implies that it is the same tree throughout. This will be true, the suggestion is, whether or not the successive property instances are causally related in the way required by the complex view. And we can know it to be true without knowing that they are so causally related. This may seem prima facie plausible. But it is not clear how the simple theorist can be in a position to say it. Crucial to the argument is the claim that, as I have put it vaguely, the existence of a tree at a place and time requires its existence at nearby places at nearby times. This implies that the persistence of trees requires spatio-temporal continuity. (I think that it is arguable that the requirement of spatio-temporal continuity stems from the fact that persistence requires causal relations and that causal chains must be spatiotemporally continuous.3 But I need not rely on this here.) Considerations about causality aside, it is unclear how a proponent of a simple view about the identity of a certain kind of objects can hold that the careers of such objects must be spatio-temporally continuous. For one cannot hold this without holding that the persistence of such things partly consists in spatio-temporal 3

If it is allowed that there can be causal paths that are not spatio-temporally continuous, it is a plausible consequence that there can be careers that are not spatio-temporally continuous.

128

sydney shoemaker

continuity – and that seems to be something that a proponent of the simple view cannot hold. We should also notice that while it is true that the property instances that make up the career of a persisting thing must include instances of sortal properties, the instantiation of sortal properties supervenes on the instantiation of other properties. It is because of other properties a thing has that it counts as being a tree. The property of being a tree will be realized in physical properties, and, ultimately, in microphysical states of affairs that realize instances of these physical properties. And it will be the causal profiles of these more basic physical properties and microphysical states of affairs that make them realizers of the property of being a tree. While the instantiation of the property of being a tree cannot be instantaneous, the instantiations of the physical properties in virtue of which something is a tree can be. It will be series of instantaneous physical property instances that realize instances of the sortal property of being a tree. So at one level of description, the series of property instances that constitutes the career of the tree will be the series of the physical realizers of the property of being a tree.4 They are realizers of that property because their causal profiles include its causal profile. The reason why the instantiation of the property of being a tree at a time constrains what is true at other times is that this property is partly individuated by causal features that contribute to determining what is true at subsequent times. When this property is instantiated one or another of its realizer properties will be instantiated, and its causal features will affect what is true at subsequent times. The effects will include the instantiation of properties that are themselves realizers of the property of being a tree. So the physical properties of a tree at a time, properties that realize the property of being a tree, will be such as to generate, or contribute to the generation of, slightly different physical properties at a later time, and these will also realize the property of being a tree. And because of the way the latter properties grow out of the former, the two sets of properties will belong to the same tree. The point that the existence of trees cannot be instantaneous rests on the fact that trees necessarily have careers that exhibit these kinds of causal connections. So it can hardly be used against the claim that the persistence of trees consists in the existence of such careers.

4

If realizers must be sufficient for what they realize, then an instantaneous property instance cannot be a realizer of a non-instantaneous property instance. What can be a realizer of the latter is a series of instantaneous property instances. In such a case let us speak of the instantaneous property instances in the series as quasi-realizers of the non-instantaneous one. So what I should say here is that what constitutes the career of the tree is the series of quasi-realizers of the property of being a tree.

Against simplicity

129

Further, if we focus on the microphysical states of affairs that are the ultimate realizers of tree property instances, we see that those that realize the property instances in the career of a tree form a series that can be described without any use of the concept of a tree. There is a good sense in which this series constitutes the career of the tree, and so constitutes the persistence of the tree over the interval in question. If this series is described in physical terms, it will not be a priori that a series of this description constitutes the career of a tree. But that it does so will be an a posteriori necessary truth – one that holds partly in virtue of the empirical fact that the causal features of the microphysical states of affairs are such as to make them realizers of the tree-property instances in the career, and partly because of the fact – which I think will be a priori – that these, and the relations among them, are collectively such as to constitute the career of a tree. This gives us a set of facts whose specification does not require the use of the concept of a tree but which constitute a fact about tree identity and tree persistence. That seems enough to show that tree identity consists in something, contrary to what a simple view about it would claim. I have claimed that on the simple view there could be a series of property instances that is the career of a persisting object despite there being no causal relations, and no relations of counterfactual dependence, between its successive members. It might be replied that all that the simple theorist is committed to is that there being such relations cannot be, or be part of, a sufficient condition for a series of events being the career of a persisting thing, and that this does not prevent him from holding that it is a necessary condition for this. The same might be said of the requirement that the career of a persisting thing must be spatio-temporally continuous. But it is not clear how these could be necessary conditions without being parts of a sufficient condition. If we deny that a series of property instances is the career of a persisting object of a certain sort on the grounds that it is not spatio-temporally continuous, or that its members are causally unrelated, we are saying that it does not have what it takes to be a career of an object of that sort. And that surely implies that there is something that constitutes the persistence of objects of that sort, something that is lacking if there is no spatio-temporal continuity or no causal connectedness. The diachronic unity relation cannot be simple and unanalyzable if its obtaining requires spatio-temporal continuity or causal connectedness. iii Most simple theorists about identity over time are not simple theorists about the identity of trees – or of rivers, automobiles or statues. So they

130

sydney shoemaker

could accept what I have said against the simple view about tree identity. Their claim concerns the identity of persons. But it would seem that what I have said about tree identity should carry over to the identity of persons. Consider the career of a person, and consider a series of person-states (person-property instances) that duplicates this career with respect to what properties are instantiated and in what order, and with respect to how the instantiations are related causally. The denial that this duplicate series would have to be itself the career of a person is, to say the least, extremely implausible. It seems to lead to an extreme sort of skepticism about third-person identity judgments and, indeed, about judgments about other minds. It is hard to see how, on this view, the phenomena we take to be evidence of the persistence of a person over an interval of time could really be so. And it is also hard to see how, on this view, the phenomena we take to establish the existence of mental states in other persons could really do so – for the behavioral evidence for the existence of mental states consists in phenomena occurring over a period of time and depends on successive stages of the behavior being manifested by one and the same person. I think, therefore, that the simple view is just as implausible in the case of persons as it is in the case of trees, rivers and the like. What partly underlies the feeling that there is a difference here is the memory access persons have to their own pasts. First-person memory judgments are not themselves identity judgments, but in making them one commits oneself to the truth of trans-temporal identity judgments about oneself. If I say that I remember having eggs for breakfast this morning, I commit myself to being the same as the person I remember eating eggs this morning. And that identity judgment appears to be one I know to be true without the use of any criterion of identity. My knowledge of it seems to be direct and not grounded on evidence of identity of any sort. We do not have this sort of knowledge of the truth of trans-temporal identity judgments about other things (including persons other than ourselves). And this can make it seem that where we do base identity judgments on evidence, the status of this evidence is different when the judgment concerns the identity of a person than when it concerns the identity of a ship or a tree or a river. In the former case the evidence can have the status of being a symptom of identity – something that is known to be evidence of an identity fact because it has been found empirically, using one’s memory access to one’s past, to be correlated with facts of that sort. The symptoms would include things like similarity of appearance, similarity of fingerprints and similarity of DNA – things that nearly everyone would regard as symptoms rather than criteria. But they would also include, on this view, spatio-temporal continuity, continuity

Against simplicity

131

with respect to properties of one sort or another (psychological properties, biological properties, etc.), and causal connections between successive states. The discovery of the correlation of course would involve the access each of us has in memory to facts about his or her own identity, an access that gives us knowledge of identity that is direct and not grounded on evidence. The use of this access to ground symptom-based judgments about the identity of persons other than ourselves would of course have to involve something like the argument from analogy – a projection onto others of correlations we have discovered to hold in our own case. Where our identity judgment concerns a ship or a tree or a river, we must have evidence concerning such things that does not consist of symptoms in this sense. Sometimes, to be sure, our judgments about such things would be based on symptoms, but the discovery of the correlations that tell us that these are symptoms would have to involve evidence that does not consist of symptoms – it cannot be symptoms all the way down. And it may seem that this non-symptomatic evidence must constitute criteria of identity – and that these criteria must be constitutive, and not merely epistemic. Thus reflection on the special access we have to our own pasts in memory, and on the absence of a similar access to the pasts of other sorts of things, can lead to the view that personal identity is alone in being simple – where its being simple is equated with there being no constitutive criteria for it. But the fact that we have this special memory access to facts about our identity does not at all imply that there are not constitutive criteria of personal identity. We do not use any criteria of identity in making memory-based first-person identity judgments, so of course we do not use as a criterion the fact that the memory involved stands in certain causal relations to the past action or experience that it represents. But this does not mean that this fact cannot be a constitutive criterion of personal identity that can be used as an epistemic criterion by persons other than the subject of the memory. The series of property instances that make up the career of a person will include episodes of making memory judgments, where these are made spontaneously and on the basis of no evidence, and in most cases these will correspond in content to, and will be appropriately causally connected to, events earlier in the career. On a psychological view of personal identity this can be one of the facts about such a series that makes it the career of a person; and it is compatible with this that first-person memory knowledge of personal identity is not grounded on this – or any other – criterion of personal identity. It is also compatible with non-psychological accounts of personal identity, e.g. animalism, that no criterion, psychological or otherwise, is used in making memory judgments about one’s own past. It can be held that it is

132

sydney shoemaker

just a contingent fact about human beings that they have in memory a direct access to facts about their own pasts. More generally, it is compatible with the complex view, the view that there are constitutive criteria of personal identity, that persons are capable of having direct knowledge of their own pasts that is not grounded on such criteria. But it is often urged against complex accounts of personal identity that they are inescapably circular. This is especially urged against psychological accounts. It is an old objection to memory accounts of personal identity that the notion of memory implicitly involves the notion of personal identity, it being a necessary condition of a state’s being a memory of a past event that its subject be identical with a past subject or witness of that event. In previous work I have countered this objection by arguing that the work done by the notion of memory in psychological accounts can be done by the notion of “quasi-memory,” which is like the notion of memory but does not include the requirement that there be an identity between the person who quasiremembers an event and a past subject or witness of that event – also by saying that much more than memory continuity goes into the psychological continuity that constitutes personal identity (Shoemaker 1970). Since my present purpose is not to argue for a psychological account but to argue that some complex view must be true, I will not pursue this matter further here. But E. J. Lowe has raised an objection to the complex view of personal identity that, while explicitly directed against the psychological version of the complex view, would if cogent go against other versions as well. He says that any criterion of personal identity on Lockean lines “cannot be deemed informative in the sense in which any adequate criterion of identity is required to be so. For conscious states – and, indeed, mental or psychological states quite generally – cannot themselves be individuated or identified save in terms which presuppose the identity of persons (or conscious subjects) whose states they are” (Lowe 2009, p. 134). Now there certainly is a good sense in which the individuation of conscious states involves the identity of persons. A conscious state of a person is necessarily a state of that person, so any situation in which it exists will be one in which that same person is its subject. But it is not only conscious states, or psychological states, whose individuation involves the identity of persons in this sense. Any state, psychological or not, necessarily belongs to the thing of which it is a subject, and so gets its identity from the identity of its subject. And it is not only states of persons of which this is true. So if this sort of involvement of the identity of persons makes a psychological account of personal identity circular, any account of the identity of any sort of persisting thing in terms of relations between successive states would be circular. If this

Against simplicity

133

were right, Lowe’s argument against the complex view of personal identity would generalize into an argument against the complex view of any sort of persisting thing – which would be contrary to his intent, since he thinks that in the case of entities other than persons there can be non-circular constitutive criteria of identity. But as I noted earlier, in talking about the identity of trees, it is a confusion to think that this sort of involvement of identity in the individuation of states, property instances, thing-stages, etc. makes circular accounts that hold the trans-temporal identity of things to consist in relations between such items. For what is involved in such individuation is neither the diachronic identity nor the synchronic identity that is at issue in discussions of personal identity (or in other discussions of the identity conditions of the possessors of property instances and states). In such discussions what will be at issue is what it is for different states, either occurring at the same time or occurring at different times, to belong to one and the same person. Each of the states will necessarily belong to a person, and what is at issue is what it is for it to be the same person all of them belong to. That the identity of each of the states is fixed by the identity of the person who has it is in no way incompatible with the claim that it is relations between the states (no doubt involving relations to yet other states) that make it the case that they belong to one and the same person. It is not altogether clear why it is natural to use the term “identity” to make the point that states, property instances, etc. necessarily have the subjects they have. Perhaps the identity here is identity across possible situations – across “possible worlds.” If a state in a given situation is the same as one in another, what has it in the one situation is the same as what has it in the other. This of course has no bearing on questions of diachronic or synchronic identity. Lowe thinks, mistakenly, that accounts that frame criteria of personal identity in terms of person-stages assume a four-dimensionalist view of persisting things, a view he repudiates.5 But there is no need to equate person-stages, as he does, with “time-slices” of persons – one can think of a person-stage as just the total state of a person at a time. A state of a thing is not a part of it, and the fact that a series of person-stages has temporal parts does not imply that persons have temporal parts. Putting aside his rejection of fourdimensionalism, Lowe says that he can see no prospect of our being able to introduce the notion of a “person-stage” in a way that would make the individuation and identification of person-stages possible without reference to the persons whose stages they would supposedly be. In 5

Lowe (2009, p. 137, fn 58) identifies me as “one leading advocate of the four-dimensionalist approach.” In fact, I have always been a staunch opponent of this approach.

134

sydney shoemaker

consequence, I can see no prospect of our being able to formulate an informative, because non-circular, criterion of personal identity framed in terms of personstages. (Lowe 2009, p. 137)

This is the same mistake as before. Person-stages are indeed individuated by reference to the persons of which they are stages. But that leaves it open whether different person-stages are stages of one and the same person, and no circularity is involved in holding that whether they are is determined by relations, including causal ones, between them. It is such an account that I have offered.

iv But there is a version of the circularity objection that has more substance than the ones I have considered. And it is an objection that applies as much to complex accounts of the trans-temporal identity of other sorts of things as it does to complex accounts of personal identity. This is an objection I raised in a paper written over thirty years ago (Shoemaker 1979). I did not think then, and do not think now, that it shows that there cannot be informative accounts of what the trans-temporal identity of various sorts of things consists in. What I did claim is that it precludes a “reductive” account of the identity of such things. It remains to be considered what that claim amounts to. The argument rests on the claim that there is an internal relation between the persistence conditions for things of a given kind and the causal profiles of properties characteristic of things of that kind. Properties are individuated in part by the effects their instantiation has on the future career of the things that have them. This is most evident in the case of dispositional properties like fragility, malleability and elasticity. For something to be malleable is for it to be such that, when subjected to certain forces, it – that same thing – acquires a different shape. For something to be elastic is for it to be such that, when subjected to certain forces, it – that same thing – will change in shape, and that when that force is removed, it – that same thing – will revert to its former shape. But it is also true of the properties that ground such dispositions that their causal profiles include tendencies to affect the future careers of the things that have them. Here the nature of the properties is explained in part in terms of the trans-temporal identity of the things that have them. But on a “complex view” of the identity of such things, the identity of things over time consists in relations, e.g. relations of causal dependence, between instances of these properties occurring at different times in their careers. This seems to pose a threat of circularity; the trans-temporal identity of things is explained

Against simplicity

135

in terms of relations between instantiations of properties, yet this transtemporal identity figures in the explanation of the nature of these properties. In fact, I think, what we have here is not circularity but a kind of holism. The nature of the properties and the nature of the persistence conditions of the things that have them cannot be explained independently of one another. It is compatible with this that we can give an informative account of what the persistence of a given sort of thing consists in. The account will be grounded in an account of what sorts of properties things of a given sort have and what properties are essential to them. Take, for example, the current debate between psychological (neo-Lockean) accounts of personal identity and “animalist” accounts that hold that personal identity consists in the sort of biological continuity that characterizes the career of a persisting organism. These are competing complex views. In differing about what personal identity consists in they are at the same time differing about what is involved in the causal profiles of the properties persons can have. To put it very roughly, on the psychological view the causal profiles of mental properties are such that the successor states their instances contribute to generating will be states of the same person, even if the successor states of the biological properties possessed by that person earlier belong to someone other than the possessor of those mental states, while on the animalist view the causal profiles of biological properties are such that the successor states their instances contribute to generating will be states of the same person, even if the successor states of the mental states possessed by that person earlier belong to someone other than the possessor of those biological states. These views give different accounts of what personal identity consists in, each of which if true would be an informative account of it. It should be noted that in most cases our concept of a property is far from including a full specification of its causal profile. So the internal relation here is not between the concept of the trans-temporal identity of a kind of thing and the concepts of the causal profiles of the properties things of that sort can have – it is between the nature of the trans-temporal identity and the nature of the causal profiles of the properties. This shows, I think, that there being this internal relation does not make accounts of trans-temporal identity in terms of continuous series of property instances circular. It is true that in some cases our concepts of properties do include quite a bit of information about their causal profiles. On a functionalist view this is the case with many of our concepts of psychological properties. To the extent that this is so there may be a kind of circularity involved in psychological accounts of personal identity. But it is a benign circularity, reflecting a conceptual holism. It may be compared with circularities we run into in

136

sydney shoemaker

specifying the causal roles of mental states, e.g. that characterizing the causal role of a belief requires mention of relations to desires, while characterizing the causal role of a desire requires mention of relations to beliefs. What such circularities point to is the need for a “package deal” account, and that is what we need here. We might take reductionism about trans-temporal identity to be the view that the persistence of the things through time consists in relations between entities that can exist independently of their involvement in the careers of persisting things. Whether the interdependence of trans-temporal identity and the causal profiles of properties stands in the way of such a reductionist account depends on whether properties have their causal features essentially. If they have them only contingently, instances of them could exist without being involved in the careers of persisting things. Even this would not mean that the actual instances of them could have existed without being so involved; only in worlds governed by different laws could the properties have different causal profiles, and only there could instances of them occur without being involved in careers in the way instances of them are here. In any case, my own view is that properties have their causal profiles essentially. So I reject reductionism on this interpretation. What I have been arguing is that the rejection of reductionism is compatible with the complex view, taken as the view that there are informative criteria for the trans-temporal identity of things, including persons.

v One of the central claims of this chapter is that such difficulties as there are in giving an informative and non-circular account of trans-temporal identity are difficulties that arise no matter what sort of persisting thing we are dealing with. None of them is such as to make the complex view about personal identity problematic in a way complex views about the identity of trees, rivers, etc. are not. As I suggested earlier, a major source of the view that personal identity is unique in being simple is the special access we have in memory to our own identities. But, as I argued, this is not a good reason for the view – complex views of personal identity can easily accommodate this special access. Advocates of the simple view have often been dualists. But such reasons as dualists may have for holding the simple view are not reasons for the rest of us. And I think there are no good reasons for the rest of us.

chapter 7

The probable simplicity of personal identity E. J. Lowe

My aim in this chapter is to show that personal identity is, in all probability, “simple” rather than “complex,” in the sense that there is no informative and non-circular criterion of personal identity. This will require me to explain what I understand a criterion of identity, quite generally, to be. Some philosophers, of course, are skeptical about the very notion of a criterion of identity – but that, I think, is simply because they misunderstand the notion, which is really perfectly straightforward and clearly has application. Others would want me to be more specific in my characterization of “simplicity,” insisting that what is strictly at issue is the existence or non-existence of an informative and non-circular criterion of diachronic personal identity. However, in my view, this also is to misunderstand the notion of a criterion of identity. Where persisting things of a certain kind K are concerned, an adequate criterion of identity for Ks should certainly provide an account of their identity across time, but can do so only in conjunction with an account of their identity at a time. It is an error to suppose that “diachronic” and “synchronic” identity are different sorts of identity and so demand different identity criteria. Identity itself is univocal and unanalyzable. It applies to any kind of entity whatever, whether it persists through time or not. My procedure will be as follows. First of all, I shall try to motivate the quest for a criterion of personal identity, partly in order to subvert the attacks of those philosophers who are skeptical about the very notion of an identity criterion. Then I shall say something about identity itself, from a logical point of view. Following that, I shall explain what I take a criterion of identity to be, and how, in my view, identity criteria are properly formulated. Subsequently, I shall focus on the specific case of persons, concentrating my discussion on Locke’s conception of personal identity and modern developments of that approach – often called the “psychological” approach. I shall argue that this approach offers little prospect for a satisfactory “complex” account of personal identity. Then, more briefly, I shall discuss other approaches which favor a “complex” account, but dismiss 137

138

e. j. lowe

them too, leaving a “simple” account as the most plausible option. I shall also try to explain why we should not be surprised by this finding. Finally, I shall address an important objection to my principal reason for rejecting a complex account.

why should we seek a criterion of personal identity? Persons, I take it, are prime examples of minded beings. However, it might be supposed that, since the notion of identity is a universal one, there can be nothing special to say about personal identity as such, beyond saying that it involves the application of this notion to minded beings of a certain kind. Some philosophers would undoubtedly agree with this view. They would urge that the theory of identity is exhausted by an account of the logical properties of the identity relation, which reduce to the fact that it is a reflexive relation governed by Leibniz’s law: that is, by the principle of the indiscernibility of identicals. This is just the principle, taken to be a necessary truth, that things that are identical share all their properties, or that whatever is true of something is true of anything identical with that thing. If that view were correct, then there would be nothing to be said about personal identity beyond the banality that persons, like anything else, can be said to be identical only if they are indiscernible from one another. On this view, there is nothing more to be said regarding the hypothesis that I am identical with, say, Julius Caesar than that it is true only if I differ from Julius Caesar in no discernible way. It might be supposed that this is then the end of the matter, since it is surely just obvious that there are indeed discernible differences between Julius Caesar and me, such as that he conquered Gaul but I did not. But why should anyone be so confident that I did not conquer Gaul? The reply may be offered: because I obviously did not even exist at the time of that conquest. But why should anyone be so confident of that? It can only be because something is being presupposed about the nature of persons which constrains the possibilities of identifying one person with “another,” such as that I cannot be identical with a person none of whose experiences I can remember having, or with a person whose body was destroyed before the body that I have now was created. It is presuppositions like these that make it seem “obvious” that I cannot be identical with Julius Caesar: but they have nothing to do with Leibniz’s law as such, since they relate specifically to the presumed nature of persons, as opposed to things of various other kinds.

The probable simplicity of personal identity

139

This suggests, then, that much more needs to be said about personal identity than can be captured simply by applying the logical properties of the identity relation to the particular case of persons. Specifically, what seems to be called for is a principled account of the identity conditions of persons – or, to use John Locke’s helpful phrase, an account of what their identity “consists in.” In modern parlance, what we might hope to establish is a criterion of identity for persons. However, before we can examine any particular proposal concerning the identity conditions of persons, we need to look more closely at identity and identity criteria considered quite generally. Furthermore, we should bear in mind that we cannot just assume in advance that our quest for a criterion of personal identity is bound to be successful, if pursued hard enough. Indeed, as I indicated at the outset of this chapter, one of my main aims in what follows is to show that such a quest is very probably doomed to failure, while at the same time urging that this is no reason to regard the concept of a person with suspicion. identity from a logical point of view Now, clearly, the expression “is identical with” – symbolized in logic and mathematics by the equality sign, = – is a relational expression and hence may naturally be supposed to denote a certain relation in which things can stand to one another. However, if so, then it is a very peculiar relation, in that it can never literally hold between one thing and another thing, but only between a thing and itself. Other relations can, of course, hold between a thing and itself, such as the relation of loving: someone can obviously love him- or herself. But this relation can also hold between different things, as when Peter loves Jane. Identity is peculiar as a relation in that it necessarily holds only between a thing and itself, and, indeed, this has led some philosophers to deny that it is “really” a relation at all. However we classify it, though, it can certainly seem strange. Since, of necessity, everything is identical with itself and with no other thing, one might wonder how facts of identity can fail to be utterly trivial and uninteresting. Part of the solution to this puzzle is provided by distinguishing, as we must do in any case, between identity and identification. Identification, in the sense of the term that I now have in mind, is a cognitive act and a far from trivial or easy one. One and the same object may often be identified in different ways, even by the same thinker, and it may not be evident to such a thinker that, indeed, he or she has identified the same object in two such ways. To be able to identify an object is, typically, to be in possession of some descriptive information which applies uniquely to that object. But, as

140

e. j. lowe

Frege (1960 [1892]) pointed out, a thinker can be in possession of two such pieces of information without necessarily thereby knowing that they apply to the same object. To cite his famous example, it was an astronomical discovery of considerable magnitude that the Evening Star (Hesperus) is the Morning Star (Phosphorus). Similarly, it would be a remarkable discovery to find out that the conqueror of Gaul (Julius Caesar) is the author of this chapter (E. J. Lowe). One function of identity criteria is to impose certain constraints on what can count as an intelligible act of identification. Before we discuss criteria of identity, however, we need to say more about identity itself. Identity, as I have already remarked, is a reflexive relation – a relation which, of necessity, holds between every thing and itself. We can formalize this as follows: ð∀x Þðx ¼ x Þ: As I also remarked earlier, identity is subject to Leibniz’s law, which for our purposes may be formalized in the following way: ð∀x Þð ∀y Þðx ¼ y → ð∀F ÞðFx ↔ FyÞÞ: Here F stands for any condition that may hold true of an object, so that the above formula effectively affirms that, for any things x and y, if x is identical with y, then anything true of x is also true of y, and vice versa. From the foregoing two principles, it is easy to derive two other logical properties of the identity relation – its symmetry and its transitivity, expressible by the following two formulas: ð∀x Þð ∀y Þðx ¼ y → y ¼ xÞ: ð∀x Þð ∀y Þð∀z Þððx ¼ y &y ¼ z Þ → x ¼ zÞ: Together, these four formulas exhaust the properties of the identity relation, from a purely logical point of view. They pin that relation down uniquely, as being not only an equivalence relation – reflexive, symmetrical and transitive – but also, more specifically, as being the only such relation all of whose equivalence classes are necessarily singlemembered, with each such member being an ordered pair of a thing and itself, of the form 〈x, x〉. To make this latter point clearer, observe that each equivalence class of the same height relation is the class of all those pairs of objects that share a certain height and, clearly, while it might happen to be the case that only one object has a certain height, it is also possible for more

The probable simplicity of personal identity

141

than one object to have the same height. Hence, some of these equivalence classes may contain many ordered pairs of different objects, such as 〈Peter, Jane〉, 〈Jane, Mary〉 and 〈Peter, Mary〉, assuming that Peter, Jane and Mary all have the same height. But the equivalence classes of the identity relation are all perfectly uniform, each having a unique member such as 〈Peter, Peter〉 or 〈Jane, Jane〉 – because, obviously enough, Peter is identical only with Peter, Jane only with Jane, and so on. These rather austere logical points are not made idly here, since they have a direct bearing on what can qualify as a satisfactory criterion of identity for things of a given kind. We may sum up the situation by saying that while an equivalence relation such as the same height relation may be described as being an exact similarity relation, the identity relation is necessarily stricter than that, in that it can fail to hold even between objects which are exactly similar in every qualitative respect.

criteria of identity: their forms and adequacy conditions As I understand it, a criterion of identity is a principle which specifies, in a substantive way, a logically necessary and sufficient condition for the identity of objects of a given sort or kind, K. The qualification “in a substantive way” is needed to exclude principles that are trivial, uninformative or circular. Such a principle may take one or other of two different forms and, depending on which it takes, it may be described as being either a “one-level” or a “two-level” identity criterion (Lowe 1989c, 1997). Onelevel criteria take the following form: ð∀x Þð ∀y ÞððKx & KyÞ→ðx ¼ y ↔ RK xyÞÞ: Here, RK denotes what we may call the criterial relation for objects of kind K. And note that such a relation must, of course, be an equivalence relation – reflexive, symmetrical and transitive – because identity itself is an equivalence relation and RK has to hold between Ks just in case they are identical. The best-known example of such a one-level identity criterion is the axiom of extensionality of set theory, which tells us that if x and y are sets, then x is identical with y if and only if x and y have the same members – so that in this case having the same members is the relevant criterial relation. Observe that this relation is, as required, an equivalence relation. However, Frege – who founded the formal theory of identity criteria – favored two-level identity criteria, which may be written in the form:

142

e. j. lowe ð∀x Þð ∀y Þðf K ðx Þ ¼ f K ð y Þ↔RK xyÞ:

Here, fK denotes what might aptly be called the K-function. The best way to illustrate this is by means of Frege’s own famous example of such an identity criterion, his criterion of identity for directions (Frege 1953 [1884], p. 74). A direction (in the geometrical sense of the word) is always a direction of something, namely, a line. And Frege’s criterion of identity for directions is just this: the direction of line x is identical with the direction of line y if and only if line x and line y are parallel. So, in this case, the K-function is the direction of function and the criterial relation for directions is parallelism between lines. Observe that the relation of parallelism between lines is again, as required, an equivalence relation. It should be easy to see why the two different forms of identity criteria receive their respective names. A two-level criterion specifies the identity conditions of things of a kind K in terms of an equivalence relation holding between things of another kind – thus, in the case of Frege’s criterion, it specifies the identity conditions of directions in terms of an equivalence relation holding between lines. In contrast, a one-level criterion specifies the identity conditions of things of a kind K in terms of an equivalence relation holding between the very things of kind K whose identity is at issue – thus, in the case of the axiom of extensionality, it specifies the identity conditions of sets in terms of an equivalence relation holding between the very sets whose identity is at issue. We shall see that this difference between the two forms of identity criteria is significant in the context of a search for an adequate criterion of personal identity. For a two-level criterion of personal identity will be appropriate only if we can and should think of persons as being entities of a “functional” kind, in the sense that directions are. Something now needs to be said about the requirement that a criterion of identity be substantive and hence non-circular. Clearly, it would be blatantly circular to allow the criterial relation in a one-level criterion of identity for Ks simply to be the relation of identity itself. It is true, but just trivially so, that if x and y are Ks, then x is identical with y if and only if x and y are identical. But sometimes a putative identity criterion can be circular in a less obvious way: for example, the putative identity criterion for sets which states that if x and y are sets, then x is identical with y if and only if x and y include exactly the same sets. It is indeed logically necessary and sufficient for the identity of sets x and y that x and y include exactly the same sets (bearing in mind that every set includes itself), but since what we are seeking is an informative way of specifying the identity conditions of sets, it

The probable simplicity of personal identity

143

is clearly unsatisfactory to do so by appealing to a criterial relation – in this case, the relation of including the same sets – which is itself defined at least partly in terms of sameness (that is, identity) between sets. Another example of such circularity is provided by Donald Davidson’s well-known proposal regarding the identity conditions of events, namely, that events x and y are identical if and only if x and y have the same causes and effects (Davidson 1980 [1969]). For, since he takes causes and effects themselves to be events, this proposal amounts to the circular claim that events x and y are identical if and only if the same events cause both x and y and the same events are caused by both x and y (Lowe 1989a). A criterion of identity for Ks should never appeal to or rely upon, in its formulation of the criterial relation for Ks, sameness (that is, identity) between Ks. Unfortunately, circularity of this kind in a putative identity criterion is not always easy to discern and sometimes needs considerable work to discover. This is a problem that afflicts certain well-known attempts to formulate an adequate criterion of personal identity, as we shall see. One final point should be made about identity criteria in general. This is that they are here understood to be metaphysical principles, not merely epistemic or heuristic ones. Thus, for example, while it is true in the case of human persons that having the same fingerprints provides strong empirical evidence for identity between such persons, it certainly is not true that human personal identity consists in having the same fingerprints – for, quite apart from anything else, a human person can obviously survive the loss of his or her fingerprints (by losing his or her fingers) and indeed can even, in these days of modern medicine, acquire someone else’s fingerprints (as a result of a hand transplant). So it cannot be true, quite generally and of necessity, that human persons x and y are identical if and only if x and y have the same fingerprints. what is a person? Locke very wisely observed that “This being premised to find wherein personal identity consists, we must consider what Person stands for” (1975 [1690], ii, xxvii, 9). We cannot hope to formulate an adequate criterion of identity for objects of a kind K unless we have an adequate idea of what Ks are. But what exactly are we asking when we ask a question of the form “What are Ks?”? The correct answer, I consider, is that we are enquiring into the nature or essence of Ks. As for what the word “essence” means in this context, we again do well to cite Locke, who said that “in the proper original signification” of the word “it denotes the very being of any thing, whereby it

144

e. j. lowe

is, what it is” (1975 [1690], iii, iii, 15). From this we may gather that, at the very least, we do not know what a K is unless we know to what ontological category Ks belong. Unfortunately, in the case of persons this immediately gives rise to a difficulty: namely, that different philosophers over the ages and across cultures have had very different views as to what, in this sense, persons are. Some have held that persons are essentially immaterial substances (“spirits” or “souls”), some that they are “combinations” of such a substance with a material one (a “body”), some that they are purely material substances (such as living animals), some that they are “phases” of such substances (rather as caterpillars and butterflies are different “phases” of the same kind of insect), some that they are non-substances (such as “bundles” of experiences, or “functional roles” that substances can occupy), some that they are not even individual entities of any kind but rather universals of a certain type, some that they are “transcendental” entities which cannot be identified with items of any kind that are located in the world of space and time, and some that they are literally non-entities having a purely “fictional” status. What is the source of this remarkably wide difference of opinion concerning the nature or essence of persons? Perhaps this: the key ingredient in anyone’s conception of a person seems to be the conviction that he or she herself is a person. Thus, possession of the first-person perspective is at the heart of anyone’s conception of a person, whatever else may also be part of it. A person is, first and foremost, something that conceives of itself as thinking, feeling or doing various things (Lowe 1996, ch. 1). Such a conception is one that requires the deployment of the first-person pronoun, “I,” or some expression equivalent to that, for its articulation. But the peculiar feature of this pronoun, from a semantic point of view, is that its competent use apparently does not require of the user any very specific conception of what kind of thing it designates. This is why Descartes (1984 [1641], ii) could famously claim to be certain of the truth of the cogito – I think – and thereby certain of his own existence, while still professing uncertainty as to what he was. In the end, of course, he concludes that he is essentially a thinking thing, a substance whose essence is thought (in the broadest sense of that term) and which excludes any physical property such as shape or mass. Locke is less prescriptive concerning the nature or essence of persons, saying only that a person is “a thinking intelligent Being, that has reason and reflection, and can consider it self as it self, the same thinking thing in different times and places” (1975 [1690], ii, xxvii, 9). This definition of personhood certainly builds in the notion that a person is a self-aware subject of thought and experience, but we should not take Locke to be

The probable simplicity of personal identity

145

implying, by his use of the capitalized word “Being,” that persons are substances, much less that they are essentially immaterial substances. In fact, it appears that Locke held human persons to be, strictly speaking, non-substances, their ontological status being that of highly complex modes (“mode” being Locke’s preferred term for an individualized property, or what metaphysicians nowadays call a “trope”). This is because, while he believed that thoughts and other mental “modes” have to be borne by substances and that these substances are in all probability “spiritual” rather than “material” in nature, he held that an individual human person who has such thoughts cannot be identified with any such substance, since that human person could in principle survive a change with respect to the substance bearing its thoughts at different times in its life. This, of course, is connected with Locke’s own theory of personal identity and his preferred criterion of personal identity, to which we shall return shortly. So the problem is that, while practically everyone might agree that, whatever else a person is, a person is something that is, or at least is capable of being, aware of itself as having thoughts, this formulation apparently leaves it almost entirely open what kind of thing this “something” is. In fact, it even seems to leave open the possibility that there need be no one kind of thing that a person could be. If that is the case, however, then it would appear to be misguided to search for a criterion of personal identity as such, since persons of different kinds could be expected to comply with the identity criteria, whatever they might be, associated with the kinds in question. For example, if it is held that human persons – as opposed, say, to android persons of science fiction lore – are animals of a certain kind and thus that I, as a human person, am identical with such an animal (a biological organism of the species Homo sapiens), then it should be concluded that my identity conditions are just those of one such animal – that I began to exist when it did and will cease to exist when it does. This view – known as animalism – is currently fairly popular amongst metaphysicians (Olson 1997a), perhaps on account of its thoroughly naturalistic flavor and perhaps, too, because it effectively does away with many of the traditional problems of personal identity of the sort that Locke’s account generates. On the other hand, the idea that persons are not really a single kind of thing and thus that things of radically different kinds, with quite different identity conditions, could all qualify as persons is prima facie counterintuitive and even rather alarming in its apparent moral implications. As Locke so aptly put it, “person” is “a Forensick Term appropriating Actions and their Merit” (1975 [1690], ii, xxvii, 26): it is indispensable for our moral

146

e. j. lowe

and legal practices of apportioning praise and blame and offering rewards and punishments. One’s natural presumption is that each person has and should have a moral concern for his or her own future and, more generally, for the futures of all other persons. But if there is no unified conception of what would count as “the future” of a person as such, because persons of different kinds can have quite different identity conditions, it may be hard to see what exactly could be the basis of such a universal moral concern. Indeed, if animalism were true regarding human persons such as you and I, why, after all, should I have any moral concern for your or my future as such, given that the animals that you and I supposedly are have identity conditions which do not entail that those futures are ones in which you or I exist as persons at all? Reflections such as these suggest that it is strongly built into the common-sense conception of a person that all persons are essentially persons, so that my ceasing to be a person would entail my ceasing to exist altogether. Locke’s definition of personhood, whatever its defects, is clearly intended by him to have this consequence and to that extent seems to be more in tune with common sense than a view like animalism is. This, in any case, is a good point at which to look more closely at Locke’s own proposed criterion of personal identity, which has remained highly influential. Even if in no other regard, I certainly want to follow Locke in maintaining that “person” denotes a specific kind of thing, governed by a single criterion of identity if it has such a criterion at all.

locke’s criterion of personal identity According to Locke, [S]ince consciousness always accompanies thinking, and ’tis that, that makes every one to be, what he calls self; and thereby distinguishes himself from all other thinking things, in this alone consists personal Identity, i.e. the sameness of a rational Being. And as far as this consciousness can be extended backwards to any past Action or Thought, so far reaches the Identity of that Person; it is the same self now as it was then; and ’tis by the same self with the present one that now reflects on it, that that Action was done. (1975 [1690], ii, xxvii, 9)

It is a matter of dispute amongst Locke scholars exactly how this passage should be interpreted (Lowe 1995, ch. 5; Lowe 2009, ch. 8), but most commentators take it to be expressing a memory-based criterion of personal identity, on the understanding that the kind of memory that we are here concerned with is what is sometimes called “autobiographical” or

The probable simplicity of personal identity

147

“experiential” memory (for example, remembering seeing a certain film some years ago), as opposed to the mere memory of impersonal facts (such as remembering that the Battle of Waterloo was fought in 1815). Here is one way in which one might attempt to frame Locke’s proposed criterion in the form of a “one-level” identity criterion, as such criteria were formulated earlier: ð∀xÞð∀yÞððx is a person & y is a personÞ→ ðx ¼ y↔ð∀t1 Þð∀t2 Þð∀eÞððx has e at t1 → y remembers e at t2 Þ & ð y has e at t1 →x remembers e at t2 ÞÞÞÞ; where t1 and t2 are any two times at which both x and y exist (with t1 being earlier than t2) and e is a variable ranging over individual conscious experiences, such as a conscious experience of having a particular thought or undertaking a particular action. Locke himself, it should be remarked, explicitly appeals to consciousness rather than to memory when formulating his criterion of personal identity in the passage just quoted above, though his talk of “extending consciousness backwards” is very naturally interpretable as referring to conscious memory. What the foregoing formula says, in plain English, is just this: if x and y are persons, then they are the same person if and only if any conscious experience had by x at any earlier time is remembered by y at any later time, and vice versa (restricting ourselves here to times at which both x and y exist, of course, since no person can experience or remember anything at a time at which he or she does not exist). This criterion entails, obviously, that a person must always remember every conscious experience that he or she ever formerly had. That, however, is extremely implausible. Indeed, its implausibility was fairly soon exploited by Thomas Reid (1975 [1785]) to construct a refutation of Locke’s proposed criterion, by means of his famous “brave officer” example. As Reid points out, we can readily imagine there being an elderly general who remembers saving the regiment’s standard when in battle as a young officer and who, as a young officer, remembered stealing apples as a boy. But it also seems quite conceivable that the elderly general has entirely forgotten the boyhood episode. The nub of the problem, it seems, is that identity is a transitive relation, whereas Locke’s proposed criterial relation for personal identity appears not to be transitive. However, according to many modern neo-Lockean theorists of personal identity, all that we need to do to save Locke’s criterion of personal identity from Reid’s objection is to replace Locke’s proposed criterial relation by the

148

e. j. lowe

so-called ancestral of that relation, which is logically guaranteed to be a transitive relation. The elderly general will certainly stand in this relation to the boy, provided that there is a finite chain of persons, beginning with the elderly general and ending with the boy, such that each person in this chain stands in Locke’s proposed criterial relation to the preceding member of the chain. It might be worried, though, that by thus quantifying over persons in formulating a revised version of Locke’s criterial relation, we are introducing circularity into the revised Lockean criterion of personal identity, since we are now appealing to the very class of entities for which we are attempting to provide an identity criterion. But this is a worry that I shall set aside here, concerning though it is, as deeper problems for neo-Lockeanism are still to be encountered. Now, the fact that the ancestral of Locke’s proposed criterial relation is a transitive relation evidently does not guarantee that it is, as required, an equivalence relation, since to have that status it needs also to be reflexive and symmetrical. That it is reflexive might seem to be relatively uncontroversial, but that it is symmetrical is certainly not, owing to the seeming possibility of fission cases, in which two distinct persons, A and B, existing at a time t2 both stand in this relation to a single person, C, existing at an earlier time t1. Of course, it may be objected that the imagined cases of personal fission (and likewise those of fusion) are purely fictional and not really possible. But that is too big a debate to be entered into here. Suffice it to say that such cases present at least a prima facie problem for the revised Lockean criterion. Another prima facie problem – first raised by Joseph Butler (1975 [1736]) – leads to the accusation that Locke’s criterion, both in its original and in its revised form, is implicitly circular. The point here is that the criterion appeals to the notion of a person, P, remembering some past experience, e. However, it is a logically necessary condition of P’s genuinely remembering e (in the first-personal, autobiographical sense of “remembering”) that P him- or herself should actually have experienced e. That being so, then – as Butler urged – memory presupposes personal identity and hence cannot be what constitutes it. The standard modern response to this objection is to concede it, but then to revise the Lockean criterion still further by appealing instead to the notion of “quasi-memory,” where this is understood to be a mental state with all the features of autobiographical memory except that it is not a defining condition of the state that one can “quasi-remember” only experiences that one had oneself (see, for example, Parfit 1984, ch. 11). It must be acknowledged, however, that the notion of quasi-memory is far from being uncontroversial, since a good many philosophers doubt whether it really makes any sense (see, for example, Wiggins 2001, ch. 7).

The probable simplicity of personal identity

149

Suppose, however, that we set aside all such doubts. Does this mean that some suitably revised version of Locke’s criterion – call it a neo-Lockean criterion – may be expected to satisfy all the requirements of an adequate criterion of personal identity? Very arguably, the answer is “no,” no matter what further modifications to Locke’s original criterion are made. For, as we are about to see, any version of Locke’s criterion seems vulnerable to another charge of implicit circularity of a quite different sort from Butler’s. the fatal circularity in any neo-lockean criterion Recall again that any version of the Lockean criterion appeals to the notion of a person, P, remembering – or, if one prefers, quasi-remembering – some past experience, e. And it must be emphasized here that it is vital that the experience, e, that is had by some person at an earlier time is the very same experience that is later remembered by P, if P is to be identified, in accordance with the criterion, with the person existing at the earlier time. So the criterion certainly presupposes some account of the identity conditions of experiences. But that means that we now need to ask ourselves how experiences themselves are individuated. Such items are mental states or events. So what are their identity conditions? We can already rule out the Davidsonian criterion of identity for events as a way of settling this question, because we found it to be implicitly circular. It was so because it sought to identify events on the basis of the sameness of their causes and effects, while also taking these causes and effects to be events themselves: so it defined “sameness of events” in terms appealing to sameness among events – a blatantly circular procedure leaving us no clearer as to what the identity of events consists in. But any neo-Lockean criterion likewise appears to be implicitly circular, albeit in a rather more roundabout way. For it is strongly arguable that the only adequate criterion of identity for mental states and events will be one which makes reference to their subjects – which, in the case of personal experiences, will be the persons who have those experiences (compare Strawson 1959, ch. 3). Let us focus then on the specific case of experiences, although the same reasoning will apply to mental states or events of any kind, so that the following objection extends to any “psychological” account of personal identity. On the view now being recommended, part of what makes an experience of mine numerically distinct from a qualitatively indistinguishable experience of yours is the very fact that it is mine as opposed to yours. The only other possible distinguishing feature seems to be the time at which an experience

150

e. j. lowe

occurs. In short, the following seems to be a very plausible criterion of identity for personal experiences: ð∀xÞð∀yÞððx is a personal experience & y is a personal experienceÞ→ ðx ¼ y ↔ ðx and y are qualitatively indistinguishable & ð∃P1 Þð∃P2 Þð∃t 1 Þð∃t 2 ÞðP1 has x at t 1 & P2 has y at t 2 & P1 ¼ P2 & t1 ¼ t2 ÞÞÞÞ; where P1 and P2 are variables ranging over persons and t1 and t2 are variables ranging over times. In plain English, what this formula says is just this: if x and y are personal experiences, then they are the same personal experience if and only if x and y are qualitatively indistinguishable experiences had by the same person at the same time. It is quite clear that the criterial relation invoked by this criterion is, as required, an equivalence relation. But, equally, it is obvious that it appeals to the notion of sameness between persons and hence presupposes that notion. Accordingly if, as I strongly suspect is the case, this is the only adequate criterion of identity for personal experiences, then any neo-Lockean criterion of personal identity is implicitly circular in that it will need to rely on the foregoing criterion for a specification of the identity conditions of the experiences to which it appeals for the purposes of identifying persons. Clearly, at any rate, we cannot both individuate persons in terms of their experiences (as any neo-Lockean criterion attempts to do) and individuate personal experiences in terms of the persons having them (as the foregoing criterion does). And to the extent that the foregoing criterion of identity for personal experiences looks to be in good order, it is any form of neo-Lockean criterion that must be rejected as inadequate (compare Lowe 2009, ch. 8). Note, incidentally, that I use the term “individuation” in this context in the way that Davidson does in his paper “On the Individuation of Events” (Davidson 1980 [1969]), whereby a criterion of identity for Ks is taken to be a principle of their “individuation.” I do not mean to imply by this that I favor this usage myself, but I adhere to it for present purposes owing to its continuing popularity and thus in order to avoid possible confusion. Now, some philosophers will undoubtedly consider that my reasoning in this section is fallacious, or at least suspect. Since this is likely to be a common reaction, I shall return to the matter shortly, discussing a quite specific objection to my line of argument. Before doing that, however, I want to broaden our perspective somewhat by considering, albeit rather

The probable simplicity of personal identity

151

briefly, some non-Lockean approaches to the question of personal identity. I shall then introduce and try to motivate the simplicity thesis that there is in fact no informative and non-circular criterion of identity for persons.

some non-lockean approaches and the simplicity thesis I have focused so far on the neo-Lockean approach because it is, deservedly, by far the most prominent one in the modern literature on personal identity, whether it is being endorsed or being attacked. But something should now be said about some alternative approaches. First of all, we have so far considered only the prospects for a one-level criterion of personal identity. But on some views of what persons are, a two-level criterion might seem more appropriate – for instance, if persons are taken to be functional states or roles that objects of appropriate kinds can occupy. Thus, one such view would be that a person’s body, or a special part of that body, such as its brain, is the object that occupies the functional role in question. Suppose that being a person is a functional role of a brain (for example, it might be taken to be the role of being a producer of first-person thoughts). Then a criterion of personal identity could be expected to take something like the following two-level form, where the variables of quantification range over brains: ð∀xÞð∀yÞðthe person of x ¼ the person of y ↔ x and y are RP -relatedÞ; where “RP” denotes a certain equivalence relation holding between brains. (In plain English: the person of brain x is identical with the person of brain y if and only if brain x and brain y are RP-related.) Indeed, on one view, this equivalence relation would simply be identity itself. There would be no circularity in the criterion on this account, since it would simply be defining personal identity in terms of brain identity, and persons and brains are here being taken to be items of quite different kinds. So this approach is by no means identifying a person with his or her brain. The brain-identity criterion of personal identity just implies that a person’s identity tracks that of the person’s brain – so that, for example, if a person A’s brain is transplanted into the evacuated head of another person B’s body, then person A acquires person B’s body; and if person A’s brain is switched with person B’s, then we have a body swap, with person A acquiring person B’s body and person B acquiring person A’s body. The scenario is really very similar to Locke’s famous imaginary example of the prince and the cobbler,

152

e. j. lowe

who supposedly undergo a body swap – although what Locke envisaged was that the soul of the prince entered the cobbler’s body and the soul of the cobbler entered the prince’s body (1975 [1690], ii, xxvii, 15). However, although Locke thought that this scenario was in principle possible, he did not, of course, subscribe to a soul-identity criterion of personal identity, because he thought that the same person could in principle have different souls at different times and that the same soul could, at different times, be the soul of different persons. Incidentally, an interesting modern defense of a soul-identity criterion is offered by Richard Swinburne (1986, ch. 8). I have nothing to say in recommendation of a two-level approach like the brain-identity criterion, although it will clearly appeal to some philosophers and psychologists. Such an approach clearly seems inappropriate if we regard the term “person” as denoting a distinct kind of substantial being – an individual substance – rather than a certain kind of state or role that such a substance can occupy. Certainly, common sense and ordinary language strongly suggest the former view. I feel myself to be some thing, with distinctive properties such as thought and feeling, rather than my being merely some property or feature of some other thing, such as my brain. But how, then, can it be explained why a satisfactory one-level criterion of personal identity that supports this conviction appears to be so very elusive? My tentative suggestion in answer to this question is that personal identity is just so basic in our ontological scheme that we should not really expect to be able to formulate such a criterion. A crucial point here is that, as we have seen, one-level criteria of identity for objects of a kind K always appeal to entities of other kinds in specifying a criterial relation for K-identity. They must do so in order to avoid circularity. This, for example, is why the principle that if x and y are sets, then x is identical with y if and only if x and y include exactly the same sets is defective as a criterion of identity for sets, as indeed is Davidson’s criterion of identity for events. But if persons really are fundamental in our ontological scheme, as I very much suspect they are, then we simply should not expect to be able to appeal to other entities of suitable kinds in their case. Rather, persons will be among the things that criteria of identity for other less fundamental kinds of entity appeal to – entities such as mental states, for example. That being so, we should probably conclude that personal identity is primitive and “simple,” in the sense that nothing more informative can be said about the identity of persons than that in some cases it just obtains and in others not (Lowe 2009, ch. 8). Certainly, it seems clear that not all entities can be provided with informative and non-circular criteria of identity, once again on pain of

The probable simplicity of personal identity

153

circularity – not necessarily circularity in any single criterion of identity, but at least circularity in the entire set of identity criteria putatively governing, between them, all the entities that there are. This kind of circularity is exhibited in the conflict, discussed earlier, between any neoLockean criterion of personal identity and the criterion of identity for personal experiences proposed in the previous section. That persons are what P. F. Strawson (1959) called basic particulars is in fact very plausible, for all the reasons he gave. It is something very close, if not identical, to that proposal that I am recommending here. And it should be clear that, in maintaining for these reasons that there is no informative and noncircular criterion of personal identity, I am not in the least voicing any kind of skepticism about the reality of persons or their persistence through time. Obviously, it is very difficult to prove a negative thesis like the simplicity thesis, since this would require us to refute any conceivable “complex” account of personal identity. The most that one can realistically hope to do is to show that leading complex accounts are fatally flawed and that the simplicity thesis is well motivated, both of which tasks I hope to have carried substantially forward in this chapter, even if I have not by any means said the last word on the matter. reply to an objection I now return to an objection to my argument (pp. 149−50) that any neoLockean criterion of personal identity is subject to a fatal circularity. Commenting on an earlier presentation of this line of argument (in Lowe 2009, ch. 8), Sydney Shoemaker (Chapter 6 of this volume) voices the following complaint. He accepts, first of all, that “there certainly is a good sense in which the individuation of conscious states involves the identity of persons.” But he points out that the same holds with regard to any state and the thing that is its subject or bearer: “Any state, psychological or not, necessarily belongs to the thing [which is its subject], and so gets its identity from the identity of its subject.” He concludes: So if this sort of involvement of the identity of persons makes psychological accounts of personal identity circular, any account of the identity of any sort of persisting thing in terms of relations between successive states [of such things] would be circular. If this were right, Lowe’s argument against the complex view of personal identity would generalise into an argument against the complex view of any sort of persisting thing – which would be contrary to his intent, since he thinks that in the case of entities other than persons there can be noncircular constitutive criteria of identity.

154

e. j. lowe

Now, I agree with the first sentence in this passage, but not with the second. I do indeed think that “any account of the identity of any sort of persisting thing in terms of relations between successive states [of such things] would be circular.” But the second sentence does not follow logically from the first. I hold that there can be non-circular “constitutive” criteria of identity for persisting things of certain kinds which are not framed in terms of successive states of the things in question. Instead, they can be framed, for instance, in terms of the persisting parts of such things. For example, I am entirely happy to follow Locke in saying that if x and y are (what he called) masses or parcels of matter, then x is identical with y if and only if x and y are composed of the same “atoms.” As he puts it: “the Mass, consisting of the same Atoms, must be the same Mass, let the parts be never so differently jumbled: But if one of these Atoms be taken away, or one new one added, it is no longer the same Mass or the same Body” (Locke 1975 [1690], ii, xxvii, 3). As for the persistence of atoms – or, more precisely, what physicists now call “fundamental particles” – I hold that it is simple, like the persistence of persons. But Shoemaker’s objection to my line of argument runs deeper than this. He considers that I have misunderstood what is really at issue in discussions of personal identity (and indeed in discussions of the identity conditions of things of other kinds). As he puts it: In such discussions what will be at issue is what it is for different states, either occurring at the same time or occurring at different times, to belong to one and the same person. Each of the states will necessarily belong to a person, and what is at issue is what it is for it to be the same person all of them belong to.

However, this suggests that Shoemaker is implicitly seeking what I earlier called a two-level criterion of identity for persons, of the following form, where the variables of quantification range over person-states: ð∀xÞð∀yÞðthe person of x ¼ the person of y ↔ x and y are RP -relatedÞ: In plain English: the person of person-state x is identical with the person of person-state y if and only if person-state x and person-state y are RP related. Now, clearly, person-state x and person-state y will be, as Shoemaker puts it, states belonging to the same person just in case the person to which person-state x belongs is identical with the person to which person-state y belongs, that is, just in case the person of person-state x is identical with the person of person-state y – which is precisely what the lefthand side of the above biconditional says. Hence the above biconditional would certainly appear to express the kind of truth that Shoemaker is

The probable simplicity of personal identity

155

looking for when he seeks to establish “what it is for it to be the same person [that different person-states] belong to.” This appears to confirm that what he is implicitly seeking is a two-level criterion of personal identity of the above form. I have already explained why I do not favor a two-level approach myself in the case of persons: namely, because I do not regard persons as “functional” entities of any kind but rather as individual substances. But, in any case, two-level identity criteria are no less subject to the requirement of noncircularity than are one-level identity criteria. If Frege had tried to formulate a two-level criterion of identity for lines in terms of some equivalence relation holding between their directions, this would have vitiated his own two-level criterion of identity for directions in terms of the parallelism of their lines, because advocating both criteria together would have committed him to a circularity of individuation whereby neither lines nor directions get their identity conditions properly determined. This is not to say that such a pair of principles – if both are true – must be entirely uninformative in the sense of being just trivial, although they will indeed be uninformative as to what the identity conditions of the entities in question actually are, precisely on account of the circularity that they harbor. In any case, the important point is that adequate identity criteria, whether considered singly or together, need to avoid circularity. Hence, if I am right that person-states, including the mental states of persons, have their identity conditions determined at least in part by reference to the persons whose states they are, then a two-level identity criterion for persons of the form indicated above will indeed introduce a fatal circularity into an account of the identity conditions of persons, no less than in the case of a one-level “neo-Lockean” criterion of the kind discussed earlier.

chapter 8

Reply to E. J. Lowe Sydney Shoemaker

The following is a response to what E. J. Lowe says in the last section of Chapter 7 in reply to criticisms I made of his view in Chapter 6. Lowe thinks that holding that there is what he calls a two-level criterion of personal identity is incompatible with holding that persons are substances. I think there is no good reason to believe this. It would be closer to the truth to say that the status of persons as substances requires that there be such a criterion – i.e. that there be a diachronic unity relation that holds between the successive states of a person. Lowe also thinks that the view that there is such a criterion violates the requirement that an account of the identity conditions be non-circular. This too I see no good reason to believe. What he thinks is a reason is the fact that states of persons get their identity from the persons that have them, it being a necessary condition of states being identical that the persons having them are identical. But what the two-level criterion of personal identity gives is an account of what the trans-temporal identity of a person consists in, and the identity of persons that figures in the identity of states is not trans-temporal identity. To be sure, a full account of the identity conditions for persons should include an account of synchronic identity, i.e. of what it is for simultaneous states to belong to one and the same person. Such an account might be along functionalist lines – states are “co-personal” (to use Russell’s term) in virtue of being such that they are apt to have certain joint effects in certain circumstances. This would have to be combined with an account of diachronic identity, since some of the effects would have to be future states of the subject of the states that produced them. What the account offers is an account of what different states having the same possessor consist in, and it surely is not made circular by the fact that each individual state necessarily has the possessor it has.

156

chapter 9

The non-descriptive individual nature of conscious beings Martine Nida-Ru¨melin

introduction According to an idea that has been around for centuries, the individual nature of individual things transcends any description.1 We cannot grasp, according to this intuition, what makes a particular thing that particular thing. Individuals do not have a “qualitative essence.” Every particular is, according to this idea, the individual thing it is in a primitive and nonreducible manner; it is not in virtue of any property or combination of properties that an individual is the one it is. According to this intuition it is impossible to get “a cognitive hold” of an individual thing as such in any “descriptive manner.” These formulations are far from clear and they only hint at an intuition which, however, is not difficult to “see.” Let us use the following formulation to express the idea: individual things have a nondescriptive individual nature. The proposal of the present chapter may be expressed as follows. Not all individual things have a non-descriptive individual nature: tables, mountains and other concrete material objects do not. However, conscious individuals do have that particular status. I will propose an account of what it is for an individual to have a non-descriptive individual nature 1

Famously Duns Scotus argued for the view that the individuality of an individual is to be explained by its “haecceitas,” which he apparently took to be a non-qualitative property specific to the individual at issue (Cross 2010 [2003]; Gracia 1994). The idea that, in a sense, the individual in its individuality cannot be conceived of at least by the human mind and can become the object of one’s thought only in some kind of “ostension” appears to occur, as far as I understand it, in the fourteenth century and in particular in the work of Pierre Auriol (Suarez-Nani 2009). For contemporary discussions of the idea that individual things have, in some sense, a non-descriptive nature, see, for example, Black (1952), Adams (1979) and Swinburne (1995). I am not claiming that all these people had the same underlying claim in mind; nor is the notion of a “perfect individual” introduced in the present chapter meant to capture the underlying common idea. But I do believe that all these ideas have a common intuitive source; to show this would, however, require a serious and cautious analysis of the relevant texts. The notion of a “perfect individual” here proposed might be helpful to capture an aspect of that common intuitive source.

157

158

martine nida-ru¨ melin

and I will motivate the claim that conscious beings, contrary to ordinary material objects without consciousness, have a non-descriptive individual nature.2 I hope that I might at least succeed in rendering the intuitive attraction of that metaphysical claim about the individual nature of conscious beings clearly visible to the reader. What I am going to say is closely related to the view about trans-temporal personal identity which has been called the “simple view” and which has been defended by, among others, Bishop Butler, Thomas Reid, Roderick Chisholm, Richard Swinburne and E. J. Lowe (Butler 1975 [1736]; Reid 1975 [1785]; Chisholm 1970a, 1989; and the contributions of Swinburne and Lowe to this volume). A difference is, however, that I will not restrict the relevant claim to the human domain. All reasons for the view that people are perfect individuals apply with the same force, or so I will argue, to any other conscious being. I will not explicitly address the issue about trans-temporal identity in what follows. It will, however, become quite obvious for the reader that the so-called simple view about trans-temporal personal identity and the claim that conscious beings have a non-descriptive individual nature share their intuitive core. Presupposing the truth of these views one may say that they are based on the same deeper insight. the constitutional basis and metaphysical bases of an individual’s existence During a walk on a beach near a small town called Kiola in Australia in July of this year, I found a beautiful small flat black stone and I took it with me to Fribourg. It is now lying in front of me on the left-hand side of the keyboard I am writing on. Let us call that black stone “Nero.” Suppose our planet had developed in a very different manner. Suppose that – due to some difference in volcanic activities or to a clash of the earth with a huge meteor – the distribution of continents over the earth’s surface had evolved differently and quite different biological species had come into existence. Let us now consider the following question: what would have to be the case under circumstances of that kind for it to be true that this 2

A similar view is defended in Nida-Rümelin (2006, 2010, 2012). My view has, however, changed in two respects: in those earlier publications I defended the claim that our conception of ourselves and other subjects of experience as “perfect individuals” (in a sense similar to the one introduced here) has its origin in special features of first-personal self-directed thought which carry over to the way we think of others. I now believe that it is a mistake to see these features of first-personal self-directed thought as fundamental. Furthermore, I have changed my view with respect to whether conscious beings have a non-trivial metaphysical basis (in the sense defined in this chapter). The denial that there is such a basis is no longer part of my view.

Consciousness and non-descriptive individuality

159

particular stone, Nero, had nonetheless come into existence in these circumstances? To address this question is to ask for Nero’s individual nature, as I will use the term. The question “What is Nero’s individual nature?” can be translated, without loss, into the following question: what are the features of a situation such that a situation having those features is ipso facto a situation in which Nero exists? To search for Nero’s individual nature is, in other words, to search for those conditions that make it the case that a situation is a situation in which Nero exists. Let us call those conditions the constitutional basis of Nero’s existence. If C is the constitutional basis of Nero’s existence, then any situation fulfilling these conditions is ipso facto a situation in which Nero exists. If C is the constitutional basis of Nero’s existence, then a description of C fully expresses what it takes for Nero to exist under given factual or counterfactual circumstances. Any full description of C then expresses what the fact that Nero exists consists in. In Nero’s case, or so I suggest, the constitutional basis of Nero’s existence is something like this: there is an object with roughly the same history, composed of roughly the same individual particles arranged in roughly the same way, having roughly the same form. It is the existence of such a stone in a given situation, of some stone having those commonalities with the real stone Nero, which renders it true that Nero, this stone, exists in that situation. That there is such a stone in a given situation is what it takes for Nero to exist in that situation, or so I suggest. (An argument for this claim will follow.) I wish to distinguish another, similar notion from the notion of the constitutional basis of an individual’s existence: the notion of a metaphysical basis of an individual’s existence. A condition is a metaphysical basis of, for example, Nero’s existence if and only if its satisfaction in a given situation metaphysically implies that Nero exists in that situation. In other words: if M is a metaphysical basis of Nero’s existence then M metaphysically necessitates Nero’s existence. Circumstances where M is satisfied and yet Nero does not exist are metaphysically impossible.3 Let us, for instance, consider a situation in which there is a perfect counterpart of Nero. A perfect counterpart of Nero is a stone (existing in a counterfactual situation under consideration) which came into existence in that situation in exactly the same way as Nero did in the real world; it consists of the very same particles 3

The notion of metaphysical necessity used here is of course problematic and there is no consensus about its precise meaning. I use the term here in roughly the following way: a situation is metaphysically possible if it is compatible with the nature of the things, properties and relations at issue. To say that p metaphysically necessitates q is then to say that there is no situation compatible with the nature of the things, relations and properties at issue where p but not q is the case.

160

martine nida-ru¨ melin

and these particles are arranged in that stone in exactly the same manner. The resulting stone looks like Nero and it has the same form. This condition, or so I will argue, is a metaphysical basis for Nero’s existence. There is no possible situation in which there is a perfect counterpart of Nero and yet Nero does not exist. The constitutional basis of an individual’s existence is, obviously, also a metaphysical basis for that individual’s existence. If C is what makes it the case that a given real individual exists, then there are no metaphysically possible situations where C is satisfied and yet the individual does not exist.

non-descriptive constitutional basis for an individual’s existence I propose to understand what it is for an individual to have a non-descriptive individual nature in the following way: a thing has a non-descriptive individual nature if and only if the constitutional basis for its existence is non-descriptive. The constitutional basis of a thing’s existence is nondescriptive if and only if there is in principle no way to say in a noncircular manner what constitutes the existence of that particular thing. If this is so, then any attempt at formulating a constitutional basis for the thing’s existence fails in the following way: we can only find formulations of such conditions that already contain in a trivial manner the assumption that this particular thing exists. I will use Nero’s example to make this notion of what it is for an individual to have a non-descriptive individual nature intuitively more accessible. There is always an easy way to formulate metaphysically sufficient conditions for a thing’s existence. For instance, every possible world in which Nero lies in a particular spot in Australia today is a world where Nero exists; that Nero lies in that spot today is a metaphysical basis for Nero’s existence. However, the name “Nero” is used in this description as a rigid designator: it is used to say of this particular stone that it is lying in a particular place in the counterfactual situation; it follows trivially that this object in front of me exists in the circumstances so described.4 We have succeeded in formulating a metaphysical basis for Nero’s existence but the condition is trivially 4

For readers not familiar with the notion of rigid designators introduced in Kripke (1981) the following rough explanation will make the notion clear enough for present purposes. A rigid designator is a term used to refer to an individual in the real world and to talk of that individual in the description of counterfactual circumstances.

Consciousness and non-descriptive individuality

161

sufficient for Nero’s existence.5 In a similar way we can formulate trivially sufficient conditions for Nero’s existence demonstratively referring to Nero saying that this stone has certain properties in those conditions. In general, there is always an easy way to describe a metaphysical basis for an individual’s existence: using a rigid designator D which refers to that individual the counterfactual situation is described as one fulfilling the condition “D exists” or “D has property P” (where P is a property the referent of D can have). The use of a rigid designator for the thing at issue trivializes any such proposal. Suppose now that, in a given case, we can only formulate a metaphysical basis of the thing’s existence by rigidly referring to that thing in a way which trivializes the proposal. We then know that there is no non-circular way to formulate a constitutional basis for that thing’s existence either (since every constitutional basis is a metaphysical basis). In that case there is no way to say in a non-circular manner what it takes for that thing to exist in counterfactual circumstances. The fact that this thing exists in a given situation is then an irreducible simple fact which can only be expressed rigidly referring to that thing and saying of it (directly or indirectly) that it exists. In that case the thing at issue has a non-descriptive individual nature. This notion can now be defined as follows: an individual X has a non-descriptive individual nature (is a perfect individual) if and only if the constitutional basis of its existence can only be described using a rigid designator which refers to X.

the individual nature of a stone Is Nero’s individual nature non-descriptive? If so, then we must rigidly refer to Nero in any description of the constitutional basis of Nero’s existence. There is no reason to think that this is the case, or so I will argue. Let us first consider situations where there is a perfect counterpart of Nero. In those situations there is a stone which looks just like Nero, consists of the same material (the same individual particles) arranged in the same way, and came into existence (in that counterfactual situation) on the basis of the same causal processes. This could be so, for instance, if the situation we consider is only minimally different from the real one. Is it metaphysically possible that all these assumptions are satisfied in a counterfactual situation and still Nero does not exist in that situation? The answer, I suggest, must be “no.” A situation in which there is a stone exactly like Nero in all respects (same 5

The triviality is due to the well-known semantic property of names: names are rigid designators, as has been argued famously and convincingly by Kripke (1981).

162

martine nida-ru¨ melin

form, same material, same particles arranged in the same manner, same origin) ipso facto is a situation in which Nero exists and the stone satisfying all those criteria in the counterfactual situation is the one lying here next to me. Is this already the constitutional basis of Nero’s existence we are searching for? Not yet, since the condition is too strong. A slightly different stone (it has, for example, lost a few particles) would still be Nero. But we may safely assume that the constitutional basis of Nero’s existence is some weaker condition than the one just considered: it is a condition implied by the existence of a perfect counterpart of Nero which does not imply the existence of a perfect counterpart of Nero. If we wish to decide whether Nero has a non-descriptive individual nature we have to decide whether any formulation of that weakened condition requires rigid reference to Nero. If we can see that the strong condition (that there is a perfect counterpart of Nero) can be formulated without rigid reference to Nero, then we know that the constitutional basis of Nero’s existence can be described without rigid reference to Nero as well. So if the condition that there is a perfect counterpart of Nero can be formulated without rigid reference to Nero, then Nero does not have a non-descriptive individual nature. Let us therefore have a closer look at the way that stronger condition can be described. There is rigid reference involved in the description: we refer to Nero (using a name or some other rigid designator) when we say that, in the situation we wish to describe, some object shares the relevant properties (its form, the stuff it is made of and its origin) with Nero, the stone next to me now.6 However, it is in principle possible to free that description from rigid reference to Nero: rigid reference to Nero does not occur essentially – as one may put it – in that description. It is practically impossible for us to introduce names for every single particle involved in Nero’s genesis and it is practically impossible for us to name every single particle making up the material that stone is constituted of. But it is clear that it would in principle be possible to eliminate rigid reference (and indeed any reference) to “Nero” in the description of the relevant condition. The condition only concerns the way a huge number of particles interacted with one another in those events that led to Nero’s formation and it concerns the way a huge number of particles were finally arranged. So, in principle, there is a way to formulate the condition that there is a perfect counterpart of Nero without rigidly 6

Another way to see that rigid reference to Nero does not occur essentially in that description is this: to describe a situation in which Nero has a perfect counterpart we only have to pick out Nero in the real world. Therefore, it does not make a difference whether we use a rigid or a non-rigid designator for that purpose.

Consciousness and non-descriptive individuality

163

referring to Nero.7 Rigid reference to Nero does not occur essentially in the description of that condition and we can conclude: Nero is not a perfect individual. Is there any reasonable way to resist this conclusion? Can someone, with convincing reasons, insist that Nero and other material things have a nondescriptive individual nature? In the present chapter I cannot try to address theoretical arguments a metaphysician might develop in favor of the view that material things have a non-descriptive individual nature. However, if the way in which I propose to understand what it is for an individual to have a non-descriptive nature is accepted, then the above reasoning seriously undermines that view. There are, in principle, two ways to avoid the conclusion. The first is to doubt the explication proposed of what it is to have a non-descriptive individual nature. According to this response, an important sense of the underlying idea has not been captured by my explication. I need not exclude that this might be so in the present context.8 My aim here is to render visible a difference between ordinary material objects (without consciousness) and conscious beings, a difference which I take to be important and “deep.” I will try to express the difference I have in mind by motivating the view that conscious beings but not ordinary material things have a non-descriptive individual nature in the sense here explicated. The success of this enterprise does not depend on whether or not the notion here proposed of what it is to have a non-descriptive individual nature already fully captures the underlying idea. So I can set this first response aside.9 A second response to the above argument is to doubt that there is a noncircular way to describe the constitutional basis of Nero’s existence. Any description of that constitutional basis would require, according to that response, rigid reference to Nero. An opponent who wishes to defend that view would have to argue that any non-trivial candidate for a constitutional basis of Nero’s existence fails to capture the particular feature of a situation 7

8

9

This is not to say that any rigid reference can be eliminated. We need rigid reference to parts of Nero, or to the stuff it is made of, or to the particles that constitute Nero. But we do not need rigid reference to Nero. For instance, someone may agree with the reasoning presented and yet insist that Nero has a nondescriptive individual nature in virtue of being composed of that particular stuff. I would respond that this is only a derivative way of “having a non-descriptive individual nature” which is not grounded in Nero’s nature as such. To free myself from the assumption that I have been able to fully capture the idea of an individual’s non-descriptive individual nature, I could abandon using this expression and just use the term “perfect individuals” (where what it is to be a perfect individual is, by stipulation, fully captured by the definition proposed).

164

martine nida-ru¨ melin

which constitutes that Nero exists in that situation. (A non-trivial candidate here is a condition which can be described without rigid reference to Nero.) If this were so, however, then even the strongest condition one might propose for a non-trivial constitutional basis − the condition that there is a perfect counterpart of Nero − would fail to capture that feature.10 But what could that feature consist in? Do we have any positive understanding of that supposed feature? It is evident, or so I claim, that we do not have any positive understanding of the feature that must be added to a world in which there is an object exactly like Nero, made up of the same material and having the same origin, in order to make it a world where Nero is that particular object. It clearly looks as though we have described a world in which Nero exists when we have described a world in which there is an object having all those relevant properties Nero has in the real world. It “clearly looks like that” for a simple reason: we have no conception of the feature that might still be lacking in those conditions in order for Nero to exist and in order for Nero to be identical with the particular stone so described. At this point the reader might suspect an illegitimate move from a claim about positive conceivability to a substantial metaphysical thesis. It might look as if the argument can be put as follows: there is no positively conceivable feature that might still be lacking for Nero’s existence in situations fulfilling the relevant condition. Therefore, there is no such feature that can be lacking in a situation fulfilling that condition (“can” in the sense of metaphysical possibility). This, however, is not the argument I have in mind. I am not trying to establish in a direct way that there is no such “hidden feature” underlying the existence of Nero which is not yet captured in the assumption that there is a perfect counterpart of Nero. My claim is weaker: I am trying to convince the reader that we have no reason to assume the existence of such a hidden additional feature. We have no conception whatsoever of that supposed additional feature which might still be lacking for the existence of Nero in a world in which there is a perfect counterpart of Nero. Therefore, we have no reason to assume that there is such a hidden feature; and, therefore, we have no reason to assume that Nero has a nondescriptive individual nature. The situation is different, or so I will argue, in the case of conscious individuals. In the case of conscious individuals we do have a positive conception of the relevant additional feature. 10

The reasoning here is this: if that strongest condition does capture that feature, then the opponent must be wrong. Either he has to admit that this strongest condition is itself the constitutional basis or that it implies the constitutional basis. In both cases the constitutional basis can be formulated without rigid reference to Nero.

Consciousness and non-descriptive individuality

165

perfect counterparts of conscious individuals If you, the reader, are a perfect individual, then there is no way to formulate the constitutional basis of your existence without rigidly referring to you. To judge whether this is so, it will be helpful to consider situations in which there is a perfect counterpart of you: a child came into existence on the basis of exactly the same biological processes that led to your existence, it was born to the same parents under the very same conditions, and it is genetically just like you. The child grows into an adult who looks like you actually look, who has your actual character, preferences and convictions, and who has chosen the same profession, the same friends and partners. Satisfying all these conditions is to be a perfect counterpart of you. In the description of these conditions I have been referring to you. But that rigid reference does not occur essentially in that description. In order to say, for instance, that the child has the biological origin you have in reality, we certainly need to rigidly refer to biological cells, but this is not to rigidly refer to you. All reference to you in the description of the counterpart’s properties could in principle be eliminated. If the description of a situation where there is a perfect counterpart of you fails to capture what it takes for you to exist in a counterfactual situation, then you have a non-descriptive individual nature. This can be seen by the following consideration. Any non-trivial constitutional basis of your existence is a condition which is weaker than the assumption that there is a perfect counterpart of you. Therefore, if the stronger assumption that there is a perfect counterpart of you does not capture what it takes for you to exist in a counterfactual situation, then no candidate for a non-trivial constitutional basis captures what it takes for you to exist. (What a stronger condition cannot capture cannot be captured by a weaker one.) It follows that every formulation of the constitutional basis of your existence requires rigid reference to you, which means, according to the definition proposed, that you have a non-descriptive individual nature. So does the condition that there is a perfect counterpart of you capture what it takes for you to exist in a given factual or counterfactual situation? Here is the beginning of an argument for a negative answer to that question. For simplicity I suppose that, if you exist in a situation where there is a perfect counterpart of you, then you are that perfect counterpart. It follows that you do not exist in such a situation if you are not that counterpart. Now let us try to make sense of the difference between the two following situations: S1 is a perfect counterpart of you and you are that counterpart, S2. There is a perfect counterpart of you, yet you do not exist. Here is a

166

martine nida-ru¨ melin

way to understand the difference: in S1 you have the counterpart’s body, you see that body as your body when that person stands in front of a mirror, you are the one who is active when that person acts, you feel pain when that body is hurt: in short, it must be the case that you are the one who lives that person’s life. In S2 you never “wake up” in order to discover the world, you never experience anything, you do not exist. Pointing out this difference between S1 and S2 we have singled out a potential feature which constitutes the difference between S1 and S2. And we have a clear positive understanding of that feature.11 I will come back to the way in which we conceive of that feature, but before doing so I must clarify an important detail. It will be objected that S2 is not a genuine possibility − that no metaphysically possible situation fulfills that description. The opponent might, for instance, believe in the necessity of origin, in the claim that any individual has its specific origin essentially (no other individual could possibly come into existence on the basis of the processes that led to its existence). The principle of the necessity of origin excludes the metaphysical possibility of S2. It is important to note that I need not argue against that view; it is not my aim to establish the metaphysical possibility of S2. My present aim is to show that we have a clear positive understanding of a particular feature which would distinguish these two possibilities, if S2 were metaphysically possible. In other words: we have a clear positive understanding of what it would take for S2 to be realized if it was metaphysically possible. We have a clear positive understanding of a feature such that its presence in S1 and its absence in S2 would constitute a genuine difference between S1 and S2 (if S2 were possible). This claim is compatible with the thesis that S2 is metaphysically impossible and it is sufficient for my argument. Compare the situation with Nero’s case. In Nero’s case we may describe two situations in an analogous manner: S10 is a perfect counterpart of Nero and Nero is that counterpart, S20 . There is a perfect counterpart of Nero and Nero does not exist. In this situation we do not have a clear positive understanding of some feature such that its presence in S10 and its absence in S20 would constitute a genuine difference between S10 and S20 if S20 were metaphysically possible. We cannot form any positive understanding of an additional feature that is necessary for Nero’s existence and not yet mentioned in the assumption that there is a stone just like Nero, with the same origin and constituted of the same concrete stuff.

11

Some readers will be reminded of the last paragraph of Nagel (1965). For comments on that citation and its relation to the view proposed here, see Nida-Rümelin (2012).

Consciousness and non-descriptive individuality

167

the central conceptual disanalogy In Nero’s case, as argued before, there is no reason to suppose that this particular stone has a non-descriptive individual nature. The reasoning was this: if Nero had a non-descriptive individual nature, then there is a feature of possible situations which constitutes Nero’s existence in those situations, which, however, is not yet captured in the assumption that there is a stone just like Nero with respect to its origin, form and internal constitution. However, upon reflection it is quite easy to see that we do not have any positive conception of what that feature might be. Since we do not have any such conception, we do not have a reason to suppose that there is such a feature. Now let us return to your case. Let us consider again a situation in which there is a person with the same biological origin as you have in the real world, a person who lives a life exactly like the one you are actually living, a person who looks like you and is like you in all psychological and physical respects. Is there anything required for your existence which is not yet mentioned in that description? The answer, or so I suggest, must be “yes”: the feature required and not yet mentioned is, quite simply, that you are the one who lives that counterpart’s life, that it was you who came into existence when that counterpart emerged. The description, by itself, does not exclude that you do not exist in the situation so described. It allows for the situation to be one where there is nothing at all from your perspective.12 We understand the description of a case where someone is a perfect counterpart of you and still you never came into existence. We have, as one might say, the conceptual resources to understand that description; we are able to positively conceive of a situation satisfying it. This is more than to say that there is no contradiction involved in the description of that case. There is no difference in this respect (lack of contradiction in the description of the case) between the example about Nero and the example about you. There is no contradiction involved in the description of a situation where some stone is a perfect counterpart of Nero and still Nero − this stone − does not exist. The description might well be said to be in conflict with plausible ontological claims about the nature of stones and their identity across possible worlds. However, the description in itself is non-contradictory. The same applies to your case. So far, the two cases are analogous. 12

I here presuppose that not every metaphysical consequence of a description is part of the content of the description. The description might metaphysically exclude that you do not exist; still, what it says about the situation (its content) is silent about whether or not you exist.

168

martine nida-ru¨ melin

The difference lies in the fact that in a specific way we cannot make sense of the idea that there is a perfect counterpart of Nero and yet no Nero while we can make sense of the parallel idea in a case concerning you. To see this clearly it is helpful to go back and forth in one’s thought between considering a situation where there is a perfect counterpart of Nero which is Nero and considering a situation where there is a perfect counterpart of Nero which is not Nero. Doing this one can realize that it is impossible to get a grip on what the difference might consist in. We do not appear to be able to grasp a genuine difference between the two cases. Of course we may demonstratively refer to Nero and repeat “Well, the difference is that in the first situation this one is there and that in the second situation this one is not there.” But doing so does not help. We cannot thereby get a grip on what would make the case that, in the second situation, this one is not that perfect counterpart. No difference in the perfect counterpart we can in any way positively conceive of might constitute the difference between the counterpart being Nero and the counterpart not being Nero. The situation is quite different when we come back to the case concerning you. Here again, to see the difference, it is helpful to go back and forth between considering a situation in which there is a perfect counterpart of you and the counterpart is you and a situation where there is a perfect counterpart of you and the counterpart is not you (you do not exist at all). What does the difference consist in? In the first situation you are the one living that person’s life, you are the one who experiences the world from that person’s perspective, you have that person’s body, and you are the one who enjoys that person’s pleasures. In the second situation, you never come into existence, there is no world from your perspective: you do not exist. There is a feature we can positively conceive of which constitutes the difference. Considering a perfect counterpart of Nero we have, as pointed out earlier, no positive conception of any feature that counterpart could possibly lack such that it is not Nero, the stone here beside me. Considering a perfect counterpart of you we have a clear positive conception of a feature that counterpart could possibly lack so that it would not be you. If the counterpart has that feature, then, if we were in that situation, we would see you when looking into that counterpart’s eyes; if the counterpart lacks that feature then there is no way of meeting you in that counterfactual situation (you do not exist, you never started discovering the world). Some people immediately agree with the disanalogy claim presented in this section: they find the difference I have been describing obvious. Others, however, hesitate or spontaneously deny that there is such a difference between the case of the black stone Nero and the case concerning you.

Consciousness and non-descriptive individuality

169

This disagreement, I believe, is not a genuine disagreement. It is based, or so I hope, on misunderstandings. I will address these possible misunderstandings in the next section. clarifications of the conceptual disanalogy claim What I wish to render “visible” to the reader may be put like this: we have a clear positive understanding of the difference between a case where a perfect counterpart of you in a counterfactual situation is you and a case where a perfect counterpart of you in a counterfactual situation is not you. We understand that difference by referring to you, in the real world, and conceiving of a situation where you have the body of that counterpart and live that counterpart’s life as opposed to a situation where “there is nothing from your perspective” and you do not exist. In a sense, when conceiving of that difference, we take “your perspective” in thought, and conceive of the difference by considering the difference “from your perspective.” This description may help to attract attention to the feature at issue the absence or presence of which constitutes the difference between the two cases. However, it should be clear that the sense in which “we take your perspective” is not psychological: it is not taking your perspective by imagining being like you in important respects or by imagining being in your situation. The kind of “taking your perspective” at issue may itself be described as “non-descriptive.” One may put it like this: in taking your perspective in the conception of the difference, we conceive of the difference “in the first-personal way,” “from your perspective,” or, as one might say, “as if we were you.” In a sense we imagine being you, but we do so without imaging being like you in any respect. We refer to you, that subject of experience, and conceive of the difference between a case where this subject exists and a case where this subject does not exist. This is a fundamental difference for that subject (for you, in that case). We understand the difference by understanding the difference it would make for you. One might call this a “first-personal way” of grasping the difference. But it is not first-personal in the sense of involving reference to oneself or reflection on one’s own mental states: we are, after all, talking of you. It is first-personal in the sense that grasping the difference requires seeing what the difference consists in “from your perspective” in the sense commented on above. To say that we consider what the difference consists in from your perspective should not, however, be taken to involve the idea that the

170

martine nida-ru¨ melin

difference exists, according to that conception, only for you. On the contrary, according to our understanding, the difference is an objective difference, a genuine factual difference – a difference which is not dependent on the way anyone experiences or judges the situation. One might get confused here since the objective feature at issue concerns, in a sense, “the subjective realm.” It is a fact concerning your “access to the world,” concerning what you experience “from your perspective”; it involves subjective facts about you. But this does not imply that the relevant feature itself is in any sense only subjective. It would be a genuine fact about a perfect counterpart who is you that he or she is you, and it would be a genuine fact about a situation where there is such a counterpart non-identical to you that you do not exist in that situation. One might say that we have in a sense a first-personal conception of the difference between S1 and S2. The only way to grasp that difference is to grasp it in that specific first-personal way, in thinking of you by nondescriptively taking your perspective. That mental activity (nondescriptively taking perspectives) deserves its name: it does not involve imaging oneself as satisfying a certain description; it does not involve imagining being like you or being in your situation; it involves nothing descriptive; it only involves reference to a subject in the real world and considering possibilities “from its perspective.” This is just a metaphor, but hopefully a helpful one. Non-descriptive taking of perspective is not restricted to people or human beings. One may refer to a dolphin or to a raven and ask the very same kind of question in the very same sense “from its perspective.” We have a clear positive understanding of the difference, for instance, between a situation where a perfect counterpart of the dolphin Delfina is Delfina and a situation where an equally perfect counterpart of Delfina is not Delfina: the conceptual resources used in that understanding are the same as when we engage in the analogous conception with respect to a case concerning you − or so I suggest. A first misunderstanding of my disanalogy claim has already been mentioned. The claim that we have a clear positive understanding of the difference between S1 and S2 should not be confused with the metaphysical claim that there is such a feature which constitutes the difference between genuine possibilities S1 and S2. I do not claim that both situations are metaphysically possible: my claim does not presuppose the genuine possibility of S2. It may be a deep modal fact a metaphysician might discover that the existence of a perfect counterpart of you necessitates your existence. It might, for instance, be a metaphysical truth that only one specific experiencing subject (and no other) can possibly come into existence on the basis of a given concrete

Consciousness and non-descriptive individuality

171

fusion of cells. The claim here proposed is compatible with this hypothesis. As already said in the preceding section, the suggestion may be put like this: we have a positive understanding of what the difference between S1 and S2 would consist in (if both were genuine possibilities). This is why the claim that the existence of a perfect counterpart of X necessitates the existence of X is a substantial claim if you are the individual X at issue. It is no substantial claim if X is the stone Nero. (It is rather trivial in Nero’s case that Nero’s existence is metaphysically necessitated by the existence of a perfect counterpart. We have no reason to assume the contrary, not even a reason to withhold an opinion, since we have no positive conception of any additional feature a perfect counterpart might lack for being Nero.) I have been using the locution that “a feature constitutes a difference” or that it “constitutes your existence.” This locution may invite a second misunderstanding. Someone might reason as follows. We are assuming that there is no difference between the counterpart of you in the first situation (in which the counterpart is you) and the counterpart of you in the second situation in which the counterpart is not you. The human organism at issue consists of the same stuff in both situations and it has developed in the very same way from the same biological origin, the counterparts do not differ in any psychological or bodily property and they are phenomenally alike. So there is no difference between the counterparts in the two situations in virtue of which the one in the first situation is you and the one in the second situation is not you. So we cannot possibly have any understanding of the difference in virtue of which you are the counterpart in the first situation but not in the second. There is nothing to be said against this reasoning and I agree with its result. There is no difference between the counterparts in the two situations in virtue of which the counterpart is you in the first situation but not in the second. The reason is that the situations considered do not differ in any other respect: they only differ with respect to the perfect counterpart’s identity; they only differ with respect to who is the perfect counterpart. When directed against the disanalogy claim the above reasoning is based on the following misunderstanding: the claim is that we have a clear positive understanding of the feature which constitutes the difference; this is not to be confused with the different (and trivially false) claim that we have a clear positive understanding of a further difference in virtue of which the counterparts are not the same person. As this discussion shows, to clarify the view I am proposing it is helpful to clearly distinguish between what it is for a feature to constitute the difference between S1 and S2 and what it means for S1 and S2 to be different in virtue of

172

martine nida-ru¨ melin

a certain feature. It is quite natural to understand the second locution (but not the first) in the following way: when we search for a feature in virtue of which S1 and S2 are different, we are looking for a describable qualitative difference between the counterparts in virtue of which they are different people; we are searching for some independently understandable feature, a feature that can be described without already including in its description the assumption of identity or non-identity with you. But there is no such feature as the case is described, and this was precisely the point. The example could not serve its purpose if there was such a feature included in or compatible with the description of the example. On the other hand, the feature constituting the difference, as I understand that locution, is not submitted to those constraints. When we talk of the feature that constitutes the difference, the feature we are looking for may be an irreducible fact about personal identity (in my view it is such an irreducible fact). Using this locution we leave it open that the feature we are looking for is no other difference than the one we can only describe saying of you (rigidly referring to you) that the counterpart is you in the first situation and not you in the second. We are not here searching for a feature that can be described without already including in its description the assumption of identity or nonidentity with you.13 Since there is no difference between the two counterparts with respect to all their physical and psychological properties, with respect to their social environment and with respect to the kind of life they are living, one might be tempted to conclude that the alleged feature which constitutes the difference must be elusive − “thin” − and that there cannot be anything positive to grasp (just as in the parallel case of Nero). Yet, it is interesting to note, this is not so. According to the way we conceive of that difference, the difference is not elusive or “thin” at all. There is something positive to grasp here when we consider the difference between the two cases. But how is this possible, one might ask, despite the fact that there is no descriptive difference between the two cases (no difference in virtue of which the counterparts are different subjects)? The answer, I think, must be this: the difference can be grasped in the first-personal way described earlier. We can grasp the difference on the basis of non-descriptively taking your perspective. Of course, from the perspective of the perfect counterpart in the first situation and from the perspective of the perfect counterpart in the second situation all is exactly alike. But rigidly referring to you and assuming that in the first 13

The two locutions are sometimes used interchangeably. The distinction made here need not describe common usage.

Consciousness and non-descriptive individuality

173

situation the counterpart is you while in the second it is not, we become clearly aware of a relevant difference. The first situation is a situation where you are living a rich human life, while the second is a situation where you never start experiencing anything. We have a rich conception of that difference. The difference, of course, does not lie in the counterpart’s rich life in the two situations (which is exactly alike). The difference lies in nothing but who is experiencing that life. In order to grasp this difference we must engage in what I called first-personal thought applied to others. Doing so we attain a rich conception of that difference. the constitutional basis of the existence of conscious individuals Considering cases of the kind discussed we can discover facts about the way we think. We can discover what we implicitly take to be essential for being a subject. According to our understanding, subjects are the kind of individuals with respect to which it is adequate to engage in non-descriptive perspective-taking. They are the kind of individuals with respect to which it is adequate to think in a first-personal way. Thinking about subjects in this particular way we can furthermore discover how we think about the constitutional basis of their existence. We discover that, according to our notion of a subject, the constitutional basis of any subject’s existence is non-descriptive. One may summarize the result of that reflection as follows: according to our understanding of what it is to be an experiencing subject, it is essential for being an experiencing subject that one’s individual nature is non-descriptive. This assumption is built into our concept of an experiencing subject. All these observations about our own concept of conscious beings do not imply that conscious beings have a non-descriptive individual nature. It could be, after all, that false assumptions are built into our concepts. But our concept of a conscious being is a special case. It is formed on the basis of being an experiencing subject. On the basis of being an experiencing subject we have access to what it is to be an experiencing subject. This is why, in that particular case, discoveries about our own concept can lead to insights about the referent of the concept, or so I will argue. implicit concepts and philosophical insights My argument will rely on a certain view about concepts and their role in the acquisition of philosophical insights that cannot be developed here. But I have to mention a few elements of that view to render the following

174

martine nida-ru¨ melin

argument more accessible and more convincing. These elements will be briefly presented without any argument in favor of the view they characterize. Many of those concepts that play a central role in our thinking and in the way we perceptually and emotionally experience the world, ourselves and others are only implicit and have no linguistic counterpart. They organize and structure our mental life and they are relevant to the way things cognitively, perceptually and emotionally appear to us. These concepts often incorporate assumptions about the nature of the phenomena (the kind of thing, the property, the relation) they refer to. However, in some cases it is quite difficult to bring those implicit assumptions to the surface, and to make them consciously accessible by finding an adequate expression in language. Having these concepts manifests itself in how things cognitively appear to be (intuitions) and in what we can or cannot positively conceive of. In many cases we must make use of these implicit concepts in any adequate description of how things appear in perception, agency and emotion. Ways to bring those implicit notions and the implicit assumptions about the nature of their referents that they incorporate up to the surface of conscious understanding are, among others, the testing of intuitions with respect to concrete and counterfactual cases and the “method” applied in the present chapter: finding out, by reflection upon counterfactual cases, what we can positively conceive of and in what kind of conceptualizing we are thereby engaged. Another such way to bring them to the surface is to carefully enquire as to the content of emotions, perceptions or the experience of agency. As the examples mentioned will already suggest, the strategies available to us to access those central, fundamental but implicit notions and the assumptions about the nature of the referent they incorporate have little to do with conceptual analysis in the traditional sense. They are far more rich and complex and they do not aim to discover analytical truths (truths in virtue of linguistic meaning). Obviously, this cannot be the aim when implicit concepts which have no counterpart in natural language are at stake. When we discover that a certain assumption about the nature of a kind of thing (about what it takes to belong to that kind, about what is essential for membership in that kind) is incorporated in a central implicit notion, then we cannot in general conclude that the things belonging to the kind the notion refers to actually satisfy that assumption. The implicit notion might be inadequate. This can be so if the reference of the notion is somehow fixed independently of the relevant implicit assumption. In that case it might

Consciousness and non-descriptive individuality

175

be that the members of the kind the notion refers to do not in fact satisfy the condition which is, according to our notion of the kind, essential for membership in that kind. In some cases the implicit assumption about what is essential for membership in the kind referred to fixes the reference to the relevant kind. In this case the assumption cannot turn out to be false, but we may find out that there are no things belonging to that kind. access to the nature of conscious beings on the basis of being a conscious being In the present case the implicit notion at issue is the notion of an experiencing subject or a conscious being. The terms “experiencing subject,” “experiencing being” or “conscious being” are technical terms which have no clear counterpart in natural language. They are introduced as expressions for an implicit notion which plays a central role in our thinking and in the way things appear to us (cognitively, emotionally, perceptually and in agentive experience). We can see, or so I claim, that the assumption that experiencing subjects are perfect individuals is incorporated in our implicit notion of a subject and is thus deeply incorporated in the way we think and experience, and in the way things cognitively and experientially appear to be. One might try to argue that the incorporated assumption at issue plays a crucial role in fixing the reference to the kind referred to (which is the category of conscious individuals). If this is so, then the denial that humans and other animals are perfect individuals implies that they are not conscious beings. This may sound like a reductio ad absurdum but the philosophical opponent will not be impressed. He or she will rather deny that the “inbuilt” assumption about our individual nature plays a significant role in fixing the reference to the kind we are talking about. He might say that the reference is rather fixed by our linguistic practices in the application of mental vocabulary, or something along these lines. It follows that conscious beings might very well not be perfect individuals, although this assumption is incorporated into our notion of what it is to be an experiencing subject. To this one might answer that the illusion the opponent thereby attributes to our thoughts and experiences is so fundamental and omnipresent that it undermines his or her view.14 A more direct way to argue for the claim that conscious beings are perfect individuals will be briefly sketched in what follows. 14

For a version and an elaboration of this kind of argument, see Nida-Rümelin (2006, sections 3.15–3.20 and 5.2–5.3).

176

martine nida-ru¨ melin

Our implicit notion of conscious individuals has its origin in what we are aware of in experience. In any experience we have an implicit awareness of ourselves as the subject of the experience, in any action we have an implicit awareness of ourselves as the one who is active in that action, and in any first-person memory we have an implicit awareness of ourselves as persisting across time. We should, or so I suggest, accept a thesis of revelation with respect to what it is to be an experiencing subject: in a non-conceptual and implicit manner we are aware of what it is to be an experiencing subject by being an experiencing subject; we thus have implicit knowledge of what it is to be an experiencing subject given that we are subjects of experience; the nature of what it is to be an experiencing subject is revealed to us by being a subject of experience. Our implicit notion of what it is to be a conscious individual is based on the special kind of access we have to what it is to be a conscious individual by being a conscious individual. This is the best and most direct access one could possibly have to what it is to be a subject of experience. Our implicit notion of what it is to be a subject of experience is formed on the basis of that particular access. We therefore have a right to suppose that the notion is adequate. A notion is adequate if the assumptions about the nature of the referent that are built into the notion are true. Therefore, if our notion of a subject of experience is adequate, as we should, in my view, suppose, then revealing these assumptions by careful reflection is a way to reveal the nature of the kind the notion refers to. If so, then the ontological claim that conscious beings are perfect individuals is justified by the insight that this assumption is deeply incorporated into our concept of what it is to be an experiencing subject. This argument presupposes a plausible claim which, however, needs to be carefully elaborated and argued for. In order to find out what it is to be an experiencing subject, we need not wait for any scientific discovery: we simply have to uncover what we implicitly knew all along.

part iii

Reconsidering simplicity

chapter 10

Personal identity: a not-so-simple simple view Lynne Rudder Baker

A number of different issues travel under the banner of “the problem of personal identity.” My interests, like those of many other philosophers, are metaphysical. In the first instance, I am not concerned with what is called “narrative identity,” or with how we re-identify a person, or with psychological aspects of personality, or with ascriptions of the word “person.” “The problem of personal identity over time,” as formulated by Harold Noonan, “is the problem of giving an account of the logically necessary and sufficient conditions for a person identified at one time being the same person as a person identified at another” time (Noonan 2003). If, as I believe, you and I are essentially persons, then we are persons at any moment that we exist. (We could not change into non-persons and still exist.) This is not uncontroversial, but I have argued at length for this view of persons elsewhere (Baker 2000, 2007d). In any case, a metaphysical account of personal identity should be tied to an account of the nature of persons, and that is how I shall proceed. what is a simple view of personal identity? Typically, accounts of personal identity over time are classified as being either simple or complex. A simple account takes personal identity to be “an ultimate unanalyzable fact, distinct from everything observable or experienceable that might be evidence for it” (Noonan 2003, p. 16), or a “further fact” that “does not just consist in physical and/or psychological continuity” (Parfit 1984, p. 210; Noonan 2003, p. 16). If personal identity over time is, as simple views hold, unanalyzable, then there are no informative or non-circular necessary and sufficient conditions for a person identified at one time to be the same person as a person identified at another time. Complex views specify informative necessary and sufficient conditions for personal identity over time – for example, persistence of body and brain, psychological continuity, etc. The difference between simple and complex views of personal identity depends on whether they 179

180

lynne rudder baker

are informative or not. Informative conditions of personal identity can be stated without presupposing the existence of persons over time. (I shall say more about informativeness later.) Here are my characterizations of “simple” and “complex” views of personal identity: An account of personal identity is a simple view if and only if the account offers no informative necessary and sufficient conditions for a person identified at one time to be the same person as a person identified at another time. An account of personal identity is a complex view if and only if the account does offer informative necessary and sufficient conditions for a person identified at one time to be the same person as a person identified at another time.

Most mainstream views (e.g. those of Shoemaker, Parfit and Lewis) are complex. Complex views (as far as I know) take personal identity to consist in a relation between items – such as mental states or brains or bodies – construed in sub-personal or non-personal terms. (That is what makes them informative.) By contrast, typically simple views (e.g. those of Foster and Swinburne and sometimes Chisholm) appeal to some kind of immaterialism, according to which a person can exist independently of any body.1 Here I shall offer a simple view, but of a different sort from the views of those who take persons to be immaterial substances. On the one hand, I am not a reductionist about persons: I do not believe that personhood can be understood in sub-personal or non-personal terms, and hence I do not believe that there are non-circular informative necessary and sufficient conditions for a person identified at one time to be the same person as a person identified at a different time. On the other hand, I also reject the view that persons are “separately existing entities” that can exist apart from bodies or that are immaterial minds or souls. I reject both standard views, and describe a third way. a first-personal approach According to my third way, we are fundamentally persons,2 who are necessarily embodied, but we do not necessarily have the bodies that we 1

2

Parfit makes this point by saying that simple viewers hold that a person is “a separately existing entity from brain and body” (Parfit 1984, p. 251). I think that “separately existing entity” is ambiguous: it may mean “not identical to brain and/or body” or it could mean “exists independently of brain and/or body.” On my view, we are persons on the first reading, but not on the second. (I suspect that many philosophers would not distinguish between the two readings, but there certainly is a conceptual distinction.) Throughout this chapter, I am talking about what I call “non-derivative persons.” Your body is a person derivatively as long as it constitutes you; but you are a person essentially and non-derivatively.

Personal identity: a not-so-simple simple view

181

in fact have. I take person to be a primary kind. On my view, every concrete individual in the natural world is of some primary kind or other (Baker 2007d, pp. 33–9). An entity x’s primary kind F is the answer to the question “What, most fundamentally, is x?” Everything is of its primary kind essentially. What makes us the kind of beings we are is not what we are made of; rather, what makes us the kind of beings that we are is a particular conceptual ability: an ability to think about ourselves and what we are doing in the first person, without identifying ourselves by any third-person device – like a name or description or a third-person pronoun. Here is an example to show that the existence of a particular thing does not depend on what it is made of. Artifacts are what they are in virtue of their intended functions (abilities), not in virtue of what they are made of. What makes something a watch is that it is produced in order to tell time. Its primary kind – watch – depends on its intended function, which is both relational and intentional. What it is made of is wholly irrelevant to its being a watch. If it is a wristwatch, there is a constraint on what it is made of: it must fit on a wrist. Similarly, a person may be made of something – silicon, organic material, what have you. But what makes the thing a person is a conceptual ability, not what it is made of. But if the thing is a human person, there is a constraint on what it is made of: it begins existence constituted by a human organism. Just as whatever the watch is made of should support the ability to tell time, so whatever the human person is made of should support the ability to think of oneself as oneself – i.e. a robust first-person perspective.3 Furthermore, just as the intentionality required for there to be manufactured items that can tell time may have come about by natural selection, so too may the mental mechanisms required for there to be natural objects that have first-person perspectives have come about by natural selection. So this account of persons should be congenial to Darwinians. We are essentially embodied, but our bodies can be made of anything as long as they provide the mechanisms that support our person-level activities and states. The relation between us and our bodies is constitution – the same relation that a painting has to the canvas and paint that constitutes it. We are constituted by our bodies, and the bodies that constitute us now are organisms. With enough neural implants and prosthetic limbs, we may come to be constituted by bodies that are partially or wholly non-organic. 3

I am speaking of what I call a “robust first-person perspective.” I am omitting complications about babies and higher non-human animals, which I take to have rudimentary first-person perspectives. For details, see Baker (2007d, ch. 4).

182

lynne rudder baker

Our uniqueness lies in the fact that our persistence conditions – which we have essentially – are first-personal. If someone says, “I wonder how I am going to die,” or “I wish I were a movie star,” her first-person perspective is exemplified – that is, she (that person) exists. Everything else besides persons – animals, artifacts, artworks – has third-personal persistence conditions. Any view that takes a person’s persistence conditions to be biological, or physical, or “somatic” leaves out – must leave out – what is distinctive about persons: the first-person perspective. This is so because, as Darwin emphasized, the animal kingdom is a seamless whole, and the persistence conditions of organisms are third-personal. What is unique about a human person is an essential first-person perspective, which I take to be non-Cartesian. (Indeed, my philosophy of mind is externalist: Baker 2007b, 2007c.) On my view, human persons are emergent – as are organisms and other constituted things.4 Your first-person perspective is a property instance that cannot be divided or duplicated. So, a molecule-for-molecule replica of your body would not have your first-person perspective. Hence, fission problems (mercifully!) do not arise. So, here is my not-so-simple simple view: a person is a being with a firstperson perspective essentially and persists as long as her first-person perspective is exemplified. To allow for the possibility that persons are temporally gappy, I should say: a person exists when and only when her first-person perspective is exemplified. To put this in terms of possible worlds: at any time t and in any possible world w, I exist at t in w if and only if my first-person perspective is exemplified at t in w (Baker 2007a, pp. 237–9).5 This condition for personal identity over time is not informative inasmuch as reference to the person is made in the explicans: “her first-person perspective” or “my first-person perspective.” Hence, this is a simple view. However, there is also a necessary condition for exemplifying a robust firstperson perspective: a person must be constituted by something with mechanisms (e.g. neural mechanisms) adequate to support first-person reference to oneself as oneself.6 4 5

6

The relation between persons and bodies is not a supervenience relation; nor is it a part−whole relation. Constitution – not supervenience, not mereology – is the glue of the universe. Harold Noonan notes that we can take the problem of diachronic personal identity to be “the problem of specifying the relation which body a existing at time t must bear to body b existing at time t0 , if the person occupying a at t is to be identical with the person occupying b at t0 ” (Noonan 2003, p. 94). Since, on my view, human persons are necessarily embodied, my view may be expressed in Noonan’s terms: the person occupying a at t is identical to the person occupying b at t0 if (and only if) the person occupying a at t and the person occupying b at t0 have the same first-person perspective. I think that this is close to what Noonan calls “the unoccupied view” (Noonan 2003, p. 97).

Personal identity: a not-so-simple simple view

183

why not-so-simple? I have called my view “not-so-simple.” Why? Although the view conforms to the characterization of simple accounts that I started with, it differs from other (perhaps more standard) simple views. For example, John Foster uses the term “simple view” to apply to accounts of persons as non-physical subjects: “Jones and the non-physical subject to which the pain is attributed in the philosophically fundamental account are one and the same” (Foster 1991, p. 238). My not-so-simple simple view differs from any simple view that supposes that there are immaterial substances that can exist without bodies in the natural world. The not-so-simple simple view differs from other simple views in other ways as well. Simple viewers who invoke immaterial substances like souls often take personal identity over time to be determinate. That is, for any person P and time t, P either definitely exists at t or definitely does not exist at t. For example, Thomas Reid said that when identity is applied to persons, it “admits not of degrees or of more or less” (quoted in Noonan 2003, p. 16). However, the not-so-simple simple view rejects Reidian determinacy about persons and other ordinary objects. Here is the reason that I reject Reidian determinacy. Because I take human persons to be natural objects, I take them to come into existence gradually. I think this follows from the empirical fact that no concrete object of any kind – a natural object or an artifact – comes into existence instantaneously or goes out of existence instantaneously. (Our solar system took eons to come into existence. Fertilization of an egg by a sperm takes up to twenty-four hours.) Hence, we have good reason to believe that human persons do not come into existence instantaneously; like everything else, they come into existence gradually. But if persons, who are essentially persons, come into existence gradually, there is a time at which their existence is indeterminate. However, as I have argued in The Metaphysics of Everyday Life (Baker 2007d), indeterminate existence (of anything) depends on the determinate existence of that thing. It is indeterminate whether some person P exists at t only if there is some other time, t0 , such that it is determinate that P exists at t0 . Let me explain by an analogy. Suppose that you had a house built. Consider the time t at which the foundation was laid and the frame was in place; did your new house exist at t? The answer depends on what happens subsequently: suppose that your house was completed at t0 , at which time it existed determinately. Since your house existed determinately at t0 , it existed indeterminately at t when the

184

lynne rudder baker

foundation had been laid and the frame was in place. But if a tornado had torn the structure down right after t, it would not have been the case that there had been a house that existed indeterminately at t. That is, its having existed indeterminately at one time depends on its existing determinately at some other time. Indeterminacy, so understood, is metaphysical and not just semantic. However, this construal of indeterminacy does not entail that there is indeterminate identity. For any x and y, if x and y are identical, “they” are determinately identical. If x and y have vague temporal and spatial boundaries, x and y are identical – determinately identical – only if they are both vague in the same ways to the same degree at the same times. If H is indeterminately a house at t, and H* is determinately a house at t0 , then H = H* only if there is a unique x such that x is indeterminately a house at t and x is determinately a house at t0 . Identity is determinate, atemporal and necessary; the existence of temporal beings like houses and persons is not (Baker 2007d, pp. 226–33). Identity is all or nothing; temporal existence – when an entity is coming into existence, or going out of existence – is not. Since the first-person perspective depends on the proper functioning of neural mechanisms, and the proper functioning of neural mechanisms is subject to indeterminacies, there are indeterminacies in the coming into existence and going out of existence of persons. (Even if a person does not go out of existence at death, the earthly body – perhaps gradually – ceases to constitute the person.) However, some philosophers take it to be a merit of the standard versions of the simple view that they are committed to the determinacy of persons (Noonan 2003, p. 16). The merit of determinacy is that it seems to avoid the duplication problem. One version of the duplication problem is generated by the “split-brain” thought experiment, where a person’s brain is split and one half is put into one body (“Lefty”) and the other half is put into another exactly similar body (“Righty”). Both successor persons survive, but on pain of contradiction, they cannot be identical. So, which is the original person, Lefty or Righty? Determinacy about persons seems to assure us that there is an answer. But without determinacy, Lefty has no more claim to being the original person than Righty, and vice versa. Although I have little patience with such thought experiments, if there is a puzzle here, it can be solved by my construal of indeterminacy. There are not just two possible answers, but three possible answers to the question: which is the original person? The answers are either Lefty, Righty or neither, and the not-so-simple simple view is compatible with all three answers. We do not know which is the correct answer, but the not-so-simple simple view implies that there is a fact

Personal identity: a not-so-simple simple view

185

of the matter that depends on whether Lefty or Righty or neither has the original person’s first-person perspective. So, we do not need Reidian determinacy in order to avoid the duplication problem. There is still another way that my not-so-simple simple view differs from standard simple views. For example, Roderick Chisholm and many other philosophers take the word “simple” to mean “having no parts,” where having parts entails conforming to a classical mereological theory of “the part−whole relation.” Thus, Chisholm and others take persons to be simple in that they have no parts.

do persons have parts? Do persons have parts?7 I cannot give a one-word answer to this question, because I think that it is ill-formed. The word “parts” is ambiguous. Persons have what I shall call “ordinary parts,” which are parts of ordinary objects like animals, clocks and people; persons and other ordinary objects do not have what I shall call “mereological parts.” Mereological parts are parts of mereological sums, and governed by axioms of classical mereological theories (Leonard and Goodman 1940). Although I have no objection to mereological theories as formal analyses of a relation called “the partrelation,” I do not believe that mereological theories merely sharpen the ordinary notion of “parts.” Rather, mereological theories introduce a different idea of parts, such that they do not apply to everyday objects or to their ordinary parts unless supplemented by constitution. The relation of mereological parts to what they are parts of is different from the relation of a cow’s tail to the cow that the tail is part of. In classical mereology, “part” is a primitive, and all objects with parts are taken to be sums. What is defined is a “sum” or “fusion”: (S) y is a sum of the xs = df every x is a part of y, and every part of y overlaps some of the xs,

where x overlaps y if x and y share a part. Mereologists believe that all objects with parts are sums.8 But on a three-dimensionalist view of ordinary objects, no instance of (S) is identical to any ordinary object. All that (S) can yield are 7

8

Throughout this discussion, I am construing parts non-derivatively (in the vocabulary of Baker 2007a). Also, I am assuming three-dimensionalism throughout. For a defense of threedimensionalism, see Baker (2007d, pp. 199–217; Baker 2009, pp. 1–14). A variant of my view would reject universal composition, relativize composition to time, construe constitution as a many−one relation, and appeal to plural quantification. Then: the xs have a sum at t if and only if there is a y such that the xs constitute the y at t.

186

lynne rudder baker

sums: sums are aggregates, not ordinary individuals.9 To see that a sum of three-dimensional objects is not an ordinary individual, consider the sum of atoms that compose my lectern. Now look at (S): exactly the same sum of atoms would be scattered about if the lectern were smashed. There is a single sum of (3-D) atoms, but before the smashing there is a lectern and afterwards there is no lectern, but the sum still exists. On a 3-D view, the definition (S) is satisfied no matter how the atoms are arranged. The relation between a particular atom in the lectern and the sum of atoms in the lectern is obviously different from the relation between that atom and the lectern: the sum exists whether the lectern does or not. (3-D mereologists may say that the xs have a sum when arranged lecternwise, but not when scattered about. But arrangements are not in the ontology.) I shall use the term “mereological part” for the term “part” in classical mereology. Consider another kind of example: in classical mereology, any things that exist have a sum.10 So there is no metaphysical difference between the sum of your head and the moon, and the sum of your head, limbs and trunk: your head and the moon compose one sum, and your head, limbs and trunk compose another sum. There is a single relation – composition – and each of the sums is a relatum of an instantiation of that relation. But if your head is related to the sum of your head and the moon in the same way that your head is related to the sum of your head, trunk and limbs, that relation is not the relation between ordinary objects and their parts. At this moment, the relation between your head and the sum of your head, trunk and limbs is quite a bit more intimate than is the relation between your head and the sum of your head and the moon. At this moment, the sum of your head, trunk and limbs constitutes you; but at no time does the sum of your head and the moon constitute anything. Composition (in classical mereology) does not distinguish arbitrary from non-arbitrary sums, but arbitrary sums are not ordinary material objects.11 9 10

11

If mereology is interpreted to apply to all material objects, then it is a false theory. Uninterpreted, it is an abstract calculus that is neither true nor false (non-Euclidean geometries). I agree with Theodore Sider that mereological universalism is true because there cannot be vague composition (Sider 2001). But from determinacy of composition – a relation between sums and their mereological parts – nothing follows about genuine objects. Composition is too weak and promiscuous to support a metaphysics of genuine objects. What can be vague is constitution: which mereological sum constitutes Mt. Everest at t? Although Peter Simons takes mereology to be an important part of ontology, he does not think that mereology alone can tell us “what collections of objects compose others” (Simons 2006, p. 613). My suggestion is that we should give up on composition as a relation in which new individuals come into existence. Sums (e.g. sums of microparticles) are constituters, but I do not take a sum to be identical to a concrete object – ever. New concrete objects come into existence by constitution, not by composition.

Personal identity: a not-so-simple simple view

187

In classical mereology, if S1 and S2 are sums and there is any difference in their mereological parts, then S1 ≠ S2. Three-dimensional classical mereology lacks the resources to have objects that gain or lose parts. I am aware that Peter van Inwagen has proposed a temporal mereology that allows change of parts. But his view entails that many sums that have no parts in common are, nevertheless, one and the same sum (Van Inwagen 2006). For example, on Van Inwagen’s account, sum A (the sum of your cells when you were born) and sum B (the sum of your cells fifty years later) are the same sum even if not a single cell in sum A is also in sum B. But the idea of sums (unlike the idea of persons) is conceptually tied to the idea of parts. Sum A and sum B constitute the same object – you – but they are not the same sum. (They have no parts in common.) To construe sums in such a way that identity of parts is irrelevant to the identity of sums seems to me to rip the word “sum” from the theoretical home that gave it meaning. In any case, I do not believe that the formal properties of “the” partrelation will give us an account of ordinary objects or their ordinary parts (Simons 2006). However, if we appeal to constitution, we can find a bridge from mereological parts to ordinary parts for ordinary objects at times. The following is a necessary condition for being an ordinary part at a time.12 (P) If x is an ordinary part of y at t, then ∃z(x is a proper mereological part of z and z constitutes y at t).

An ordinary part of an ordinary object at t – a chair leg – is a proper mereological part of a sum that constitutes the chair now. Sums have mereological parts, and have them essentially. Now we can see how different sums are from ordinary objects: sums and only sums are composed of mereological parts; no constituted object like you or me has mereological parts. Sums cannot change parts; ordinary objects routinely change parts and are as real as objects can be. So, I think that Chisholm’s view is exactly backward. It is not that genuine objects are constrained by mereological essentialism. Mereological essentialism applies only to mereological objects – i.e. sums defined by (S) – that are not genuine 12

In Baker (2007d, 2008), I offered a definition of ordinary parthood at a time in terms of mereological parthood. But I now think that the definition does not fully capture ordinary parthood. It would allow that an atom in my chair is an ordinary part of my chair. So, I have demoted my “definition” to a necessary condition. Here is an alternative construal. Define “mereological part” and “ordinary part” in terms of a primitive term “part” in a generic sense: x is a mereological part of y = df y is an aggregate and x is an item in y; and x is an ordinary part of y at t = df ∃z(x is a mereological part of z and z constitutes y at t). On this construal, any part is either a mereological part or an ordinary part.

188

lynne rudder baker

objects at all.13 No sum is identical to an ordinary object; rather, sums (of legs and tabletop, say) constitute ordinary objects. The answer to the question “Do persons have parts?” depends on how we are using the term “parts.” If we mean “ordinary parts,” yes, of course persons have parts: by (P), your head is an ordinary part of you in virtue of being a mereological part of a sum that constitutes the body that constitutes you.14 (The sum that constitutes your body is the sum of your head, trunk and limbs.) But persons (and other ordinary objects) do not have proper mereological parts (parts constrained by axioms of any extant mereological theory known to me). Only mereological sums (aggregates) have mereological parts.15 The relation between an ordinary part and an ordinary thing – like a person – is a different relation from the relation between a mereological part and a sum. So, the relation between an ordinary thing and its ordinary parts at t is not composition. So, persons (and other ordinary things) are simples if we mean that they have no mereological parts – parts constrained by a formal theory; but persons are not simples if we mean that persons have no ordinary parts. A not-so-simple simple view of personal identity has a not-so-simple answer to the question “Are persons simples?”

why there are no informative criteria of personal identity The first-personal view is a simple view because it provides no informative criteria of personal identity.16 The not-so-simple simple view provides no informative criteria of personal identity because it takes personhood to be a basic property, and not susceptible to a non-personal or sub-personal account. By a “basic property,” I mean one that either was exemplified from the beginning of the universe, or one whose exemplification is emergent. New kinds of phenomena, objects and properties emerge over time – some by 13

14 15 16

The distinction between ordinary parts and mereological parts also makes clear the distinction between constitution and composition. Composition is a mereological relation; constitution is not. Sums have mereological parts; ordinary objects, including persons, do not. Composition is a (necessary) relation that is exemplified by a sum and its mereological parts; constitution is a (contingent) relation that is exemplified by things of different primary kinds. Constitution explains sameness of parts at t of x and y when x constitutes y at t; hence, constitution should not be defined in terms of sameness of parts On my view, constitution is transitive. Like E. J. Lowe, I take us to be mereological atoms; but, unlike Lowe, I take us to have ordinary parts. Mereology has a different role in my view from the one it has in Lowe’s view. See Lowe (2001). Merricks (1998b, pp. 106–24) argues that there are no criteria of identity over time at all.

Personal identity: a not-so-simple simple view

189

means of natural processes and some by means of human ingenuity. The original property base at the Big Bang is not itself sufficient for all the later emergent items, and hence the emergents are not reducible to the original base. A full account of reality would include the original base together with the emergent items. Likely candidates for being emergents are self-replicating molecules, qualia and persons. If the first-person perspective is an emergent property, then it cannot be understood in non-personal or sub-personal terms, and it belongs in a full account of reality. It is not surprising, then, if I am right, that there are no informative criteria for diachronic personal identity. To be informative, the criterion of “having the same first-person perspective” at different times would have to be specifiable without using the idea of a person. There seems to be no prospect of doing that.17 However, rather than being embarrassed by the absence of informative persistence conditions for persons, we should insist on it. The only alternative is to construe our diachronic personal identity in non-personal terms. So, if you agree with me that we are irreducibly persons, then you too will want to avoid informative persistence conditions.

objections and replies obj. 1: If persons are not identical to animals, how can they be part of the natural order? reply 1: Biologically speaking, I am a Darwinian: I believe that there is important continuity between the most primitive organisms and us, that we have animal natures, and that biology can uncover all there is to know about our animal natures. When our biological ancestors developed robust first-person perspectives (along with grammatically complex first-person sentences), entities of a new kind – persons – came into being.18 My speculation is that grammatically complex first-person language and human persons came into existence together – both in the course of what anthropologists call the period of “cognitive inflation” of human organisms (Mithen 2004, p. 164). So, we can certainly imagine how we – with our ability to conceive of ourselves in the first-person and hence to enjoy inner lives – could have come about in the natural course of things. obj. 2: Sure, persons have abilities not shared by other kinds of beings, but why suppose that the difference between persons and other things is ontological? 17 18

See the detailed discussion of Thomas Metzinger’s view in Baker (2007b). E.g. “I wish that I had more food” or “I believe that I am getting sick” are grammatically complex sentences that indicate a robust first-person perspective.

190

lynne rudder baker

reply 2: Biology can teach us a lot about ourselves, but not everything. I do not believe that biological knowledge suffices for understanding our nature, all things considered. Even if we are the products of natural selection, ontological emergentism opens logical space for there to be more to us than our animal natures.

For one thing, we can exercise our first-person perspectives to do things that conflict with our animal natures. As Steven Pinker, a well-known evolutionary psychologist, writes, “A Darwinian would say that ultimately organisms have only two [goals]: to survive and to reproduce” (Pinker 1977, p. 541). But he also points out that he himself is “voluntarily childless,” and comments, “I am happy to be that way, and if my genes don’t like it, they can go jump in the lake” (Pinker 1977, p. 52). These remarks suggest that Pinker has a first-person notion of himself as something more than his animal nature as revealed by Darwinians. Moreover, we persons have abilities that have no discernible antecedents in any other species. Let me enumerate some of our features that, I submit, cannot be understood as simply extensions of features of non-human animals (Hauser 2009). These uniquely human features are emergent from features that we share with non-human animals: * We share with other species the ability to communicate with conspecifics, but only we human persons can have a fully articulated language with necessity and possibility. Only we worry about the paradox of the heap. * We share with other species the trait of having a perspective on our environments, but only we human persons can have rich inner lives, filled with counterfactuals (“if only. . .”). * We share with other species methods of rational enquiry (Where’s the bone? I saw it being buried over there yesterday. So, I’ll look over there), but only we human persons can deliberate about what to do and can attempt to rank preferences and goals, and try to resolve conflicts among them (and thus be rational agents). * We share with other species activities like self-grooming, but only we human persons can have self-narratives. * We share with other species the ability to make things that we need (e.g. nests), but only we human persons can make things that we do not need (e.g. SUVs). * We share with other species the property of having social organization, but only we human persons have war crimes, international courts and human rights.

Personal identity: a not-so-simple simple view

191

I believe that these are just a few of the uniquely human features that add up to an ontological difference between human beings and animals. All these differences rest on our having robust first-person perspectives. Robust firstperson perspectives bring with them a cascade of new kinds of abilities: we can see ourselves as individuals and as members of communities; we can plan for our futures; we can hold ourselves responsible for what we do; we deceive ourselves; we can write memoirs; we try to reform; we have rich inner lives. And on and on. With respect to the range of what we can do (from trying to control our destinies to daydreaming) and with respect to the moral significance of what we can do (from assessing our goals to confessing our sins), it is obvious that beings with robust first-person perspectives are unique. And the uniqueness is ontological. It follows from the theory of constitution that if x constitutes y at t, then y is on a higher ontological level than x. When philosophers speak of levels, they usually mean levels of description. I do not. I mean levels of reality (Baker 2007d). On the constitution view, sums of molecules constitute people, but sums of molecules are fundamentally different kinds of things from people. So, the difference in level between molecule-talk and people-talk is not just a difference in level of description; it is a difference in what is being talked about. The idea of constitution makes sense of this ontological difference between molecules and people. conclusion First, unlike other simple views, the not-so-simple simple view does not appeal to any immaterial substances in the natural world that can exist independently of bodies. Second, the not-so-simple simple view holds that, although persons are essentially persons, the onset of human personhood is gradual and not instantaneous. Third, the not-so-simple simple view holds that persons have ordinary parts that are not the mereological parts of standard algebraic mereological theories. Fourth, and finally, the not-so-simple simple view holds that what persons fundamentally are cannot be understood in exclusively thirdpersonal terms. It is non-reductive. It does not pretend to understand persons or their identity over time in non-personal or sub-personal terms. To be a person is to have a first-person perspective essentially, and a person continues to exist as long as her first-person perspective is exemplified.19 19

Thanks are due to Phillip Bricker and Sam Cowling for comments on a draft of this chapter.

chapter 11

Is “person” a sortal term? Christian Kanzian

introductory remarks The leading questions of my contribution are (a) whether “person” is a sortal term in a technical sense, and (b) why this is relevant for the discussion of personal identity, especially regarding the opposition between simplicity and complexity. In the first parts of the chapter I discuss (standard) no answers: “person” is not sortal; then yes answers: “person” is a sortal in a strict and technical sense, exactly like “car,” “sheep” or “Homo sapiens”; finally, views which seem to be inconsistent concerning the semantic character of “person.” Then I try to present an alternative: “person” is a semantically unique term. This uniqueness consists in a kind of incompleteness, which I intend to spell out. I discuss this view in the context of a theory of personal identity. There I start with considerations about persons as a kind, and continue by focusing on individual persons. In my center of interest lie specific personal individual forms as the founding instances of personal identity. In conclusion I borrow an argument from the philosophy of mind to strengthen my thesis of personal individual forms as simple units. At the beginning let me bring in a methodological remark on the general topic of our volume: diachronic personal identity, simple or complex? To take something as simple means in general to regard the fact in question as not subject to analysis and thus to a certain extent primitive. Conversely, complexes can be analyzed, at least as being composed of their constituents. It is such an analysis that offers the possibility of reduction: it might be the case that the complex is nothing but the sum of its constituents. Something non-analyzable does not fulfill this (necessary) condition for reducibility. Epistemologically considered, the non-analyzable cannot be explained, at least not in an informative way. The simple is what it is. Such statements of non-explainability sound suspect. The assertions of simplicity, unanalyzability and non-reducibility are often interpreted as 192

Is “person” a sortal term?

193

“not knowing what the fact in question really is” or “being unable to provide a profound, ‘real scientific’ explanation”; in short: “being metaphysical or ideological.” And thus we should stick to a complex view. My purpose is to express my conviction that such a conclusion against simplicity is problematic. That can be made clear if we consider that every complexity is not complex in an absolute sense, but only in a relative one. Someone can take a complex view relative to diachronic personal identity. But the supporter of such a complex view concerning personal identity must say what personal identity consists in. And the explanation expressing such convictions must include references to some kind of: (a) basic entities like ontological “atoms”, (b) relations combining the atoms to complexes, and (c) a procedure for reconstructing personal unities from the bottom up from (a) and (b) elements. This might be perfectly informative. But what about these relations themselves? And – we must not forget – what about the atoms themselves? Even if you try to give informative answers about the nature of the relations, as far as atoms are concerned, the end of analysis is near. Since Locke (see Locke 1975 [1690], ii, xxvii, 3) we have known that atoms either are what they are, or they are not atoms at all. The same problems can be referred to processontological reconstructions of diachronic identity, and to psychological reductions of personal persistence in a Humean–Parfitean style. All of them might start with a complex view, but end as supporters of simple, non-analyzable units, whatever they are. An absolute complex view is impossible. This is why I think that the supposed epistemological advantage of complex views is a relative matter. In other words: simplicity is unavoidable (cf. Benovski 2010). Every view has its unexplainable starting points – its “metaphysics,” if you like.1 This does not decide the complex–simple discussion on personal identity, but deprives the followers of a complex view of one line of argumentation and takes away the burden of proof from the supporters of the simple view. However, the aim of this chapter is not to deal with these fundamental questions concerning the complex–simple debate. My starting point is the semantics of “person,” the question whether it is a sortal term, and how we

1

Even David Lewis seems to accept this point as a methodological unavoidable angle: “A system that takes certain . . . facts as primitive, as unanalyzed, cannot be accused of failing to make a place for them” (Lewis 1997, p. 198).

194

christian kanzian

can understand the relevance of the question for an ontology of personhood and personal identity.

“person” is not a sortal term Let me come to the “no answer” position concerning the status of “person” as a sortal term. How can we understand such a position? Before we seek an answer, we must give at least a provisional account of what a sortal term, technically considered, is. Without being able to discuss alternatives at this occasion, I suggest defining sortal terms such that they: (a) denote those (and only those) species which are called species infimae; (b) provide a maximally precise apparatus for the identification of individuals, not only synchronic, but also diachronic, which means throughout their whole existence (since species infimae determine the identity of individuals at each moment, but also their persistence, in a maximally specific way); (c) are (necessarily and sufficiently) connected with principles for counting and enumerating. Condition (b) excludes terms denoting accidents (“being tall”), as well as so-called “phase sortals” (“caterpillar,” “juvenile”); condition (a) excludes terms for higher species, genera, and categories (“mammal,” “living being,” “substance”), as well as for classes within the species infima (“woman”); and condition (c) also excludes mass terms (“gold,” “orange juice”). Standard no positions urge that “person” lacks one of these semantic characteristics. The most promising way to argue in favor of such a no position is to stick to a (b) procedure: “person” has either an accidental character (radical no answers) or can be classified as phase sortal (weak no answers). The latter has the advantage of allowing one to link “person” with a remarkable identification apparatus, at least for the synchronic identification of some entities. The first (to my knowledge) to whom such an “accidental” understanding was imputed is Gilbert of Poitiers (1080–1155), who primarily intended to oppose Boethius’s substance-view and thus suggested taking personhood as something which is added (Lat.: extrinsecus affixa predicamenta) to the natural constitution of human beings (Gilbert of Poitiers 1966). The Council of Reims (1148) took him to task for this account, on the grounds that it is committed to an interpretation of personhood as accidental. I cannot discuss whether Gilbert really wished to become the ancestor of all (more or less radical) no answers to our question on the sortal character of “person,” but I think he actually is.

Is “person” a sortal term?

195

However, systematically considered, I think that the above-mentioned no answers are completely untenable. Weak no answers offer a sufficient argument against the radical ones: their identification function is essential for “person,” whereas terms for accidents have no identification function. Thus “person” is no accidental general term. Weak positions holding that “person” is a phase sortal term like “juvenile,” “caterpillar” and so forth, assert the identification function of the term in question, but actually restrict it conceptually or a priori to a period of time. This seems problematic, since standard uses of “person” do not provide such a priori restrictions;2 thus I think that “person” cannot be a phase sortal. To concede that “person” is a general term whose identification function is not conceptually restricted to a specific period of time means we must admit that “person” is not a phase sortal at all. The supporter of such a thesis actually gives up the phase sortal theory, and in consequence the no answer position. The problems with standard no answers (accidental, as well as phase sortal accounts) are the best reasons to explore a yes answer position: “Person” should be accepted as a general term standing for a species infima, providing a complete and maximally precise apparatus for the identification of individuals, and as connected with principles for counting and enumerating. “person” is a sortal term Boethius (480–526) can be considered to be the ancestor of this position. Boethius does not speak about general terms in a manner which can be compared with our dealing with sortal terms. Nevertheless, his effort to combine personhood with substantiality establishes persons as a substancekind, precisely to avoid the accidental character of persons; and, primarily, to explicate the Christological dogma of Chalzedon: one substance – one person and two natures within his categorical framework (Boethius 1988). And precisely this ontological view seems to go hand in hand with the thesis of “person” as a sortal term in the technical sense introduced above. The ontological and the semantic thesis are committed to each another.3 2

3

Another argument could be that the phase sortal conception concerning “person” would make “person” useless for theological contexts a priori. But I think it is also problematic to restrict human personhood to a specific period, for conceptual reasons. Here I can neglect the questions whether there are also sortals for the category of events and for the genus of non-substantial things.

196

christian kanzian

Although the substance-character of persons sounds harmless, I fear that it is not. On the contrary, it provides one of the most significant problems of the yes view: “person” is a sortal. Were persons substances in their own right, they would be things “like other things in the closed cosmos of substances” (Lutz-Bachmann 1983, p. 66; original language of the quotation is German, translated by author). Persons would exist just as amebas and sheep do. I have no problems with amebas and sheep – they occur as such, but persons do not. There are no persons simpliciter. What obviously occur in everyday life are human persons, organisms of the species Homo sapiens, who we like to regard as persons. If persons as such were substances, we would have to choose between neglecting the substantial character of organisms (in which case we could defend the view that human persons are one and only one substance: persons) or conceding that we, human persons, consist of two substances.4 I think the first alternative leads very quickly to a dead end, since the denial of the substancehood of organisms opposes all reasonable concepts of substance. The latter alternative would be a rather strong version of dualism. There is no point in rehearsing the problems of substance dualism. Before I start to ask for a third way between standard no positions concerning the sortal character of “person” and the yes view, I want to draw attention to another possible position, which we can call the “don’t care” view. Don’t care views can be accused of inconsistency concerning the logic of “person,” which does not bother them and is what provides reason for calling them “don’t care-ists.”

the don’t care view A paradigmatic “don’t care-ist” is Peter Singer in his Practical Ethics (Singer 1978). On the one hand, he is committed to the accidental character of personhood: according to Singer every (living) being that is able to gain the 4

A third alternative could be to take persons as being (nothing but) organisms of a specific kind. I am not sure if this is compatible with a yes position: “person” is a sortal term in its own right. I rather would understand such a view to be either reductionistic or an accidental view. Obviously compatible with a yes position is a “constitutional view” on persons, as Lynne Rudder Baker presented it in her Persons and Bodies (Baker 2000) and in her The Metaphysics of Everyday Life (Baker 2007d). In the latter she explicitly states: “Person – like statue – is a primary kind, one of many irreducible ontological kinds” (Baker 2007d, p. 67). I concede that Baker’s view does not commit to dualism. But I think that the premises of her constitutional view are themselves so problematic that her theory cannot be regarded as a really successful yes theory. I ask for understanding for not being able to go into details here.

Is “person” a sortal term?

197

status of person can lose it (which happens, say, when it is deprived of the intellectual capacity which is decisive for this status) and perhaps – under lucky circumstances – regain it. This is typical for an accidental account of personhood or a (radical) non-sortal view on “person.” My interpretation is supported by examples given by Singer to point out the irrelevance of possible personhood for the (ethical) status of its bearer. For example: just as Prince Charles has the potential to become the king of England, but is not the king of England now, so a being having the potential to become a person is not a person now, and thus need not be treated as such (Singer 1978, p. 120). Without discussing the example in detail, it seems clear that its explanatory power depends on the assimilation of “person” to “king,” and “king” is obviously a general term in the category of accidental properties. On the other hand we find passages in Singer’s Practical Ethics in which he seems committed to the substantial character of personhood. A person is characterized as a being with the preference to exist in future as the same person (Singer 1978, p. 78). This presupposes that a person must be able to exist (as the same person) through time, and to identify herself as the same person diachronically. The latter would be impossible if “person” were not combined with a sortal-like apparatus for identification. “Person” must be considered as a sortal, to make Singer´s characterization of persons sound. And sortals stand for substance species.5 We can conclude that Singer’s position is inconsistent with respect to the semantic role of his concept “person.”

“person” is a semantically unique term I am sure that we can criticize Singer’s theses in his Practical Ethics by a semantic analysis of his use of the concept “person.” But I do not intend to lead a debate on Singer here. I would prefer to return to my main point and ask for an alternative between standard no positions and the simple yes answer to the question of the sortal character of “person.” How, on the one hand, can we avoid the shortcomings of the no positions, yet on the other hand also keep away from the problems of the yes position? The logical nature of “person” seems to require that we use it to identify something, or rather someone: we pick out units from the manifold empirical impressions by marking them as persons. We mean to have reasons for 5

Since persons have the preference to exist in future as the same without conceptual or a priori restriction to a specific period of time, “person” must really be considered as a sortal; the phase sortal account of personhood cannot protect Singer from inconsistency.

198

christian kanzian

assuming that the units are real units, not only at the actual moment of identification, but also through time. These reasons are just that the units fall under the concept “person.” Furthermore, we use “person” to count these units, which we are able to identify and re-identify by applying the term. What could be wrong with asserting that there are thirty-two persons in a room? Nothing, I concede, but we should nevertheless be careful not to jump to the conclusion that “person” is a sortal term in a strict and technical sense. In order to make my caution plausible, let us examine in detail the identification function of terms in general, and of our term in particular. What in general is the reason why we can use an F-term for the identification of x’s? It is because we have knowledge of some criteria which we connect with the F-term and which provide reasons for assuming that something is an F-unit, both at one particular moment and through time.6 For example: why are we able to identify a living being with “sheep”? Because “sheep” provides clear and informative criteria for the decision why the tail, for instance, belongs to the sheep but that the grass on which it is lying does not; why there are two, and not only one sheep in the pen, and why we should accept that it is the same sheep there yesterday and today: maybe it is the sum of specific, sheep-like life functions, in which the parts of the organisms in question are integrated. What about “person”? Is “person” connected with informative criteria for the identification of units falling under it, like “sheep”? I do not think so.7 “Person” is not connected with corporal criteria for the identification of beings. This becomes especially clear if we consider reidentification: i.e. identification through time. We can change corporal parts without ceasing to be one and the same person. Nor are biological criteria satisfactory. Without being able to go into the details here, I want to recall John Locke’s famous thought experiment of the body change of two persons (Locke 1975 [1690], ii, xxvii, 27, 15), which can strengthen our antibiologistic intuitions. Psychological criteria are problematic, because every person can change her psychological character radically. Moreover, it is hard to see which psychological criteria we should employ and which we should not. Is there an isolated psychological criterion? I do not think so. That neither physics, biology nor psychology can solve the problem of identification criteria for “person” is due to the impossibility of translating 6 7

I am of the opinion that we cannot identify something without taking it to be an F-unit. Without sort, there is no unit. But this is not at stake here. I borrow this view from Lowe (1989b, ch. 7.3 and 2003). Other parts of my theory (the incompleteness of “person”) distinguish me from him.

Is “person” a sortal term?

199

“person” into a pure physical or physiological, biological or psychological concept. If “person” were, for instance, a biological concept, biology would be able to teach us sufficiently what units persons are – but this is not the case. But, because we have no empirical criteria connected with “person,” I think we must give up the idea of informative criteria for identification connected with this general term. We cannot explain why we identify units with “person.” Nevertheless we do actually identify units with “person,” and we have the strong intuition that there should be reasons why we do so. How can we resolve this seemingly paradoxical situation? I think the answer is that we (can only) use “person” for the identification of units, because we combine “person” with another general term, which indeed provides us with clear and informative criteria for the identification of entities. One of the best candidates for such a term is “human being” or “Homo sapiens,” which stands as a general term for a biological species.8 “Human being” is a sortal term in a strict and technical sense. It provides an apparatus for the identification of individuals, not only synchronic, but also diachronic, throughout their whole existence, and is connected with principles for counting and enumerating. Thus, what we do when we identify and count persons is to identify and to count human persons. This sounds unspectacular, but it is not, semantically considered: it makes clear that “person” is an incomplete general term, with respect to its identification function. Does that mean that we fall back into one of the standard no answer positions, taking “person” as an accidental general term or as a phase sortal? No, because we stick to the identification capacity of “person,” which is not conceptually restricted to a specific period, but add that “person” per se is not connected with informative criteria for the identification of anything. “Person” depends on the identification apparatus of another general term, of a real sortal term. If we take the informativeness of such criteria as necessary for a term to be a count term, it becomes clear that “person” as such cannot fulfill condition (c) for sortal terms, which also distinguishes our theory from yes positions. I want to conclude that “person” should be understood as a general term with a unique semantic status: it is an incomplete term, which means that it depends on other terms to yield a function to which its “logical nature” is dedicated: to identify something without conceptual or a priori temporal 8

NB: I do not pretend that there are logical reasons why “human being” should be the only candidate.

200

christian kanzian

restrictions. All wrong accounts in the semantics of “person” start by neglecting this unique status and by putting “person” with other terms into the same logical category of general terms. In the next section I shall apply the semantic idea of the incompleteness of “person” to the kind of persons, and finally to an analysis of the “individual form” of persons.

the kind of persons The first step can be taken with an enquiry into whether the kind person is a species infima in the sense introduced above: a species that determines the identity of the individuals belonging to it in a very specific way, at each moment, but also throughout the whole time of their existence. If my considerations concerning the term “person” are adequate, the answer must be “no,” if only for conceptual reasons: I defined species infimae as exactly those kinds for which sortal terms stand; “person” is not a sortal; thus person cannot be a species infima. I hold this to be true, but nevertheless, I confess it sounds tricky, and needs further argumentation. To meet this demand, let me explain in more detail just what the determination of the identity of an individual by its kind consists in (i), whether the kind person can provide such a determining function or not (ii), and (iii) what we can conclude from (i) and (ii) concerning the ontological status of the kind person. (i) The first aspect of the sortal determination of an entity is basic. I admit that it is easier to grasp if you are guided by some Aristotelian premises in your ontology: according to the Aristotelian tradition no entity (especially no substance, and we may restrict our considerations to this category here) can exist without a specific individual form, without a means by which (a “how”) its individual material is construed to make up the entity in question. No specific individual form can be imagined without being specified by a species infima; thus no entity, especially not in the category of substances, can exist without a species infima. Beneath this basic or existential aspect, we can state more concrete aspects of the determination of an individual by its kind: so entities are determined by their species infima in their continuity, spatially as well as temporally. That Susan is a sheep determines her spatial extension and the peculiar characteristics of it. The same holds for her temporal career: her persistence conditions are determined by her sheephood: general ones (e.g. that her diachronic identity cannot be interrupted) as well as special ones,

Is “person” a sortal term?

201

pertaining to the specific principles by which sheep-like organisms develop (Hirsch 1982, pp. 37–8; Oderberg 1993, pp. 22–3). Another concrete aspect of sortal determination is mereological and material composition: which parts normally or essentially belong to a unity, of which matter these parts can or must be composed. All that is determined by the kind to which the entity belongs. (ii) However, the decisive question for our topic is whether the kind person can provide such a full identity-determining function. Attempting to answer this question lands us again in a paradoxical situation. There are strong intuitions, as well as ontological reasons, supporting the view that the identity of a person is determined by her being a person. If we take the ontological status of personhood seriously, we cannot give up the idea that it has a determining influence on the identity of those individuals that belong to it. Personhood is not accidental. The ontological nature of the kind person seems to determine the identity of “its” individuals. But, on the other hand (and this yields the paradoxical situation), we cannot explain how the kind person determines the identity of persons, as it is demanded of a species infima. The first reason may be that, to determine the identity of an entity in a maximally specific way, we need informative conditions for the identity of that entity. But the kind person does not contain any informative conditions for how the entity’s spatial and temporal career or its mereological and material composition are determined. For the kind “person” is not connected with material or biological features, or with other candidates for informative conditions. The argument for this can be borrowed from our considerations concerning the general concept “person.” (iii) The only way out of this paradoxical situation, as far as I can see, is to assume that the kind person depends on other kinds, such as Homo sapiens, to get its “natural” identity-determining function. The kind person is dependent, and thus incomplete, and moreover unique in this incompleteness. Since only complete kinds can be classified as species infimae, the kind person cannot be one. This view is supported by the assumption that there are no entities that are purely and simply persons (which is a rejection of extreme dualism or of spiritism). The identities of entities that are persons are not completely (but incompletely, which means in dependence on another kind) determined by their personhood. What we do when we provide conditions for the existence, continuity and composition of persons is to appeal to those conditions that are connected with the species infima to which persons belong: e.g. humans. We determine the identity of persons by determining their identity as human persons.

202

christian kanzian

This point of view is not dualistic, but is not its price reductionism concerning the kind person? No: that a kind F depends in a central way on a kind G does not imply that F is reducible to G, since there might remain functions at F that are not covered by G. Actually my argumentation (to say it cautiously) excludes neither such functions nor their possible relevance for an ontology of human persons. Does this approach amount to “sortal relativity” concerning the identity of human persons? No: sortal relativity means that the identity of an entity is determined by two ontologically equivalent kinds or sorts, which is not the case with the kinds of persons and of humans determining the identity of human persons. Although there might be other remarks worth considering, I want to come to my next point, which is an analysis of individual persons, especially the individual form of persons. the individual form of persons My result will be that the personal individual form is a simple and incomplete unit, which depends in its function as form on other individual forms. But let me go step be step. From an explanation of what individual forms in general are and how we can grasp their ontological functions, I will proceed to a characterization of personal individual forms in particular, and argue for the above-mentioned thesis. So, what is an individual form? Recall from the previous section that we can understand the form of a substance as the how aspect of its ontological constitution. The need for such a how aspect becomes clear when we consider that there is no thing (in a technical sense) that is not formed matter. (Unstructured material parcels are not things in this sense. The first may be the material aspect of the latter.) And the individual form is the concrete formation of the material composing a thing. The concrete way in which Susan’s organism is arranged, structured and related to her environment, its possibilities of developing and reproducing, is her individual form. That the individual form of a substance is determined by its kind was also mentioned in the previous section. The concrete arrangement of Susan’s organism is given by her being a sheep. Not explicitly mentioned before, but easy to deduce from the assumptions made so far, is that the individual form is the instance of a substance which founds that substance’s identity. Susan’s identity is not founded, neither synchronically nor diachronically considered, on a specific material part or on the sum of her parts, but on the life functions which are given by her

Is “person” a sortal term?

203

individual form.9 That her tail belongs to her but that the grass on which she is standing does not is granted, because her tail, in contrast to the grass, is integrated into her life functions. That she is the same sheep as she was yesterday is not founded on the sameness of her cells, but on the continuity of life, for which her individual form is the ontological principle, so to speak. An individual form is the identity-founding how aspect of a substance, which is determined by the kind to which the substance belongs. How can we apply these general features to the specific personal individual form? What characteristics must an individual form have, determined by a kind, which is incomplete and thus dependent on other kinds in ontologically significant respects, and which is not connected with informative conditions for the existence, continuity and composition of individuals belonging to its scope? I think the only possible answer is that these individual forms must also not be informatively explainable and analyzable, i.e. simple in their ontological constitution. The same holds for personal identity insofar as it is founded in a simple individual form: it is not complex, which means that it is not analyzable into other relations, like continuity (for instance of life functions). And the personal individual form must be considered as dependent on another form in its ontological nature. As the kind person cannot carry out the ontological functions which are essential to kinds without other kinds (real species infimae), the personal individual form cannot fulfill the function of a form without other individual forms: for instance, an organic one. This becomes clear when we consider that the (primary) ontological function of a form is the composition of a whole substance, which means the integration of its material aspect into a substantial unity. The integration of complex matter cannot be done by a simple and primitive form, but only by a complex one, paradigmatically by an organic form. We can conclude that the personal individual form is a simple and incomplete unit, unique in this incompleteness. Human persons own two individual forms, one organic and one personal. “Diachronic personal identity: simple or complex?” is the leading question of our volume. It must be simple, if my considerations are sound; but dependent in a unique way on something which is not simple: for example, on human identity. It is incomplete in nature, because of its essential dependence on something complex.10

9 10

The thesis that life is the principle of the identity of organisms finds support from Locke’s famous Chapter xxvii in Locke (1975) up to Van Inwagen’s Material Beings (1990b). For more details of this thesis see Kanzian (2009, part iii).

204

christian kanzian the unity-of-consciousness argument

Individual personal forms are simple and incomplete. I have tried to find semantic and also ontological arguments for my thesis, which primarily refer to an analysis of kinds in general and of the kind person in particular. In the last section of my article I want to point out that we can also find additional support for my assumptions from considerations in the philosophy of mind, especially of consciousness. Before I come to that, let me remind the reader of my remarks at the beginning. There I argued that the supposed epistemological advantage of complex views is a relative matter. Simplicity is unavoidable. I conceded that this does not decide the complex–simple discussion on personal identity, but takes away the burden of proof from the supporters of the simple view. I think that is important, but does not release us from the necessity of looking for further direct reasons for accepting the simplicity of personal individual forms. The so-called unity-of-consciousness argument seems to be such an additional argumentative support. I do not want to bring up the manifold classical versions of the argument,11 but restrict myself to one contemporary version: that of William Hasker. I refer to his book The Emergent Self (Hasker 2001), not because I think that all of Hasker’s theses are adequate (“emergentism” itself seems to be more an explanandum than a solution of any problem), but because in one of his preliminary chapters he makes the decisive point discussable in the terminological context of the present simple–complex debate on personal identity – which is the only issue here. I start with a quotation of a passage in Hasker’s book: The point is simply that . . . awareness . . . is essentially unitary, and it makes no sense to suggest that it may be “parcelled out” to entities each of which does not have the awareness. A person’s being aware of a complex fact cannot consist of parts of the person being aware of the fact. A conjunction of partial awareness does not add up to a total awareness. (Hasker 2001, p. 128)

How can we understand Hasker’s point? Hasker starts by considering a visual field. According to standard biologistic interpretations of such a state of consciousness (as well as others like perceptions and the content of thoughts in general), the subject or the bearer of it is something bodily: 11

Descartes, Sixth Meditation; but most importantly Kant, Critique of Pure Reason, A 652. Peter van Inwagen has an interesting modern version of the argument (1990b, p. 118).

Is “person” a sortal term?

205

the brain. The brain is aware. The brain perceives. The brain thinks. The decisive point for the present discussion is that something complex is aware, perceives, thinks. Hasker is arguing against such a point of view: let us imagine the smallest part of the brain (Hasker calls it “V”), which contains the modeling of all the information, for example, of the visual field. Is it possible to say that this smallest part is aware of the complex visual field? Perhaps so, but we need to consider the composition of V . . . we can say that V is a whole composed of physical parts. Many of these parts model information from various parts of the visual field. But no proper part of V models all of this information, so it is not possible for any of these parts to be aware of the entire visual field. But if V is a whole composed of parts each of which is not aware of the visual field, how can V itself be aware of it? [Hasker illustrates his points:] . . . this would be like saying that each student in a class knows the answer to one question on an examination, and that in virtue of this the entire class knows the material perfectly! It is true that the members of the class are able, working together, to reproduce all of the information, but there may in fact be no one at all who knows or is aware of all of it. (Hasker 2001, p. 128)

According to Hasker we may assert that it is false to say that the brain or a part of it is aware of something, perceives and thinks. States of awareness (or, more clearly, states of consciousness) are essentially unitary. Something essentially unitary cannot inhere in something non-unitary or in something complex, as the brain is. Consciousness – I would restrict myself to saying human consciousness – must be grounded in something non-complex: that means in a simple unity. The result of Hasker’s analysis is that the bearer of our states of consciousness must be a simple unanalyzable unit. If that is the case, which is what I actually assume, we have the alternative of taking whole substances as such simple units, which would lead us to a rather strong version of dualism, or – and this is what I would suggest – of sticking to individual forms as the bearer of specific personal acts of consciousness. The unity-of-consciousness argument as such does not commit to substance dualism. Thus, it can be taken as the supporter of my thesis of simple personal individual forms. Perhaps we can formulate this in a stronger way: if the unity-ofconsciousness argument is sound, and we want to avoid substance dualism, we have to accept simple individual forms as the only possible bearer of specific personal intellectual capacities. I think this is good news for friends of the simple view, even if personal identity remains – let us not forget – essentially incomplete.

chapter 12

Materialism, dualism, and “simple” theories of personal identity Dean Zimmerman

“complex” and “simple” theories of personal identity Derek Parfit introduced “the complex view” and “the simple view” as names for contrasting theories about the nature of personal identity. He detects a “reductionist tradition,” typified by Hume and Locke, and continuing in such twentieth-century philosophers as Grice, Ayer, Quinton, Mackie, John Perry, David Lewis and Parfit himself. According to the reductionists, “the fact of personal identity over time just consists in the holding of certain other facts. It consists in various kinds of psychological continuity, of memory, character, intention, and the like, which in turn rest upon bodily continuity.” The complex view comprises “[t]he central claims of the reductionist tradition” (Parfit 1982, p. 227). The complex view about the nature of personal identity is a forerunner to what he later calls “reductionism.” A reductionist is anyone who believes (1) that the fact of a person’s identity over time just consists in the holding of certain more particular facts, and (2) that these facts can be described without either presupposing the identity of this person, or explicitly claiming that the experiences in this person’s life are had by this person, or even explicitly claiming that this person exists. (Parfit 1984, p. 210)

Take the fact that someone remembers that she, herself, witnessed a certain event at an earlier time. When described in those terms, it “presupposes” or “explicitly claims” that the same person is involved in both the episode of witnessing and of remembering. Purging the psychological facts of all those that immediately imply the cross-temporal identity of a person will leave plenty of grist for the mills of psychological theories of persistence conditions. Although veridically remembering may be ruled out, closely related psychological states – and causal connections between them – can be 206

Materialism, dualism, and “simple” theories

207

included. As has frequently been noted, reductionists can replace remembering with a relation of “quasi-remembering” (Parfit 1984, pp. 220–2; Shoemaker 1984, pp. 82–6; see also Shoemaker 1970). Personal identity is not explicitly invoked by describing the experiencing of a certain event by a person at one time, and the occurrence of an apparent memory of just such an event by a person later on, which recollection is caused, indirectly, by the original experience in virtue of the normal processes that go on in our brains – whatever those are, exactly. I follow Parfit in calling psychological facts that can be so described “impersonal.” Questions loom about the identity conditions for facts, and about just how fully an impersonal description must characterize a fact in order for that fact to qualify as describable without presupposing or explicitly making claims about personal identity. But I shall not pursue these questions here. The complex view favored by reductionists is contrasted with the simple view of an opposing “non-reductionist tradition.” According to nonreductionists, “personal identity does not just consist in these [psychological and physical] continuities, but is a quite separate ‘further fact’” (Parfit 1982, p. 227; see also Parfit 1984, p. 210). To be a non-reductionist, one must “reject either or both of ” (1) and (2) (Parfit 1984, p. 210), though they arguably stand or fall together.1 According to Parfit, non-reductionists come in two varieties. Some maintain that “[a] person is a separately existing entity, distinct from his brain and body, and his experiences,” such as “a Cartesian Pure Ego, or spiritual substance,” or perhaps “a separately existing physical entity, of a kind that is not yet recognized in the theories of contemporary physics.” Other non-reductionists accept a “further fact view”: they hold that, “though we are not separately existing entities, personal identity is a further fact, which does not just consist in physical and/or psychological continuity” (Parfit 1984, p. 210). The non-reductionists include, he says, “the great majority of those who think about the question” (Parfit 1982, p. 227); but this claim stands in need of qualification. He mentions “Butler, Reid, Chisholm, Geach, Swinburne” (Parfit 1982, p. 227), but one wonders who else he has in mind. I should have thought that the dominant philosophical views about persons would all satisfy Parfit’s definition of reductionism – at least, so long as attention is confined to English-speaking philosophers publishing 1

(2) entails (1), since (2) presupposes the existence of the “more particular facts” in which personal identity is said to consist, according to (1). And, as Marco Dees has pointed out to me, the relation of partially consisting in should be transitive and irreflexive, in which case (1) entails (2) as well.

208

dean zimmerman

articles or books since, say, the 1930s. Very few of these philosophers believe in anything like “a Cartesian Pure Ego” or “spiritual substance.” Almost all are materialists, committed to the view that the identity of persons, over time, must be determined by facts about psychological or physical continuities with no need to appeal to souls or any new kind of physical entities “not yet recognized in . . . contemporary physics.” They may disagree about whether biological or psychological continuities are the basis for the persistence of persons, but they typically attempt to formulate persistence conditions in their favored terms that leave no room for sameness of persons as a “further fact.” A closer look at his definition of reductionism can make one wonder whether anyone really qualifies as a non-reductionist. Would someone who believed persons are immaterial substances, or a new kind of physical particle, really deny reductionism, given Parfit’s official definition? Well, can an immaterial substance or a new kind of physical particle, and the conditions under which they persist over time, be described without “presupposing” or “explicitly claiming” anything about the identity of persons? At least some proponents of views along these lines might think so. Chisholm is one of Parfit’s paradigmatic non-reductionists, identified as a defender of the simple view. The only hypotheses about the nature of human persons that Chisholm took seriously were these two: (i) that each of us is identical with “something of a microscopic nature, and presumably something located within the brain” (Chisholm 1986, p. 75); and (ii) that each of us is “a monad” – an immaterial, indivisible substance (Chisholm 1986, pp. 73, 77). Parfit explicitly consigns both sorts of hypothesis to the category of non-reductionism. But it is far from clear that either one is inconsistent with (1) and (2). By Chisholm’s lights, the proper way for me to determine the nature of my persistence conditions is for me first to try to figure out what kind of thing actually has the psychological properties with which I am directly acquainted. Is it a special kind of physical particle or bit of matter? Then I am that particle or piece of matter, and I have the persistence conditions appropriate to particles or pieces of that sort. Is the bearer of my psychological properties an immaterial substance? Then I am an immaterial substance, with whatever persistence conditions are appropriate to such things.2 If there are facts – even brute facts – about the identity over time of bits of 2

Chisholm detects the same approach to personal identity in Bayle’s remarks about the rational selfconcern of a group of atoms that radically change their form – remarks he cites with approval: see Chisholm (1976, p. 113).

Materialism, dualism, and “simple” theories

209

immaterial substance or of some special physical particles that can have psychological properties, then the discovery that I am a simple immaterial substance or a special physical particle should lead me to conclude – with Parfit’s reductionists – that my “identity over time just consists in the holding of more particular facts,” ones that can be described without “explicitly” making claims about identity of persons. To further probe this question, I turn to some theses about the relations between persistence and immanent causation. Chisholm was convinced that, whether we turn out to be physical or immaterial, we must be fundamental substances – not mere “modes” of something else, or “entia successiva” constituted by different things at different times (Chisholm 1986, pp. 66–71; 1976, pp. 104–8). For fundamental substances (ones that are not made out of, or dependent upon, some other kind or kinds of thing), persistence through time plausibly requires the sort of causal dependence of later stages upon earlier stages that is often called immanent causation. Elsewhere, I offer a theory about what such causal dependence should look like. The theory is based upon the idea that the earlier intrinsic features of a thing must be causally relevant to its later intrinsic features (though not in just any old way) (Zimmerman 1997). Some philosophers – call them “immanent-causal reductionists” – will think that, for fundamental things at least, the facts about persistence consist in nothing more than the facts about immanent-causal connections between stages of the things – i.e. that facts of the first kind are completely determined by, or supervene upon, facts of the second kind. (As Shoemaker points out, one can talk of “stages” in this context while remaining neutral with respect to the doctrine of temporal parts – though many immanentcausal reductionists will believe in temporal parts (Shoemaker 1984, p. 75).) The immanent-causal reductionist will have to say that, whenever there are forking paths of immanent-causally connected stages, one must give the same verdict about whether the original thing persists through the apparent fission. There are choices about what the verdict should be: for example, a fundamental thing ceases to be, replaced by two new ones; or there were two things all along; or a thing becomes multiply located; or . . . But a uniform answer to all cases of branching paths is required by the immanent-causal reductionist’s commitment to the idea that persistence, for these fundamental things, is fully determined by the facts about paths of immanent causation. (Similar choices must be made in the case of converging immanent-causal paths.) Others who think immanent-causal continuity is necessary for the persistence of fundamental things may nevertheless suppose that, at least in

210

dean zimmerman

cases of branching and convergence (and probably elsewhere, too) there are facts about the persistence of fundamental things that are not fully determined by the facts about immanent-causal dependencies. (Chisholm himself would likely have been in this camp.) In a case of branching paths, there are (at least) three genuine possibilities: the original thing ceased to be, or it went to the right while a new thing appeared on the left, or it went to the left while a new thing appeared on the right. (Some will want to allow for a further possibility: the thing became multiply located.) These philosophers posit a further fact about identity, at least for fundamental things – one that is not fully determined by immanent-causal dependencies. Now, where, within Parfit’s taxonomy, should one place an immanentcausal reductionist who believes, with Chisholm, that persons are a special kind of fundamental entity – whether physical or immaterial (however exactly that distinction is to be made out)? Parfit clearly allows that one may make use of psychological terms, when describing the facts in which personal identity is supposed to consist, without violating clause (2) of reductionism. So the immanent-causal reductionist could say: (a) a person is a thing of any fundamental kind that, at some time or another, has psychological properties, or perhaps simply could have psychological properties; and (b) a particular person’s persistence conditions just consist in the holding of immanent-causal relations between stages of the kind of fundamental entity he or she (or it) happens to be. This philosopher seems pretty clearly to accept (1) and (2), and therefore to be a reductionist. But she also seems to satisfy Parfit’s example of a paradigmatic non-reductionist: she is a believer in a Cartesian Ego or special physical substance, a “separately existing entity.” What should one say about a person who believes that we are fundamentally immaterial or physical substances while denying immanent-causal reductionism, maintaining instead that there is a “further fact” about the identity over time of such things? This person seems closer to rejecting Parfit’s (1) (and thereby denying (2)). Still, suppose this philosopher believes in further facts about identity for all fundamental things, whether or not they are capable of having psychological properties. She might very well be willing to state very general necessary conditions for the persistence of fundamental things, physical or not – namely, continuity of immanentcausal paths – but then she will say that necessary and sufficient conditions will not be forthcoming, because there remains the further question whether the very same fundamental thing is present all along the path. She regards identity over time, for fundamental things, as a kind of “brute fact” above and beyond immanent-causal dependencies. But suppose she is

Materialism, dualism, and “simple” theories

211

even-handed about the need to posit brute facts, not treating persons as in any way special within the category of fundamental things. She gives very general, though not very informative, persistence conditions for fundamental things of all kinds, ones that do not make use of terms that “presuppose” personal identity or “explicitly” invoke sameness of persons over time; and the persistence conditions for persons are subsumed as a special case. Should she really be said to deny (1) and (2)? Parfit writes as though no version of reductionism, in his sense, could be consistent with our being “separately existing entities, apart from our brains and bodies, and various interrelated physical and mental events” (Parfit 1984, p. 216). It is not clear to me, however, that Parfit’s arguments against non-reductionism must count against views like Chisholm’s, if these theories about personal identity satisfy (1) and (2) – as they would when conjoined with immanent-causal reductionism, for example. The persistence conditions assigned to a Cartesian Ego or a “separately existing” new physical particle by an immanent-causal reductionist may well consist in causal continuities that could, in principle, divide into equally viable streams. If these circumstances require a “no-branching” clause among the facts in which their persistence consists, the resulting views would confront the strongest of Parfit’s arguments for the conclusion that personal identity is “not what matters” (Parfit 1984, pp. 261–6). They may well be no less (and, for that matter, no more) vulnerable to Parfit’s line of criticism than the more usual forms of reductionism that he considers. Be that as it may, Parfit’s more informal characterizations of reductionism go beyond the idea that personal identity over time does not depend upon further “impersonal” facts; he insists that the facts upon which identity over time depends do not include any “spooky” extra facts about souls or special kinds of matter, but only facts about the mundane stuff our bodies have in common with all sorts of physical objects. The first part of Parfit’s definition of reductionism seems effectively to be replaced by: (1*) that the fact of a person’s identity over time just consists in facts about ordinary material stuff – the kinds of matter our bodies share with non-living, non-sentient things. I shall understand reductionism, or the complex view, as the combination of (1*) with Parfit’s (2) as a further condition upon the nature of the facts in which personal identity consists – it is needed to rule out a kind of further fact view to be discussed below.

212

dean zimmerman

I now articulate a reasonably precise thesis that should follow from reductionism: namely, the supervenience of personal identity upon microphysical facts, or upon microphysical facts supplemented only by “impersonal” psychological facts.

two doctrines of the supervenience of personal identity upon the impersonal Philosophers are more reluctant than they used to be to propose that something is “analytic.” Even those who accept an analytic–synthetic distinction will often admit that examples of truly compelling conceptual analyses of interesting notions are thin on the ground. But most of us go on giving philosophical theories (of knowledge, right and wrong, objective similarity, causation, etc.), and Parfit is no exception. Although we might not attribute analyticity to the core tenets of our theories, we do generally take ourselves to be searching for necessary truths that, in some sense, explain the phenomena with which we are concerned. The ethicists (many of them, at any rate) want to know what makes something right or wrong, good or bad – what do these normative properties consist in; in virtue of what is an act right or wrong, or a state of affairs good or bad? The search for the grounds for theoretically interesting truths continues in all fields of philosophy, though less frequently under the banner of “conceptual analysis.” Parfit prefers to talk of one fact “just consisting in” other facts. This terminology is, I take it, in the same line of work as talk of the “grounds” of a certain phenomenon; or of what it is “in virtue of which” something is the case.3 When a fact is grounded in facts too complicated for us to fully grasp, we may not be able to offer anything like necessary and sufficient conditions for the obtaining of that sort of fact in terms of (what we take to be) its ultimate grounds. We may nevertheless manage to indicate the kind of facts upon which we think it must depend. And where there are several rival philosophical theories about some phenomenon, they may nonetheless agree about the kind of facts that, in one way or another, ought to provide the grounds for the phenomenon. Parfit’s reductionism (or, equivalently, the complex view) is a claim of this sort: whatever the truth is about the conditions that are necessary and sufficient for a person to exist at more than one time, they must ultimately be grounded in facts about ordinary material 3

For discussion of various ways to understand a metaphysical “grounding” or “in virtue of” relation, see Fine (1994, 2001) and Schaffer (2009).

Materialism, dualism, and “simple” theories

213

stuff, and facts that can be described in “impersonal” ways – ways that do not immediately involve the concept of sameness of person. When a philosopher claims that facts about the holding of a certain relation just consist in facts about the holding of some other relations, he or she is committed to a “supervenience thesis”: the pattern in which the former relation is exemplified cannot “float free from” the patterns of exemplification of the latter relations; there could not be two worlds just alike with respect to the latter but differing with respect to the former. “Just consisting in” does not, itself, just consist in supervenience. But it entails it. Reductionism, with (1*) in place of (1), entails the supervenience of personal identity over time upon the exemplification of certain patterns of properties and relations among bits of ordinary matter. I shall take a moment to spell out two supervenience theses in the neighborhood, one of which will appeal to the physicalistically inclined reductionist; the other is the kind of supervenience that falls most directly out of reductionism. I begin with some definitions and simplifying assumptions. Reductionism includes the claim that no “separately existing entity,” in particular no kind of matter “not yet recognized in the theories of contemporary physics,” need be introduced in order to have adequate grounding for facts about persons. New kinds of particles are predicted and discovered all the time, so there is no point trying to make this clause more precise by means of a list. But the atoms in our bodies are made out of particles that can be rearranged to make non-sentient non-persons; and they are subject to a limited number of forces – gravity, electromagnetism, strong and weak nuclear forces – and these forces obey laws that, so far as we know, take no interest in whether the universe contains sentient beings. One kind of reductionism would be the claim that personal identity over time is fully determined by or grounded in the spatio-temporal relations among the particles of these types (and particles of other kinds, so long as they do not occur only in persons), plus the interactions among these particles that are governed by whatever impersonal forces show up in “final physics.” This would be a kind of physicalist reductionism, a view that takes a physicalist gamble: that the introduction of consciousness will not require new laws or new forces, let alone new kinds of matter or immaterial souls. To state a reductionistic claim about the identity over time of particular persons, rather than simply affirming a global kind of physicalism, a physicalist reductionist of Parfit’s stripe must say: the fact that some person persisted from one time to another just consists in facts about sets of these kinds of particles existing throughout that period, and their being arranged in a certain way and interacting by means of these kinds of fundamentally

214

dean zimmerman

impersonal forces. It is a thesis about the conditions under which certain batches of particles located at various times do constitute a single person existing at all of those times, and the conditions under which they do not. The thesis entails a kind of determination of identity by the microphysical: when you have various collections of particles constituting the body of a single person throughout a period of time, there must be impersonal physical phenomena going on involving those microphysical particles and their surroundings which are sufficient to insure that a single person exists during the period. A canny physicalist reductionist will not want to say that whether some particles go together to make up a certain person at a time is fully determined by their intrinsic interrelations alone. Compare a pair of twins, Toothy and Gappy, who are absolutely identical except for the fact that one of them has had the visible parts of all his front teeth removed. (I leave the roots of the teeth intact, so Gappy’s gums are not behaving differently from those of Toothy.) The particles making up all of Toothy except for the visible parts of his front teeth may be exactly the same, in their intrinsic arrangement, as the particles making up all of Gappy’s body. But, in Toothy’s case, they do not constitute a human body, only a part of one – whether they make up a whole human body depends upon what relations hold between them and other particles, like the ones nearby in Toothy’s front teeth. So, in looking for the underlying physical facts upon which personal identity might depend, one must pay attention not only to relations among the members of a group of particles when asking whether they constitute the same person as someone existing at some other time; one must also pay attention to whether they are embedded in a larger group of particles that could constitute that person instead. I shall help myself to the notion of a certain kind of property of collections of particles: a “complete physical specification” of the “impersonal particles” in our universe – that is, kinds of particles that can exist whether or not there are any sentient beings. Two collections of particles share a complete physical specification if and only if there is an isomorphism between them that preserves all the instantaneous facts about: (i) spatial relations among the members, (ii) all the interactions among them involving basic (impersonal) forces, and also (iii) the kind of physical environment in which they find themselves – for example, proximity to this or that kind of (impersonal) particle, or the presence of basic forces impinging upon them from things not in the collection. The particles making up Gappy’s body do not share a complete physical specification with the set of particles making up Toothy-minus-front-teeth. Although there is an isomorphism

Materialism, dualism, and “simple” theories

215

that preserves (i) and (ii), it does not preserve (iii).4 With this notion in hand, it is fairly easy to spell out a supervenience thesis that ought to be part of a physicalist reductionism about personal identity over time for creatures with bodies made of our kinds of matter. M is a micro-state =df M is a pair consisting of a set of particles and a complete physical specification. (D2) The micro-state M occurs =df the set of particles in M exemplifies the physical specification in M. (D3) H is a micro-history =df H is a series of micro-states, each one of which is paired with a different positive real number or zero. (D4) The micro-history H occurs =df for every pair of micro-states M and M* in H, if H assigns n to M and n + m to M*, then M occurs m seconds before M* occurs. (D5) A micro-history H is a continuous part of the history of a single person =df (a) H occurs, (b) H’s micro-states form a continuous series ordered by their numbers, and (c) there is a person who, for every time t at which a micro-state in H occurs, is wholly constituted at t by the set of particles in that micro-state. (D6) micro-histories H and H* are indiscernible =df (a) H assigns a microstate to some n if and only if H* does, (b) the micro-states which H and H* associate with any n have the same complete physical specification, and (c) if there is a particle x that plays a certain role in the micro-states that H associates with n and n*, then there is a particle y that plays that same role in the micro-states H* associates with n and n*. (D1)

Here is a first pass at a statement of the supervenience of personal identity over time upon the microphysical: (PR) If a micro-history is a continuous part of the history of a single person in one possible world, then any indiscernible micro-history occurring in any possible world is a continuous part of the history of a single person. Not every materialist will accept (PR); I know that, were I a materialist, I should ask for at least two modifications. 4

See Zimmerman (1998) for a more detailed presentation of the kind of supervenience thesis sketched here.

216

dean zimmerman

If mental states do not supervene upon the physical, then one might well think that (PR) is false – that personal identity over time does not supervene upon the physical either. But one might still accept Parfit’s more general reductionist thesis: personal identity over time supervenes upon the physical plus the “impersonal” (i.e. personal-identity-neutral) mental states (quasi-memory, etc.) that could be associated with the history of a single person at a particular time. To add these to the supervenience base, each stage of a history must include facts about whether, in the vicinity of the particles, there is a thinker with impersonal mental states of certain kinds. The specifications of mental states should, like the physical states of the particles, be complete specifications – ways that a mind could be in toto, at least with respect to the most fundamental, non-supervenient, impersonal mental properties. For this non-physicalist reductionist, a modified supervenience thesis can be constructed: (D7) M is a micro-psycho-state =df M is a triple consisting of a set of particles, a complete physical specification, and a complete impersonal psychological specification. (D8) The micro-psycho-state M occurs =df the set of particles in M exemplifies the complete physical specification in M, and also constitutes the body of a person who exemplifies the complete psychological specification. Modifying the rest of the definitions is trivial, and allows for the formulation of a more general supervenience thesis (hence the “G” in “(GR)”) – one that even non-physicalist reductionists could accept: (GR) If a micro-psycho-history is a continuous part of the history of a single person in one possible world, then, in any world in which an indiscernible micro-psycho-history occurs, it, too, is a continuous part of the history of a single person.5 5

André Gallois (1998, pp. 248–54) articulates a related doctrine he calls “weak anti-haecceitism.” It is more general than (PR) or (GR) (applying to anything made of parts, not just persons) and stronger in another way too: it implies that if a micro-history is the history of a single person in one possible world, an indiscernible micro-history that occurs in any world is the history of that same person, so long as the same particles play the same roles in each micro-state. Weak anti-haecceitism is, he thinks, “very plausible.” One might worry that qualitatively similar but numerically distinct objects and events leading up to the beginnings of a micro-history might make a difference to the identity of the whole made out of the parts caught up in the micro-history, even if the parts are the same. (PR) and (GR) remain silent on this matter.

Materialism, dualism, and “simple” theories

217

A second reason to doubt (PR) (and (GR) as well) is the fact that nothing in the definition of the occurrence of a micro(-psycho)-history requires that any particular causal relationships hold between earlier and later states in the history. Suppose the history of the particles in my body could occur, but with a causal “break” in it – a time at which the occurrence of subsequent micro-states does not depend, causally, upon the occurrence of earlier ones in the way they actually do in my history. Arguably, that occurrence of my micro-history would not be the history of a single person or even of a single organism.6 So long as the particles caught up in a micro-history do not all change at once, the reasonable demand for immanent-causal connections wherever there is identity over time will not allow my history to occur with a complete break in causal connections. Still, a further fix-up may well be needed along the following lines: (PR*) If a micro-history is a continuous part of the history of a single person in a possible world w, then any indiscernible micro-history occurring in any possible world w* is part of the history of a single person – so long as, for any n and n*, the micro-state corresponding to n is causally dependent (in one way or another) upon the microstate corresponding to n* in w* if and only if they are dependent (in that same way) in w. (GR*) will be understood to be the result of adding the same clause to (GR). Consideration of abnormal causal mechanisms illustrates the need for the parenthetical “in that same way” to qualify the causal dependence among micro-states. Suppose time-travel is possible, and God exists. If all the atoms in my body at noon today had been caused to jump away from one another exactly then, this organism (and I, if I am this organism) would have ceased to exist. This would be so even if, later on, God were to cause these atoms to travel back in time so as to seamlessly continue on the trajectories they were tracing prior to their jump. A kind of causal dependence could still hold between the pre-jumping micro-state of my body and the micro-states of the same particles, immediately post-time-travel. God notes the state of each particle in my body, just prior to its jump, and then uses that information to send my scattered parts back in time to form a duplicate. God’s activity insures that there is causal dependence between the trajectories of the particles prior to their jump and the trajectories of 6

Compare Armstrong’s discussion of “immaculate replacements” (Armstrong 1980).

218

dean zimmerman

the time-travelling particles. But the micro-states in the history of this pseudo-body would not display all the same kinds of causal dependence as would the corresponding micro-states in the otherwise indiscernible microhistory of my body in the actual world.7 r e a s o n s t o d e n y ( g r* ) Richard Swinburne accepts “the simple view” as a good name for his own position – a view he also attributes to Butler, Reid and Chisholm (Swinburne 1984, pp. 17, 19 and 26). Here is the beginning of Swinburne’s statement of the simple view (“P1” and “P2” should be taken to stand for terms that refer to some arbitrarily chosen person or persons – “Obama” and “Nixon,” say, or “Cicero” and “Tully”): [A]lthough apparent memory and brain continuity are, as they obviously are, evidence of personal identity, they are fallible evidence and personal identity is something distinct from them. Just as the presence of blood stains and fingerprints matching those of a given man are evidence of his earlier presence at the scene of the crime, and the discovery of Roman-looking coins and buildings is evidence that the Romans lived in some region, so the similarity of P2’s apparent memory to that of P1 and his having much the same brain matter, is evidence that P2 is the same person as P1. Yet blood stain and fingerprints are one thing and a man’s earlier presence at the scene of the crime another. His presence at the scene of the crime is not analyzable in terms of the later presence of blood stains and fingerprints. The latter is evidence of the former, because you seldom get blood stains and fingerprints at a place matching those of a given man, unless he has been there leaving them around. But it might happen. So the suggestion is, personal identity is distinct from, although evidenced by, similarity of memory and continuity of brain. (Swinburne 1984, p. 19)

Swinburne is surely right in attributing a similar view to Chisholm. Chisholm uses the term “criterion of personal identity” to mean “a statement telling what constitutes evidence of personal identity,” and draws a sharp distinction between the evidence one might have for identifying a person who exists at one time with a person who exists at some other time, and the “truth conditions” for such claims (Chisholm 1976, pp. 108–13). Both Swinburne and Chisholm include all kinds of physical and psychological continuities among the sorts of evidence one might have for sameness of persons over time, and they insist that personal identity itself is “logically independent” of the holding of such continuities (Chisholm 1976, p. 112). Swinburne clearly agrees (Swinburne 1984, pp. 18–20). Their view is 7

This would be a sort of “immaculate replacement” without replacement of the fundamental material parts.

Materialism, dualism, and “simple” theories

219

not the triviality that the phrase “is the same person as” means one thing, and “satisfies such-and-such elaborate conditions for being the same person” means something else. Their thesis is much more radical: they reject all philosophical theories that try to state persistence conditions for persons in terms of physical and psychological continuities – at least physical continuities based upon the behavior of the ordinary matter found in human bodies and brains, and psychological continuities that can be described in an impersonal way. Swinburne and Chisholm offer thought experiments to support this conclusion – thought experiments that would undermine (GR*). They describe one or another scenario in terms of the physical and psychological continuities that could hold between persons existing at different times, and then they invite us to see that there are two or more genuine possibilities consistent with the events as so far described – possibilities involving differences in which persons exist at which times. There are some possible worlds in which these physical and psychological facts hold, and a single person exists at the beginning and the end of the events described; and there are others, exactly the same with respect to physical and psychological continuities, in which the same person does not make it through the whole episode. Accepting the failure of (GR*) on these grounds has drastic consequences: if any statement of necessary and sufficient conditions for the persistence of persons over time could be given in terms of physical and psychological continuities, (GR*)-style supervenience would have to hold. So all theories of personal identity that try to formulate such conditions are doomed. Some philosophers report that, when they conduct these thought experiments, they do not even generate, for them, the appearance of more than one possibility. And those who feel the pull of the stories will frequently deny that this appearance is veridical or trustworthy. For the record, I agree with Chisholm’s relatively modest claim about the deliverances of these thought experiments: “They seem to me to be worthy of being taken, at least provisionally, as data in our philosophical enquiries about the person; in other words, we should affirm them until we have very good reason for rejecting them” (Chisholm 1970b, p. 188). In this chapter, however, I am not primarily concerned with the status of the thought experiments.8 I shall 8

In a recent essay, Swinburne defends the deliverances of these sorts of thought experiments. He provides an analysis of the notion of a metaphysically possible world in terms of logically consistent maximal descriptions couched in terms of what he calls “informative designators”; and he develops a subtle and interesting argument for the conclusion that our failure to see any inconsistency in the

220

dean zimmerman

instead explore the structure of a certain argument based upon the conclusion that there are two possibilities in these cases. I ask whether the conclusion – the failure of supervenience – can be used to support a robust form of substance dualism. Examination of the argument turns up a relatively clear doctrine that could plausibly be called “emergent materialism”; and I point out that emergent materialists can make their view consistent with (GR*) by positing immanent-causal differences due to “emergent properties.” two thought experiments Since the stories told by Chisholm and Swinburne have more bite when imagined from one’s own perspective, I will use the first person to state them. One thought experiment involves my fission – the division of my body and brain in a way that results in two equally qualified candidates for being me, either of which would understandably be taken for me had the other half of me been lost through an injury that simultaneously destroyed all the organs on the other side of my body. Here is a summary of the – by now familiar – story: Fission I could have been more symmetrical than I am, and could have undergone an operation that brought about my perfect fission. Were that to happen, it would be possible for me to be the person who has the left hemisphere of my brain and other organs from that side; and not to be the person who would be thinking with what had been my right hemisphere. But the reverse is also possible.9

There is another line of thought that, although not fully explicit in Chisholm and Swinburne, can be used to support the failure of personal identity to supervene upon physical and psychological continuities. Both philosophers insist that personal identity is a matter of all or nothing, immune to the effects of conventional decisions about where to draw the temporal borders of a thing. But the kinds of physical and psychological continuity that can hold between a person at one time and a person at

9

thought experiments proves that there are metaphysically possible worlds, in his sense, that are physically and psychologically indistinguishable while differing with respect to facts about personal identity. See Swinburne (2007). Versions may be found in Chisholm (1969, p. 106; 1970a; 1976, pp. 111–12) and Swinburne (1973–4; 1984, pp. 13–20).

Materialism, dualism, and “simple” theories

221

another time “may be satisfied to varying degrees” (Swinburne 1984, p. 16). When considering just how much of one’s body or brain could be lost at once, or just how much sudden psychological change could be endured by a person, the proponents of persistence conditions for persons stated in terms of biological or psychological continuity inevitably end up facing borderline cases of the kind of continuity they have selected. The presence of these borderline cases will lead to a failure of supervenience in the following way. Fuzziness Take any kind of mental or physical continuity that could be used in a theory attempting to state necessary and sufficient conditions for the persistence of persons. It will always be possible to describe a series of similar events involving the loss of the relevant kind of continuity, a series that satisfies these constraints: (i) at one end of the series, the event in question is a loss of continuity that seems irrelevant to my surviving; (ii) at the other end of the series, the event is a loss of continuity that seems very likely to be incompatible with my surviving; and (iii) the difference in degree of continuity between each pair of neighboring cases in the series is always too small for it to be necessary that in the one I definitely survive, while in the other I definitely do not. Still, there must always be a definite fact of the matter as to whether I survive any particular episode. So, for at least one of the events in such a series, there are possible worlds in which I survive that degree of loss of continuity, and there are also possible worlds in which I do not survive that very same degree of loss of continuity.

One way to resist fission would be to say either: in fission, I would be partially the one, partially the other; or, in fission, I would necessarily die. One way to resist fuzziness would be to say: although the differences between some neighboring cases in the loss-of-continuity spectrum may be too small to necessitate that, in one case, I definitely survive and, in the next, I definitely do not, nevertheless, there will be a transition from cases of continuity sufficient to necessitate definite survival to cases that merely necessitate its being less than definitely true that I survive. Chisholm and Swinburne emphasize the difficulties of these lines of resistance (see e.g. Chisholm 1976, pp. 110–13; Swinburne 1984, pp. 13–19). What the two thought experiments most immediately support is a nonsupervenience claim: where persons go is not settled, given all the “impersonal facts” – personal identity could “float free from” such facts, and so (GR*) is false.

222

dean zimmerman fission and fuzziness vs. supervenience upon the impersonal

Fission stories are supposed to undermine both (PR*) and (GR*). In the case of perfect symmetry, (PR*) and (GR*) require that either the person dies or goes both ways; but he or she cannot go one way in one possible world and the other way in the other. In fuzzy cases, the facts about ordinary matter and impersonal psychological continuities included in micro-histories are imagined to vary smoothly along all the dimensions that could be relevant to sameness of person; but whether something is the body of the same person is (we are supposed to feel) sharp. So there must be worlds in which the same physical and psychological continuities hold, but in one of them the micro-psychohistory is the history of a single person, and in the other it is not. Clearly, accepting the conclusion that (PR*) and (GR*) fail due to these possibilities is to treat personal identity over time as involving some kind of “further fact.” But does countenancing these possibilities support the idea that, in addition to the particles that make up the ordinary matter in our bodies, there is also a further substance – be it immaterial or not – that either is the person or is crucial to the ongoing existence of that person? It has always seemed to me that recognizing these seeming possibilities as genuine does support a robust dualism according to which a person is not made entirely of the kinds of matter that appear in the micro(-psycho)histories of our bodies. I have accepted what I shall call the “conditional necessity of (GR*)”: necessarily, if persons were entirely material, (GR*) would be true. And I am not alone in being attracted to arguments based upon this idea.10 Here are two, based upon fission and fuzziness: Fission argument for dualism 1. A perfectly symmetrical division of my brain and body could result in two indiscernible micro-psycho-histories, which would overlap until the division, and then follow the two halves of my body into different rooms; the histories would be mirror images of one another with respect to causal dependencies among stages; and the procedure would have (at least) the 10

Martine Nida-Rümelin makes use of it, appealing just to the case of fission (Nida-Rümelin 2010, esp. pp. 208–9). She does not regard her support for (her version of) premise 2 as consisting of a “conceivability argument,” but rather as an argument that denial of two genuine possibilities in the case of fission would imply that we (along with any other beings capable of attributing conscious thought to themselves or others) would be radically mistaken about all sorts of important matters.

Materialism, dualism, and “simple” theories

223

following two genuinely possible (incompatible) outcomes: (a) the left half of my body ceases to be a part of me, and I survive with a body composed entirely of the matter that formerly constituted my right half; and (b) the right half of my body ceases to be a part of me, and I survive with a body entirely constituted by (what was) my left half (based on fission). 2. If I had no parts other than the particles making up the ordinary matter in this body, then the perfectly symmetrical division resulting in indiscernible micro-psycho-histories (with similar causal dependencies) would not have the following two genuinely possible (incompatible) outcomes: (a) I survive the loss of my left half, becoming constituted by my right half; and (b) I survive the loss of my right half, becoming constituted by my left half (from the conditional necessity of (GR*)). Therefore, I have a part (perhaps an improper part) other than the particles making up the ordinary matter in this body. Fuzziness argument for dualism 1*. There is a possible continuation of my actual micro-psycho-history in which all the same kinds of causal dependencies hold among stages, but there are two genuinely possible conclusions: (a) I definitely still exist at the end of it; and (b) I definitely do not still exist at the end of it (based on fuzziness). 2*. If I had no parts other than the particles making up the ordinary matter in this body, then there could be no continuation of my actual micro-psycho-history with the same causal dependencies and two genuinely possible conclusions: (a) I definitely still exist at the end of it, and (b) I definitely do not still exist at the end of it (from the conditional necessity of (GR*)). Therefore, I have a part (perhaps an improper part) other than the particles making up the ordinary matter in this body. Most materialists accept a kind of global supervenience of all facts about persons upon the fundamental physical facts about the matter that makes up our bodies; and that sort of supervenience will lead to acceptance of the conditional necessity (and actual truth) of both (PR*) and (GR*). But most materialists are willing to resist these arguments by simply denying their initial premises: fission and fuzziness have their seductive appeal, but they only seem to be possible. A few materialists take a different approach. They regard physical fission as a genuine possibility for creatures relevantly similar to us; and they believe that, even if both resulting persons were equally similar to the original, there would be a “further fact” about which (if either) was the same as the person

224

dean zimmerman

who underwent the division. (I find less explicit discussion of fuzziness by the materialists I have in mind; but their responses to fission suggest analogous things that could be said about fuzziness.) So these materialists accept the initial premise of the fission argument. But they deny the conclusion. I will appropriate the over-used, multivalent label “emergent” as a qualifier for their brand of materialism about persons. The question I want to ask is: Where do the emergent materialists think the arguments go wrong?

persons as entirely material but “emergent” One natural place for the emergent materialist to resist is by giving up (GR*) (and the conditional necessity of (GR*)), enabling her to reject premises 2 and 2* of these arguments. (In the next section, I offer the emergent materialist a different strategy: accept (GR*), and its conditional necessity, but deny the first premises of the arguments from fission and fuzziness.) Let me put some words into the mouth of this sort of emergent materialist: I have none but physical parts, consisting of ordinary types of matter. Nevertheless, at least in cases of fission and fuzziness, one and the same micro-history or micropsycho-history can be the history of a single person in some possible worlds, but not in others. There is a further fact – whether I go to the left or the right in a fission case, or when exactly I cease to exist in a fuzzy case – that is not settled by all the impersonal facts. Since my comings and goings are not completely determined by facts about my physical parts, I can rightly be said to be something “over and above” my parts; I show it by my independent spirit. But do not take the metaphorical phrase seriously. Strictly speaking, there are no such things as spirits (or at least no human spirits, if we want to leave God out of this).

Imagine a perfectly symmetrical but otherwise human-like being who undergoes an operation that produces two mutilated but conscious persons waking up in rooms 1 and 2. There is one micro-psycho-history consisting of the micro-psycho-states of the particles in the original patient and those of the patient in room 1; and there is a mirror-image micro-psycho-history that begins with the same micro-psycho-states but is continued by particles in room 2. The emergent materialist supposes that, in one possible world, the first micro-psycho-history is that of a single person but the second is not; while in another possible world, the reverse is the case. In an operation like this, she says, there is simply a “further fact” about whether the patient loses the right half of his body and moves into room number 1, or loses the left half of his body and moves into room number 2; and there is a

Materialism, dualism, and “simple” theories

225

corresponding “further fact” about whether a new person is waking up in room number 2 or room number 1. Suppose that, as a matter of brute “further fact,” the person recovering in room 1 is the actual survivor of the operation; and that the patient in room 2 came into existence upon the operating table. If the emergent materialist were to explain the situation to the new man in room 2, the latter might well protest. He can make an equally strong case for the claim that he was the pre-operative person. But the emergent materialist will say: In many respects, you do seem to be an equally good candidate for being that man. But the fact remains that there was one person who went into the operating room, and he is now waking up in room number 1; and there is no person who went into the operating room and is now in your room.

These remarks will be more convincing if the emergent materialist can follow them up with the claim that there is nothing that went into the operating room looking like a surprisingly symmetrical human being and now lies recuperating in room number 2. But an emergent materialist who accepts the doctrine of temporal parts cannot say this – at least, not if her metaphysics recognizes arbitrary fusions of the temporal parts of persons. When the symmetrically fissioning person shrinks to half his original size and leaves the operating room for room number 1, and the mere fissionproduct wakes up in room number 2, this sort of emergent materialist must admit that there is something that went into the operating room in the form of a whole person, and that is now waking up in room number 2. Her claim must be that the only person who entered the operating room is now waking up in room number 1. The character in room number 2 will want to know: “Why does my word ‘I’ not refer to a person who existed before the operation? After all, there is a thing that existed then, and that was fearfully anticipating the operation, and that is now uttering these words; how could it fail to be a person – in fact, the person I am?” A temporal parts theorist who denies the existence of arbitrary fusions could give a perfectly good answer to the patient waking up in room 2 when he asks for a reason to think that he was not the patient who entered the operating room: “Your word ‘I’ doesn’t refer to something with those preoperative stages as parts because there is nothing that has both your current stage and those stages as parts. Those temporal parts and your temporal parts do not form a larger whole.” By contrast, an emergent materialist who accepts arbitrary fusions of temporal parts will simply have to regard being a person as a property that does not supervene upon the micro-history or micro-psycho-history of an object.

226

dean zimmerman

How bad would this be? Worse, I think, than having to accept the failure of microphysical supervenience for the reasons given by Trenton Merricks. Merricks raises difficulties for supposing that the intrinsic properties of collections of particles, plus spatio-temporal and causal relations holding just among the members of the collection, provide a sufficient supervenience basis for facts about whether some particles make up a conscious human being (Merricks 1998a). As Sider has pointed out in his criticisms of Merricks’s argument, the materialist is likely to think that being a human person is a spatially maximal property, like being a rock (Sider 2003, pp. 142– 3). A large part of a rock is not a rock – although it would be one, were it not embedded in more rock. Similarly, a large part of a human body, such as a body-minus-its-left-index-finger (to use Merricks’s example (1998a, p. 62)), might be the sort of thing that would be a person, were there not more organic matter attached to its hand. Both properties display “bordersensitivity”: whether something qualifies as a rock or (given Sider’s materialist assumptions) a human person “depends on what is going on around its borders” (Sider 2003, p. 139). Suppose that a materialist wants to regard being conscious as intrinsic, and also to follow Sider in treating being a human person as a maximal property due to border-sensitivity as in the case of the rock – a human person is the largest of ever so many would-be persons. In that case, she must see a host of conscious non-persons in the vicinity of every human being.11 Materialists can avoid this conclusion if they follow Sider in accepting the extrinsic nature of consciousness.12 One might have hoped that an analogous move could be made by the emergent materialist who believes in temporal parts and arbitrary fusions; that she could say the failure of supervenience was due to the fact that personhood is an extrinsic matter, and a true supervenience claim can be formulated by appeal to a broader supervenience base – one that is sensitive to the presence of nearby candidates for being a certain person. But the emergent materialist’s denial that being a (single) person supervenes upon micro-histories or micro-psycho-histories cannot be rendered benign by claiming that personhood is an extrinsic matter, and only fails to supervene 11 12

The moral can be drawn from Merricks (1998a) and Olson (1995). A materialist like Sider can sensibly conclude that consciousness must be extrinsic, in virtue of its analytic ties with being a person – something a materialist will naturally take to be border-sensitive. The move is not without its costs, as Merricks points out (2003). I would argue that it is least costly for the materialist who regards thinking and experiencing as no more natural or fundamental than such obviously non-fundamental activities as executing a karate chop or shaking hands with someone. Arguably, body-minus-its-left-index-finger does not do either of the latter things, although it would be doing so were there no left index finger attached to it. Materialists about persons who are property dualists about the phenomenal may not have it so easy. But that is another story.

Materialism, dualism, and “simple” theories

227

upon micro(-psycho)-histories due to their overly intrinsic nature. A microhistory includes all the facts about the disposition of matter surrounding the collections of particles caught up in it. So one can tell, just by looking at the micro-history of the particles that make up the person who survives the operation, that his other half was removed intact and is now recovering nicely in room 2. I conclude, then, that emergent materialism and the doctrine of temporal parts make poor bedfellows, barring adoption of mereological principles disallowing arbitrary fusions. Those who reject temporal parts altogether seem to me to be in a dialectically much stronger position. It is not surprising to find that all the clearest examples of emergent materialists are also opponents of temporal parts – for example, Trenton Merricks, Tim O’Connor, Jonathan Jacobs and Lynne Rudder Baker.13 holistic causation to the rescue? In response to the arguments from fission and fuzziness, an emergent materialist must either reject (GR*) (and its conditional necessity) or find some way to deny premises 1 and 1*. She accepts fission (the case upon which I will focus): there is one micro-psycho-history that corresponds to the stages in the life of a person who loses half his body and then wakes up in room 1, and another that corresponds to this person’s pre-operative stages plus the stages of a different person who comes into existence on the operating table and wakes up in room 2. To accept fission and reject premise 1, she must find a causal difference between otherwise exactly similar micropsycho-histories. Might the causal relations holding between the whole body and just one of the immediately post-operative persons be of a categorically different nature than the ones holding between that body and the other post-operative person? If so, the two micro-psycho-histories might be indiscernible; but they would differ in the kinds of causal dependence holding between stages at the time of the surgery. (GR*) could then be accepted, but premise 1 denied. Consider the fissioning symmetrical person, who goes to the left and ends up in room 1. In the fission of this organism into two persons, it would be natural to expect that the properties of each resulting person’s body, 13

See Merricks (1997, 1998b); O’Connor and Jacobs (2010, pp. 80–2); and Baker (2000, pp. 132–41). For their opposition to temporal parts, see Merricks (1999); O’Connor and Jacobs (2010, p. 74, fn. 5); and Baker (2000, p. 22). If I understand Stephen Davis’s appeal to the will of God in determining identity (Davis 2010, pp. 26–7), he, too, is committed to the falsehood of (PR*) and (GR*). Though I lack hard evidence, I would bet that he, too, rejects the doctrine of temporal parts.

228

dean zimmerman

immediately after the operation, would be adequately causally explained by describing two things: the features of just one half of the pre-operative person, and the procedure that separated the halves. On this natural assumption, there is no additional property that belongs only to the whole pre-operative body immediately prior to surgery, and that causes the postoperative persons to have some intrinsic feature directly – i.e. not in virtue of the feature’s correlation with contemporaneous intrinsic properties belonging to just one or the other half of the person. I will call this supposition “no holistic causation.” On this hypothesis, any effect the whole person had upon the intrinsic nature of the person who woke up in room 1 would be most directly caused by just a part of the pre-operative person. And this does not seem to me to be consistent with supposing that the causal relationship between the whole person and the new guy in room 2 is of a radically different nature. Prior to the operation, the left half of the person is not itself a person; and it is causally responsible for the nature of the person in room 1. The right half is not a person either; and it is causally responsible for the nature of the person in room 2. Considered intrinsically, the two halves generate the two separate persons by an exactly similar process. Assuming no holistic causation, the whole person “works through” one of its halves to bring about the existence of a person with certain intrinsic features in room 1; he causes something by having a part that causes it. And the whole person also has an exactly similar part – his right half – which, when considered by itself, does exactly the same thing. If something (in this case, a person) uses two exactly similar agents (its two halves) to bring about two exactly similar results by means of exactly similar causal processes, and the thing stands in exactly similar relations to the two agents, then the thing does not cause these two results in significantly different ways. But relax the no holistic causation assumption, and things look altogether different. Suppose that the man is feeling queasy right before the operation; that this feeling of queasiness causes subsequent states of a similar sort; and that it does not cause them in virtue of its correlation with more direct causes of subsequent feelings, causes attributable to intrinsic properties of the brain. When the man’s pre-operative queasiness causes his post-operative queasiness, it does not do so purely in virtue of intrinsic properties of the half-brain that he kept (since, in general, queasiness is now supposed to have effects that are not entirely due to the physical properties of the man’s brain). This opens up space for a difference in the kinds of causal relations holding among micro-psycho-states in the two otherwise similar micropsycho-histories. The pre-operative queasiness of the whole man could

Materialism, dualism, and “simple” theories

229

directly cause the queasiness felt by the man who is wheeled off to room 1; while the similar queasiness presumably felt by the man wheeled off to room 2 would have to be caused in some other way – perhaps by the physical conditions of the right hemisphere that, upon removal, has become his brain. There would be a kind of causal – or at least counterfactual – dependence holding between the original patient’s pre-operative queasiness and that of the new man in room 2. If the original patient had not felt queasy, he would not have had the kind of right hemisphere that would tend to produce queasiness in the new man in room 2. But, on the holistic causation hypothesis, this would not be the same kind of direct causal dependence that holds between the post-operative queasiness of the man in room 1 and the pre-operative queasiness of the whole man. It may well be no coincidence that some of the main defenders of emergent materialism (e.g. Merricks 2001, ch. 4; O’Connor and Jacobs 2010) posit “emergent”14 properties exemplified by the person as a whole – properties that are causally efficacious, and not just in virtue of the causal powers exercised by the smaller parts of the body. Indeed, the description of a case of perfectly symmetrical fission given by O’Connor and Jacobs resembles – at least abstractly – the story I told about the whole man’s queasiness.15 Of course, if the holistic causation in question can hold between pre-operative patient and both fission-products, the emergent materialist is once again forced to deny the conditional necessity of (GR*).

fissioning simples and premise 2 of the fission argument for dualism The emergent materialist asks me – as a defender of the argument for dualism based on fission – why I believe that, if something is a mere material object, then in a case of perfectly symmetrical fission, with indiscernible micro(-psycho)-histories, the object must go either both ways or neither. The question seems especially pressing when I reflect upon the very 14

15

I should note that Merricks does not himself use the term “emergent” to describe the properties in virtue of which an “object’s causal contribution would be . . . independent of what its atoms were like and so, presumably, independent of what its atoms cause” (Merricks 2001, p. 90), but the label is often applied to such properties by other advocates and opponents of the idea that wholes have properties satisfying this description. See O’Connor and Jacobs (2010, pp. 80–1). Their “emergent level states” seem to consist mainly of the “particularities” of individuals, however (pp. 74–7), and not phenomenal properties. I am not a fan of the metaphysics of “thin particulars” which they use to explain the notion of a “particularity”; so, were I an emergent materialist, I should look for other properties that might have causal powers in some way independent of the activities of a person’s proper parts.

230

dean zimmerman

different reaction I have to the case of a simple substance – be it a particle, an intelligent monad, or what-have-you – undergoing apparently symmetrical fission.

Simple Fission Imagine a particle, x, that is buzzing along, minding its own business. Suddenly, it is zapped by some kind of energy, and two particles exactly like x appear, headed off to the left and right of x’s original trajectory in what looks like a case of symmetrical fission. Suppose the counterfactuals are the same on both sides: the particle on the left (“Lefty” will be shorthand for this descriptive phrase) would have had spin up had x had spin up immediately prior to the apparent fission, Lefty would have been moving more rapidly had x been moving more rapidly then, etc.; and similarly for the one on the right (“Righty,” another abbreviated definite description).

What, if anything, is wrong with supposing that there are three genuinely possible endings to simple fission that do not differ in the facts about the particle stages, their environments, and the relations of causal dependence holding between particle stages: (i) x ceased to be, replaced by two new particles; (ii) x is Lefty and Righty is a new particle; and (iii) x is Righty and Lefty is a new particle? Personally, I find myself inclined to say that these do represent three distinct possibilities, each consistent with the case as described. But then why should I have such a different reaction to apparently fissioning composite things, like the emergent materialist’s physical persons? If I accept these three outcomes as genuine possibilities for simple fission, have I not thereby given up the idea that, when an object is entirely physical, its persistence through time should supervene upon facts that do not directly imply anything about the persistence of the object in question and that concern only its smallest parts and the relations of causal dependence among its stages? If so, have I not lost the right to insist upon the conditional necessity of (PR*) and (GR*)? These supervenience principles describe the way complex wholes depend upon the histories of their parts. When one reaches the level of a partless simple, it is not at all clear that its persistence through time should be thought to consist in anything other than the simple’s existing at more than one time. What would a supervenience principle for the persistence of simples look like? It would have to find more fundamental facts that do not directly imply anything about the identities of the kinds of simple in question, and exhibit relations at this level of facts that could plausibly be the grounds of simple-persistence. The friends of temporal parts will have

Materialism, dualism, and “simple” theories

231

no difficulty finding subjects for such facts, namely, instantaneous temporal parts; but, as shall appear, they should probably resist the idea that simple fission has the three possible endings. Others, who do not regard temporal parts as inevitably accompanying all persistence through time, may well accept that there are three possible endings; but either insist that this is due to the partlessness or fundamentality of the object in question (which distinguishes the case from that of a complex material person), or claim that the story leaves out important details about the kind of causal dependence holding between x and its offspring, Lefty and Righty. Those who believe in the doctrine of temporal parts deny that a persisting simple really is partless. They are likely to accept the supervenience of the persistence of even the smallest particles upon facts about their instantaneous temporal parts – i.e. to be immanent-causal reductionists about simples. A temporal parts metaphysics may or may not leave room for a deep distinction between immanent causation and other sorts of causal dependence. On a very thin notion of immanent causation, the counterfactuals mentioned in simple fission would settle all the facts about what caused what, and by what means – leaving no room to suppose that the spin and speed of Lefty, say, but not Righty, depended in a special immanentcausal way upon x’s state at the point of fission. This metaphysician really should accept immanent-causal reductionism, which requires symmetrical treatment of perfect fission; she should deny that the three endings are genuinely possible. A friend of temporal parts might have a thicker notion of immanent causation; she might think that there is a special kind of causal relation that always holds between stages of the same particle, a relation that is not automatically present whenever there is the kind of counterfactual dependence described in simple fission. She need not immediately draw the conclusion that immanent causation must either fork or stop when x is zapped; for all she has been told, the counterfactuals on just one side might be due to immanent causation, the others being the product of another kind of causal relationship. If she thinks that immanent causation is, nevertheless, a relation that can hold between the earlier states of one particle and the later states of another, then immanent-causal branching can simply be added to the story. And, again, this temporal parts metaphysician should be an immanent-causal reductionist, not admitting the three possibilities for x once all the causal facts are stipulated to be symmetrical as well. Some of us do not believe that persisting through time involves, as a matter of necessity, the possession of new, short-lived temporal parts. We have a right to the reaction I reported as my own: simple fission is

232

dean zimmerman

consistent with each of the three endings. Some of us may deny that any special kind of causal dependence must hold between the way a simple is at one time and the way it was at earlier times – or that any causal dependence must hold at all. These opponents of temporal parts will reject supervenience of the persistence of particles upon other facts. I do not see, however, that doing so should be thought to undermine their confidence in principles like (PR*) or (GR*), which pertain to complex wholes. Other opponents of temporal parts have at their disposal a sort of backhanded way of saving a supervenience principle even for simples. Rejection of temporal parts is naturally coupled with the idea that a robust kind of immanent causation is involved in persistence. When there is genuine persistence of a perfectly natural kind of thing, there is a very distinctive sort of causal relation holding among stages of the thing – a way for stages in the history of a single particle to depend upon one another that is inevitably accompanied by identity, but that stands at a slight conceptual remove from identity nevertheless. The state or event of a certain fundamental object’s having a certain fundamental property at a time can have a tendency, in propitious circumstances, to bring it about that there will subsequently be a fundamental object of the same kind with a fundamental property from the same “determinable” family. But the opponent of temporal parts believes it is possible to have this sort of tendency for two very different reasons: a thing can have it in virtue of a propensity to generate a new thing of the same general type, or it can have it in virtue of a propensity to merely hang around. Simple fission does not specify whether this second sort of propensity is being exercised at all at the point of fission; nor whether, if it is exercised, the effect occurs toward the right or the left. If this propensity was not triggered by the zapping, x ceased to be – however similar to it Lefty and Righty might be, and however robust the counterfactual connections between x’s states and theirs. I find this second response quite plausible: there is a sort of necessarily immanent relation among stages in the history of a fundamental thing, it constitutes a distinguished sort of causation, and it implies the identity of the thing in question. So when I hear the story about x’s apparent fission, I see genuinely distinct possible ways in which x may have caused the stages of Righty and Lefty; and these differences seem highly relevant to the question whether x survived, and, if so, whether it is Righty or Lefty. The supervenience of physical persistence upon the microphysical and the causal, when enriched with this kind of causal dependence, becomes unproblematic. If one supposes that x is Lefty, the micro-history of x-up-to-zap-then-Lefty

Materialism, dualism, and “simple” theories

233

is not exactly like the micro-history of x-up-to-zap-then-Righty; one kind of causal dependence holds between early Lefty stages and x’s stage at the zap – immanent causation, involving the exercise of a propensity to hang around – and some other kind of causal dependence holds between the early Righty stages and x’s stage at the zap. So I have found a way to explain why my reaction to the fissioning simple does not betray a lack of commitment to the supervenience of the persistence of physical things upon microphysical histories plus (the right kind of) causal dependencies within those histories. Since the causal dependence is so closely tied to persistence through time, it would not satisfy reductionistic attitudes toward the persistence of simples. But I do not feel much pull toward the idea that identity over time, for simples, must consist in something else. The differences in direction of immanent causation used in response to simple fission suggest a way for an emergent materialist to resist the fission and fuzziness arguments for dualism without quite giving up (GR*) (or its conditional necessity): she can deny premise 1 (and 1*, though I will focus on fission) by positing an important immanent causal difference between otherwise indiscernible micro-psycho-states. In response to simple fission, I gave a sketchy account of a special kind of causation that an opponent of temporal parts might believe in, consisting in the exercise of a propensity to keep existing. Those of us who posit such a causal process may well deny that it occurs in anything other than a fundamental entity. (I should want to leave it a largely empirical question whether subatomic particles persist through time, whether the actual fundamental physical entities are small or large, and whether they include fields or a substantival space-time manifold or both. For all I know or care, all fundamental physical entities may consist of series of momentary things, just as the friends of temporal parts allege. I simply see nothing inevitable about this.) In order to give the emergent materialist a chance, here I shall suppose that complex organisms can exercise the deeply different kind of hanging-around causation that I have posited at the level of fundamental persisting things. In particular, I shall allow that it may hold between the stages of an object while that object suddenly loses a large part. If some state of the pre-operative person immanently causes some state of the person in room number 1, and no state of the man in room 2 is similarly immanently caused, premises 1 and 1* could obviously be denied by the emergent materialist. This strategy may make fission consistent with the letter of (GR*); but it is a nice question whether the supervenience principle still captures something of Parfit’s reductionism, once immanent causation

234

dean zimmerman

is posited as a distinctive kind of causal dependence.16 On the face of it, appeal to causation that only works when an object persists would seem to seriously undermine reductionism about personal identity. However, if immanent causation can be characterized without any special mention of persons, or of psychological properties that imply that a person exists, perhaps the emergent materialist will not have violated the spirit of Parfit’s reductionism after all. What is needed at this point is a more precise account than I have given so far of two things: the nature of the “hanging around” kind of causation, and the motivations for advocating reductionism. Loose ends and confessions Two of the central topics of this chapter are begging for a fuller treatment than I have been able to give them: (1) I am convinced that there is much more to be said in favor of the conditional necessity of (GR*), and against the emergent materialist who would resist arguments for dualism simply by denying (GR*); (2) I am disappointed by the sketchiness of my attempt to characterize immanent causation, and to distinguish it sharply from other species of causal dependence. I would hope to identify a kind of causal relation definable in terms of the exercise of “strongly identity-entailing conditional powers” – a notion that proved useful when I was grappling with Shoemaker’s distinction between “thin” and “thick” properties (Zimmerman 2009, pp. 699–703). I conclude with a couple of confessions. The conditional necessity of (GR*) seems to me to glow with the light of Truth, and fission and fuzziness appeal strongly to my imagination. But if the arguments from fission and fuzziness provided my only reasons for believing in a dualism of persons and bodies, I would regard the case as inconclusive. Conceivability may be an important source of evidence for possibility, but it can lead us astray – especially when used as evidence for the ascription of essential properties to particular individuals, as in the present case.17 But other considerations push me toward dualism – or at least away from all the more plausible kinds of materialism, including the versions of emergent materialism canvassed here (Zimmerman 2003, 2004, 2010). In 16 17

I am grateful to Marco Dees for posing this nice question. For my criticisms of one argument for dualism that relies upon conceivability in an unsafe way, see Zimmerman (1991).

Materialism, dualism, and “simple” theories

235

the actual circumstances, then, and despite my qualms about conceivability, the arguments from fission and fuzziness are not irrelevant. Dualism receives a small epistemic boost for me, in virtue of the fact that it respects the conditional necessity of (GR*) and allows me to trust my modal instincts about fission and fuzziness.18 18

I learned much from the questions and suggestions of participants in the conference “Diachronic Personal Identity: Complex or Simple?,” organized by the Department of Philosophy at the Theological Faculty of the University of Innsbruck, July 2010. I am especially grateful to Richard Swinburne and Hud Hudson. Much later, and in the nick of time, Ofra Magidor showed me that I had somehow begun to use “(GR*)” to mean two quite different things; a reading group at Rutgers uncovered some technical flaws; and Marco Dees helped me through some confusions. Unfortunately, other comments from these sources, and from audiences at Oxford, Yale and Buffalo, raised important issues and objections too large to be addressed before this book went to press.

chapter 13

The morphing block and diachronic personal identity Hud Hudson

the criterion of diachronic personal identity One perplexing philosophical puzzle concerns how best to complete the following sentence: (C1) Necessarily, if x is a human person at a time, t, and y is a human person at a distinct time, t*, then x = y if and only if and because ______________________________.

Defending any particular candidate for filling in the blank would amount to endorsing a constitutive (rather than an epistemic) criterion of diachronic personal identity and would yield a more informative claim than a mere necessary bi-conditional insofar as it would provide illuminating conditions that serve as the grounds for the persistence of a human person. Plausibly, however, that way of putting the puzzle unfairly privileges one metaphysics of persistence over another at the outset, for although the endurantist is now ready to go on a hunt for the blank-filler, the perdurantist (more on this distinction on p. 237) first requires a reformulation: (C2) Necessarily, if x is a momentary stage of a human person at a time, t, and y is a momentary stage of a human person at a distinct time, t*, then x is a momentary stage of one and the same human person as y if and only if and because ________________________.

Several – but not all – prominent theories of personal identity can be introduced and explored by way of this challenge to properly complete the sentence appropriate to one’s general metaphysics. My goal here, however, is not to furnish an overview of current theories of personal identity, but rather to identify and pursue a problem for many of them that are routinely classified among the so-called complex views of personal identity. As a bit of disclosure, I share Eric Olson’s suspicion that there is no single and straightforward thesis separating the simple and the complex views, despite the wide-ranging use of those terms, but whether or not the labels do in fact mark out a significant distinction between theory types, there is an 236

The morphing block and diachronic personal identity

237

intriguing worry worth investigating that threatens several of the most popular current examples of the putative complex views.1

two complex views of diachronic personal identity Introducing our targets:2 discussions of diachronic personal identity should begin, I think, with the question “What is a human person?” Although not controversy-free, I favor a materialist metaphysics of the human person, according to which a human person is identical to a certain, highly organized, material object.3 Some variety or other of materialism is the dominant view these days, but just which material object we are to identify with a human person is hotly contested. The animalists favor identification with an entire human organism, while the minimalists select the least inclusive thing that houses the relevant psychology: say, a proper part of a human brain.4 Next (as was indicated by our two options for formulating a criterion on p. 236 above) let us ask the question “How does a material object persist through time?” Again, although hardly controversy-free, perdurantism and endurantism remain at the top of the candidate answer list. Perdurantists take a persisting material object to be extended in four dimensions (three spatial and one temporal) and to be composed of temporal parts (as well as of spatial parts). Persistence across an interval of time is a matter of being partly present at the different times in that interval (i.e. of having different temporal parts located at the different moments in that interval). Endurantists take a persisting material object to be an entity extended in three dimensions (all spatial). Persistence across an interval of time is a matter of being located at the different times in that interval (i.e. of being multi-located at a series of non-simultaneous three-dimensional regions).5 1

2 3 4

5

See Olson’s “In search of the simple view” (Chapter 2 in this volume). I borrow the “if and only if and because” connective from Olson’s discussion. Hereafter I drop the hedging “putative” when referring to the complex views. The following three paragraphs describing two representative complex views contain material from Hudson (2012). For an extended discussion of exactly this question and a defense of this reply, see Hudson (2001). There are many other flavors of materialism, as well, but these two will do for now. For a representative defense of animalism, see Olson (1997b). For a critique, see Hudson (2007). For a representative defense of minimalism see Hudson (2001). For a representative defense of perdurantism, see Sider (2001). For a representative defense of endurantism, see Van Inwagen (1990a). Since human persons are extended in three spatial dimensions, I am presenting our two alternatives against that assumption, but (strictly speaking) each theorist can accommodate minima, too, and offer the very same explanations of the persistence of material objects extended in only two or one or zero dimensions.

238

hud hudson

Finally, we might ask the question “Does our preferred answer to the previous two questions give us any guidance on filling in the blanks in (C1) or (C2)? – that is, on selecting a criterion of persistence for the special kind of material object that counts as a human person, a theory that will allow us to determine the beginning and ending of a human person’s lifespan?” Animalism together with endurantism, for example, may steer us in the direction of a biological criterion of persistence for human persons, a theory that will allow us to determine the contours of a human person by locating the precise beginning and ending, shape and size of a certain organism. Alternatively, minimalism together with perdurantism may lead us to a psychological criterion of persistence for human persons, a theory that will tell us that to focus on a human person is to trace out the history of a certain psychological profile, whether it stays put in a single body, or transfers from one human animal to another, or jumps across a temporal gap, or even shifts from an organic to an inorganic host.6 Accordingly, we may hereafter focus our attention on two clear and representative examples of proponents of complex views of diachronic personal identity: the animalist–endurantist–biological criterion theorist, who will fill in the blank in (C1) with his favored way of indicating sameness of human animal, and the minimalist–perdurantist–psychological criterion theorist, who will fill in the blank in (C2) with his favored account of a psychological continuity and connectedness relation between human person-stages.7 But before generating problems for these theorists, we first have some metaphysical stage-setting to do. five theories of time A popular metaphysical picture can generate a worry for these complex views of diachronic personal identity. Consider five theories in the philosophy of time: presentism is the view according to which only present things exist. The growing block theory is the view according to which only present and past things exist. The shrinking block theory is the view according to which only present and future things exist. The disappearing branch theory is the view according to which past, present and future things exist (and in 6 7

For the arguments from animalism and endurantism to the biological criterion and from minimalism and perdurantism to the psychological criterion, see Hudson (2001, ch. 4). Olson is an example of the first kind of theorist and I am an example of the second. For the record, (C1) need not be so narrowly construed, for one could certainly think (as Olson does) that “human person” is a phase sortal and thus that a thing can cease to be a human person without ceasing altogether.

The morphing block and diachronic personal identity

239

which the future consists of a proliferation of equally real branches that suddenly disappear as soon as time flows along any path excluding them). Eternalism is the view according to which past, present and future things exist (with no additions, subtractions or disappearing branches). Other theories compete alongside this field of five, but our quintet is representative enough. The literature on the advantages and disadvantages of these competing theories is massive, and the references given throughout this chapter comprise only the tip of that philosophical iceberg which is the debate on the nature of time. In this chapter, I am interested in exploring a feature shared by three of these views in particular – the growing block, the shrinking block and the disappearing branch theories – and I will develop my thought primarily in comparison with the growing block theory, since it is the best known of the three.8 As it is standardly characterized, the growing block theory offers us a picture of the universe featuring a space-time volume that increases as time passes. At any given moment exactly one time is special – the time associated with the hyperplane on the surface of the block in the direction of its growth. The outermost surface, so to speak, is the new kid on the block; it did not exist moments before, and although it will continue to exist, it will not remain the outermost surface moments hence. During its brief instant in the spotlight – before becoming ever more imprisoned in the block’s interior – its stock of facts and events are present. Soon they will become forever past and take their eternal places frozen in the block, but for one shining moment they are privileged – balanced on the very edge of Being. (A description of the basic idea behind the shrinking block theory can be had straightaway just by thinking about its name and adjusting bits of the story just told accordingly.) On some scorecards, the growing block theory is thought to combine the best features of its primary rivals. Like presentism, it proclaims the uniqueness of the present, recognizes objective and irreducible temporal properties, takes tense seriously, and countenances the genuine passage of time. Like eternalism, it furnishes truth-makers for past truths, provides relata for cross-temporal relations, and acknowledges the existence of many objects that are not present. Of course, this very combination of commitments is also alleged to be the source of its decisive refutation, with opponents complaining that if it were true, we (absurdly) could not know it is now 8

For formulations and discussions of the growing block theory see Broad (1923) and Tooley (1997). The next two paragraphs describing the growing block theory contain material from Hudson (2012).

240

hud hudson

now or else would risk living amongst zombies in the block (i.e. amongst non-present existants with no consciousness).9 Other critiques balk at a privileged present and allege incompatibility with special and general relativity or else target the endorsement of time’s passage, maintaining that the view must be supplemented by a hypertime in which the growing of the block occurs and against which its rate of passage can be measured.10 Whatever its other defects, I judge Ned Markosian to have shown this final criticism to be in error.11 And, although I am quite intrigued by the compossibility of the growing block theory and the hypothesis of hypertime (for current purposes), I am happy to concede that if there is something metaphysically amiss with the growing block, the shrinking block, or the disappearing branch theories, it is not for want of being embedded in a second temporal dimension. So just what is so striking about these views that calls for special attention? Well, the catch-phrase associated with them is fairly remarkable: namely, that the volume of space-time differs at different times. But that is not really a careful enough slogan, since if (in the growing case) time has an infinite past or (in the shrinking or disappearing cases) time has an infinite future, the total volume need not increase or decrease even with the addition or subtraction of hyperplanes. Perhaps, then, it is better to put the shared commitment like this: on growing, shrinking and branching models, different regions of space-time exist when different times are present. That is, to the extent that these three theories of time are taken to be live metaphysical possibilities, so, too, the theses that points of spacetime can come into existence and go out of existence are also taken to be live metaphysical possibilities. But just how is this proposal to be understood? I think the most promising response to that question is to presuppose a substantivalist space-time, identify times with hyperplanes, and incorporate a certain independence or recombination principle for space-time points. Substantivalism takes space-time to be a concrete particular with an ontological status not reducible to relations between material objects (or any other entities). Such a view identifies space-time and its extended subregions with either pluralities or else fusions of uncountably many, simple, unextended, space-time points, and it countenances a perfectly natural 9

10 11

On the alleged obstacle to knowing whether it is now now, see Bourne (2002) and Braddon-Mitchell (2004). For the zombie problem, see Forrest (2004) and Heathwood (2005). For an attempt to undermine the motivation for the growing block theory, see Merricks (2006). J. J. C. Smart challenges the notion of temporal becoming and raises pointed questions about the need for a hypertime (Smart 1949, 1998). Markosian (1993) answers Smart’s challenge without resort to hypertime.

The morphing block and diachronic personal identity

241

relation of occupation that connects material objects (the guests) with such pluralities or fusions (the hosts).12 Quite unlike the proponent of Leibnizian monads, then, it would seem that the substantivalist proponent of a spacetime capable of augmentation or diminishment does not regard its point locations as modally bound to one another. Instead, different combinations of space-time points coexist at different times. Well . . . if that is so . . . there has been something of a failure of imagination when it comes to dynamic space-time. the morphing block theory Consider what appears to be a growing block that began with a slice that was a duplicate of our own block’s slice from a billion years ago. When the calendars on its outermost surface say October 14, 1066 the block has one volume and when they say April 19, 1775 it has another, such that the first plurality of hyperplanes have been joined by uncountably many others over the 709 year interval that separates the two occasions; reality is growing. Then a surprise . . . new hyperplanes steadily appear at both ends of the block continuing to duplicate the relevant ever-increasing portion of our own block’s history; reality is growing at both ends. Then a reversal . . . the latest hyperplane remains fixed as more and ever earlier ones are tacked on; reality is growing once again, but in the other direction. An alarming development . . . hyperplanes at both ends begin to disappear; reality is shrinking, shrinking, shrinking until there is but one hyperplane in existence. “Then” the world is non-spatio-temporal. If the creation-and-annihilation stories of the growing block, the shrinking block and the disappearing branch theories are live metaphysical options, then why not this block that grows at both ends or shrinks at both ends or grows at one while shrinking at another and then reverses? Better yet – why not a block that acts just like a properly behaved growing block but at some point in its development dutifully adds outermost hyperplanes while recklessly subtracting sub-regions from its inner hyperplanes, i.e. moving (roughly) from block to hourglass shape (or, rather, transitioning between their four-dimensional equivalents)? Why not a block that (perhaps tired of being the only instance of its kind) suddenly annihilates an entire inner hyperplane and thereby brings it about that actuality “now” consists of a pair of disconnected space-times, each of which grows in the old familiar way? 12

On substantivalism (and its opponents) see Earman (1989), Nerlich (1994) and Van Cleve (1987).

242

hud hudson

Perhaps here is why not: note the scare-quoted “then” and “now” in the last two paragraphs. Of course, there is no time such that the world is nonspatio-temporal then, and there is no time such that before that time actuality has but one space-time and afterwards it has two! True enough, but nothing hangs on pretending otherwise. The scare-quoted temporal terms do help to get the relevant ideas across but are dispensable (as will be shown below). Another try – perhaps here is why not: blocks do not cause themselves to undergo such alterations, and nothing else could cause a block to suffer such bizarre augmentation or diminishment either! Well, maybe God could . . . but true enough. All the business about the block’s being a reckless agent or tired of its lonely status was just rhetorical fun, but this observation about causation has not disqualified the growing block or the shrinking block or the disappearing branch theories which also feature uncaused creation and annihilation of space-time regions. Where do all those regions go that make for the shapely new hourglass figure of our morphing block? Presumably to the same graveyard that contains the shrinking block’s spent hyperplanes and the untrodden paths of the disappearing branches. Whatever annihilationist story those theorists told which managed to keep their views amongst the metaphysically possible theories of time simply gets told again. Another try – perhaps here is why not: on the growing block and the shrinking block and the disappearing branch theories we have pretty compelling stories to tell about what it takes for a world to be temporal, what it takes for a time to be the present, what it takes for time to pass, what the direction of time turns out to be, which propositions express the laws of nature, and how to understand the causal relations between the occupants of the block; your unpredictable morpher cannot compete with that! Well, given the ongoing disagreements in the literature, maybe “compelling stories” is a bit overstated . . . but true enough.13 Depending on its symmetries, a morphing block might well turn out to be a spatial-only (i.e. nontemporal) world. Alternatively, a morphing block might turn out to have more than one distinguished hyperplane that deserves the title “present.” Then again, in a morphing block time might reverse its flow or perhaps time might happen to run in more than one direction. My goal here, however, is not to provide and defend answers to such questions. I am content to introduce the morphing block theory and to 13

For a representative discussion of some of these far-ranging topics, see Skow (2007) and a number of the essays in Callender (2011).

The morphing block and diachronic personal identity

243

acknowledge some of the troublesome questions that immediately arise. As indicated above, motivation for taking the morphing block theory seriously can be had straightforwardly by reflecting on common presuppositions of the growing block, the shrinking block and the disappearing branch theories and by maintaining that the primary controversial features of the morphing block theory are already to be found amongst these rivals. Accordingly, to ignore or exclude it while embracing its cousins is to play metaphysical favorites with no adequate justification. In the remainder of this chapter I would like to explore some of the philosophical consequences of taking the morphing block theory seriously. Anyone who wishes to undermine the metaphysical possibility of a morphing block – while retaining that favored status for growing, shrinking or branching blocks and while pleading innocent to the charge of unjustified metaphysical favoritism – can simply consult the three previous paragraphs for an array of targets at which to take aim. curious consequences of morphing Recall our two representatives of the complex view from p. 238: the animalist– endurantist–biological criterion theorist, who fills in the blank in (C1) with his favored way of indicating sameness of human animal, and the minimalist–perdurantist–psychological–criterion theorist, who fills in the blank in (C2) with his favored account of a psychological continuity and connectedness relation between human person-stages. Despite their considerable philosophical differences, these theorists (along with many of their like-minded, complex-theory-of-diachronicpersonal-identity brethren) tend to share two particular commitments. First common commitment: each theorist is responsible for explaining the appropriateness of the adjective “human” in the label “human person.” Usually, this is accomplished by acknowledging some significant relation between the person and a human organism (although, as we have seen, the relation need not be identity). Moreover, human organisms are material objects that have a certain sort of history: that is, nothing that came into existence suddenly and looking for all the world like Achilles at age twenty would be a genuine instance of the kind “human person.” Even though it may exemplify all the cognitive and moral features we ordinarily associate with human persons, to the extent that it had the wrong historical properties, it would not be a human person. Second common commitment: when we ask after the animalist’s account of sameness of human animal or the minimalist’s account of the psychological continuity and connectedness

244

hud hudson

relation, we will find that they often appeal to certain theses about causation and causal dependence. Without resort to the right sort of causal dependence of later features of an enduring person on earlier features of that person (alternatively – of later person-stages of a perduring person on earlier person-stages of that person), there are no compelling ways to divide the cases of genuine persistence of a single person from cases of the seamless spatio-temporal replacement of one person by a distinct but extremely similar one. Consider a formulation of the causal requirement stated in language friendly to the minimalist’s perdurantist metaphysic: (CR) Necessarily, if a human person, P, persists throughout an open temporal interval, T, then for every instant, t, in T, there is an open interval of time, T *, with t as its point-limit such that the sum of P ’s temporal parts that exist during T * is a partial cause of P ’s temporal part at t.14

Finally, we have the ingredients for the problem facing our complex theories of diachronic personal identity. Suppose ours is a morphing block space-time which (up until now) has behaved as would a growing block space-time – monotonously and unidirectionally adding new hyperplanes to reality as the moments tick by, while also faithfully preserving everything once added since the Big Bang. Our friend, Althea, a human person, was born in 1975 and is standing here along with us (in virtue of her most current momentary person-stage) on the very edge of Being: that is, on the hyperplane that is presently on the outermost surface of the block. In five minutes, however, our morphing block will show its true colors and will no longer masquerade as a mere growing block, for in five minutes, a portion of the hyperplanes identified with the 1980s will disappear. In particular, some portion of those hyperplanes that contain all of Althea’s 1980 temporal parts will simply be annihilated. Moreover, in ten minutes, a portion of the hyperplanes that contain all of Althea’s 1970 temporal parts will be annihilated as well. I suppose a number of people would not regard the obliteration of certain portions of the 1980s or the 1970s as a tragedy, but what of our friend Althea? What will become of her five and then ten minutes hence? For starters, we should note that at the five-minute mark our space-time will 14

This principle of immanent causation for human persons is a modified version of a more general principle in Zimmerman (1999), in which “human person” is replaced by “material object.” Whereas I regard the general principle appropriate for most familiar sortal terms, “material object” can apply to fusions of objects with no causal connection whatsoever. On this last point, see Hudson (2006, ch. 5).

The morphing block and diachronic personal identity

245

have a new shape – “a block with an internal cavity” is not quite right, since cavities are themselves located at regions, but it gets the idea across. With that rough and ready observation on shape change, then, let us turn our attentions to Althea. First let us note that the 1980s will still exist; the hyperplanes identified with those times will have a hole in them (so to speak), but otherwise they will be as they were when they occupied the outermost positions on the block. If we could examine the contents of the block from 1975 to the present, we would find a temporally connected object spanning some moment in 1975 to the first moment of 1980 that looks much like a human person from conception to the tender age of five. We would also find a temporally connected object spanning the first moment of 1990 to the present that looks much like a human person from age fifteen to thirtyseven. We would be inclined to declare that these are two temporally extended person slices of one and the same human person (and indeed, were we to make that assertion now – premorph – we would be unproblematically right). But in just a few minutes we would face an obstacle – namely, that there will be no temporally connected object with the appearance of a five- to fifteen-year-old to be found anywhere in the 1980s that has the slightest chance of standing in the slice-of-the-same-person-as relation with our other two objects. If Althea (who will still be standing with us five minutes hence) was born in 1975 (as she will continue to profess), then she will be a temporally gappy person who will suffer from false memories of an alleged ten-year period in her life that simply never took place. Some clarification: since (by hypothesis) the regions that contain Althea’s 1980 temporal parts will be obliterated, all the objects and events they contain (e.g. her first kiss, her middle-school graduation, that scar from her hiking accident) will also be removed from the world’s stage. Significantly, though, nothing at all happens to her temporal slices that occupy the 1990s onward. Accordingly, whereas she can now recall that first kiss almost as vividly as she experienced it, in five minutes she will be mistaken to believe that she had that experience at all or that she attended middle school or that the scar above her left eye is the result of an unfortunate tumble down a cliffside. Is she fated, then, to be temporally gappy and deluded by apparent memories? Two questions there. Whether or not she is temporally gappy will be determined (in part) by whether or not she will still satisfy the causal requirement introduced above. According to (CR) this would require that “the-conception-to-five” slice is the right kind of partial cause of “thefifteen-to-thirty-seven” slice, despite not having temporally intermediate

246

hud hudson

objects to carry the requisite causal signals and despite the presence of many features in the latter object that seem to lack a cause altogether. Whether she will be deluded by apparent memories is a tricky question. Note that on the model on offer what it takes (from the perspective of some time, t) for an event to have happened is that when t is the present moment, the event is located in the block at times earlier than t. Accordingly, as things stand now – premorph – the memory of her first kiss is veridical; that event really happened. And accordingly, five minutes hence – post-cavity – the memory of her first kiss will be merely illusory; it will be true that that event never really happened. A sad state of affairs, but Althea is a strong human person – perhaps she can come to live with it. But even worse things are in store for her. At the ten-minute mark, our space-time will have morphed again, and we can ask after Althea a second time. Again, although a decade’s worth of more hyperplanes will now sport holes (of a sort), the 1970s and 1980s will still exist, and otherwise those times will be as they were when they successively occupied the outermost positions on the block. Undoubtedly, Althea will still report 1975 as her birth year (but she will be wrong). Lots of things will have happened in the 1970s but none of them will have happened to Althea, for she will not have any temporal parts located in any portion of that tenyear span. In fact, it will be true that Althea came into existence in 1990 looking and thinking and feeling much like a fifteen-year-old girl (albeit one who is wildly mistaken about her past). She will not be temporally gappy, but she will not be human either – and this, despite the fact that nothing will have happened to her. Why will she be non-human and how can we say nothing will have happened to her if she becomes non-human? Two questions there: as we noted above, our persistence theorists share a commitment to explaining the appropriateness of the adjective “human” in the label “human person,” and this is accomplished by somehow relating a human person to a human organism. But human organisms are material objects that have a certain sort of history. Even though a thing may exemplify all the cognitive and moral features we ordinarily associate with human persons, to the extent that it had the wrong historical properties, it would not be a human person. Althea will have the wrong historical properties – the wrong origins: that is why she will be non-human. Yet nothing will have happened to her. It is not as if she will undergo some change or transition from human to non-human (where “change” is here read as qualitative difference between her temporal parts). For reasons exactly similar to those which explain why five minutes hence she will be

The morphing block and diachronic personal identity

247

deluded by apparent memories, ten minutes hence it will be true that she has never had any pre-1990 temporal parts, for when the hyperplane identified with the time ten minutes hence is present, that is all of Althea the block will contain. And the features had by the block precisely when that moment is present completely determine what is true of the past when that moment is present. These problems seem to be rooted in a simple observation. Just which temporal parameters (and parts) a thing has are sensitive to the machinations of a morphing block, and the consequences of such sensitivity are severe.

severity and skepticism Let us return to our discussion of Althea’s fate five minutes hence, for it is that scenario in which we most clearly face the difficulties for our complex theories of diachronic personal identity. To recap: if growing or shrinking or branching blocks are taken with all metaphysical seriousness, then a morphing block is worth considering seriously as well. But such consideration has consequences for judgments of diachronic personal identity. In our morphing block scenario, whether it will be true that Althea was born in 1975 depends on whether she will satisfy the causal requirement (CR) – or at least that is what our minimalist–perdurantist–psychological–criterion theorist (from p. 238) contends. Our animalist–endurantist–biological criterion theorist (also introduced on p. 238) maintains that it depends instead on whether there is the right sort of causal dependence of the (enduring and wholly present) woman who will be standing next to us five minutes hence upon a certain (enduring and wholly present) little girl in 1979. But whether it ultimately relates to mere slices or whole persons, the shared commitment to causal dependence remains the best strategy for these theorists when it comes to telling apart cases of genuine persistence from cases of seamless replacement by a doppelgänger. Taking the morphing block theory seriously, however, may generate rather unpalatable skeptical scenarios. Anyone who accepts as an epistemic possibility that his is a morphing block world will shortly find himself in the uncomfortable position of being in serious doubt about whether he was in fact born in the year he currently believes is his birth year. Moreover, this is no “maybe-the-world-was-created-five-minutes-ago” hypothesis, for – and this is the truly surprising bit – such doubt can reasonably arise even if we realize that he currently knows perfectly well the year of his birth. That is, it

248

hud hudson

will put him in the position of endorsing an analogue of this speech offered by one of Althea’s worldmates: Even if I assume that I currently know that Althea exists here and now and that I also know that she is numerically identical to that little girl who existed there and then, five minutes hence I may not possess that knowledge (despite the fact that nothing of significance will have happened to me during the interval and despite the fact that “Althea” and “that little girl” will continue to successfully refer to individuals in space-time).

A morphing block scenario would, of course, generate a large number of such skeptical theses, but this seems a particularly impressive one. To the extent that we think such a skeptical doubt in Althea’s worldmates would be untenable (or crazy) we face a choice: either reject the metaphysical possibility of the morphing block that gives rise to the causal worry or else reject the theories of personal identity that share the causal dependence requirement. Should the metaphysical possibility of the morphing block seem secured by the metaphysical possibility of the growing or the shrinking or the branching blocks (and by their relevant similarities), we thereby acquire some motivation for a simple view of diachronic personal identity. A proponent of a simple view need not resort to the causal dependence claims articulated in (CR), since the little girl may simply be the grown woman, and that is all there is to say – there simply are no further, deeper facts that are relevant to judgments of identity. Consequently, a proponent of a simple view can countenance a wider range of theories about the nature of time without facing bizarre threats to the identification of herself and her friends with certain individuals in the past. That is a modest but interesting advantage in the personal identity wars.15 15

Thanks for delightful comments, conversation and criticism to Ross Cameron, Peter Forrest, Meghan Sullivan and Ryan Wasserman.

References

Adams, Robert 1979. “Primitive Thisness and Primitive Identity,” Journal of Philosophy 76: 5–26. 1987. “Divine Necessity,” in Robert Adams, The Virtues of Faith and Other Essays in Philosophical Theology. New York: Oxford University Press, pp. 209–20. Armstrong, David 1980. “Identity Through Time,” in Peter van Inwagen (ed.), Time and Cause: Essays Presented to Richard Taylor. Dordrecht: Reidel, pp. 67–78. Baker, Lynne Rudder 2000. Persons and Bodies: A Constitution View. Cambridge University Press. 2005. “Death and the Afterlife,” in William J. Wainwright (ed.), The Oxford Handbook of Philosophy of Religion. Oxford University Press, pp. 366–91. 2007a. “First-Person Externalism,” The Modern Schoolman 84: 155–70. 2007b. “Naturalism and the First-Person Perspective,” in Georg Gasser (ed.), How Successful is Naturalism? Heusenstamm: Ontos Verlag, pp. 203–27. 2007c. “Social Externalism and the First-Person Perspective,” Erkenntnis 67: 287–300. 2007d. The Metaphysics of Everyday Life. Cambridge University Press. 2008. “A Metaphysics of Ordinary Things and Why We Need It,” Philosophy 83: 5–24. 2009. “Identity Across Time: A Defense of Three-Dimensionalism,” in Ludger Honnefelder, Edmund Runggaldier, and Benedikt Schick (eds.), Unity and Time in Metaphysics. Berlin: Walter de Gruyter, pp. 1–14. Barnett, David 2009. “Is Vagueness Sui Generis?,” Australasian Journal of Philosophy 87: 5–34. 2010. “You Are Simple,” in Robert C. Koons and George Bealer (eds.), The Waning of Materialism. Oxford University Press, pp. 161–74. Bealer, George 1996. “On the Possibility of Philosophical Knowledge,” in James Tomberlin (ed.), Philosophical Perspectives, vol. x. Cambridge: Blackwell, pp. 1–33. Benovski, Jiri 2010. “Relational and Substantival Ontologies, and the Nature and the Role of Primitives in Ontological Theories,” Erkenntnis 73: 101–21. Black, Max 1952. “The Identity of Indiscernibles,” Mind 61: 153–64. 249

250

References

Boethius, Anicius Manlius Severinus 1988. Contra Eutychen et Nestorium, in Theological Tracts, ed. M. Elsässer. Hamburg: Meiner. Bolinger, D. 1972. Degree Words. The Hague: Mouton. Bourne, Craig 2002. “When Am I?,” Australasian Journal of Philosophy 80: 359–71. Braddon-Mitchell, David 2004. “How Do We Know It Is Now Now?,” Analysis 64: 199–203. Broad, C. D. 1923. Scientific Thought. London: Kegan Paul. Butler, Joseph 1975 [1736]. “Of Personal Identity,” in John Perry (ed.), Personal Identity. Berkeley and Los Angeles: University of California Press, pp. 99–106. Callender, Craig (ed.) 2011. The Oxford Handbook of Philosophy of Time. Oxford University Press. Chalmers, David 1996. The Conscious Mind. Oxford University Press. Chisholm, Roderick M. 1969. “The Loose and Popular and the Strict and Philosophical Senses of Identity,” in Norman S. Care and Robert H. Grimm (eds.), Perception and Personal Identity. Cleveland, OH: Case Western Reserve University, pp. 82–106. Chisholm, Roderick M. 1970a. “Identity Through Time,” in Howard E. Kiefer and Milton K. Munitz (eds.), Language, Belief, and Metaphysics. Albany: State University of New York Press, pp. 163–82. 1970b. “Reply to Strawson’s Comments,” in Howard E. Kiefer and Milton K. Munitz (eds.), Language, Belief, and Metaphysics. Albany: State University of New York Press, pp. 187–9. 1976. Person and Object: A Metaphysical Study. La Salle, IL: Open Court. 1986. “Self-Profile,” in Radu J. Bogdan (ed.), Roderick M. Chisholm. Dordrecht: Reidel, pp. 3–77. 1989. On Metaphysics. Minneapolis: University of Minnesota Press. Cross, Richard 2010 [2003]. “Medieval Theories of Haecceity,” in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall edition), available at: http://plato.stanford.edu/archives/fall2010/entries/medieval-haecceity/. Davidson, Donald (ed.) 1980 [1969]. “The Individuation of Events,” in Donald Davidson, Essays on Actions and Events. Oxford: Clarendon Press, 163−80. Davis, Stephen T. 2010. “Resurrection, Personal Identity, and the Will of God,” in Georg Gasser (ed.), Personal Identity and the Resurrection. Farnham: Ashgate, pp. 19–31. Descartes, René 1984 [1641]. “Meditations on First Philosophy,” in John Cottingham, Robert Stoothoff and Dugald Murdoch (eds.), The Philosophical Writings of Descartes. Cambridge University Press. Earman, John 1989. World Enough and Space-time. Cambridge, MA: MIT Press. Evans, Gareth 1978. “Can There Be Vague Objects?,” Analysis 38: 208. Field, Hartry 2003. “No Fact of the Matter,” Australasian Journal of Philosophy 81: 457–80. Fine, Kit 1994. “Ontological Dependence,” Proceedings of the Aristotelian Society 95: 269–90. 2001. “The Question of Realism,” Philosopher’s Imprint 1: 1–30.

References

251

Forrest, Peter 2004. “The Real but Dead Past: A Reply to Braddon-Mitchell,” Analysis 64: 358–62. Foster, John 1991. The Immaterial Self: A Defence of the Cartesian Conception of the Mind. London: Routledge. Frege, Gottlob 1953 [1884]. The Foundations of Arithmetic, trans. J. L. Austin. Oxford: Blackwell. 1960 [1892]. “On Sense and Reference,” in Peter T. Geach and Max Black (eds.), Translations from the Philosophical Writings of Gottlob Frege, 2nd edn. Oxford: Blackwell, pp. 56−78. French, Steven 2006. “Identity and Individuality in Quantum Theory,” in Edward N. Zalta (ed.), Stanford Encyclopedia of Philosophy (Spring edition), available at: http://plato.stanford.edu/archives/spr2006/entries/qt-idind/. Gallois, André 1998. Occasions of Identity: A Study in the Metaphysics of Persistence, Change, and Sameness. Oxford University Press. Geach, Peter 1976. “Two Kinds of Intentionality?,” The Monist 59: 306–20. Gilbert of Poitiers 1966. The Commentaries on Boethius, ed. N. M. Häring. Toronto: Pontifical Institute of Mediaeval Studies. Gracia, Jorge J. E. 1994. Individuation in Scholasticism: The Middle Ages and the Counterreformation. New York: State University of New York Press. Hasker, William 2001. The Emergent Self. Ithaca, NY: Cornell University Press. Hauser, Marc 2009. “The Mind,” Scientific American (September edition): 44–9. Hawthorne, John 2006. “Epistemicism and Semantic Plasticity,” in John Hawthorne (ed.), Metaphysical Essays. Oxford University Press, pp. 185–211. Heathwood, Chris 2005. “The Real Price of the Dead Past: A Reply to Forrest and Braddon-Mitchell,” Analysis 65: 249–51. Hick, John 1990. Philosophy of Religion, 4th edn. Englewood Cliffs, NJ: Prentice Hall. Hirsch, Eli 1982. The Concept of Identity. New York: Oxford University Press. Hossack, Keith 2006. “Vagueness and Personal Identity,” in Fraser MacBride (ed.), Identity and Modality. Oxford University Press, pp. 221–41. Howard-Snyder, Frances 2005. “It’s the Thought That Counts,” Utilitas 17: 265–81. Hudson, Hud 2001. A Materialist Metaphysics of the Human Person. Ithaca, NY: Cornell University Press. 2006. The Metaphysics of Hyperspace. Oxford University Press. 2007. “I Am Not an Animal!,” in Peter van Inwagen and Dean Zimmerman (eds.), Persons: Human and Divine. Oxford University Press, pp. 216–34. 2012. “Transreceptacle Identity,” in Shieva Kleinschmidt (ed.), Mereology and Location. Oxford University Press, in press. Kanzian, Christian 2009. Ding – Substanz – Person. Heusenstamm: Ontos. Kripke, Saul 1981 [1972]. Naming and Necessity. Oxford: Blackwell. Leonard, Henry S. and Goodman, Nelson, 1940. “The Calculus of Individuals and its Uses,” Journal of Symbolic Logic 5: 45–55. Lewis, David 1976. “Survival and Identity,” in Amelie Rorty (ed.), The Identities of Persons. Berkeley: University of California Press, pp. 17–40. 1979. “Attitudes De Dicto and De Re,” Philosophical Review 88: 513–43.

252

References

1980. “A Subjectivist’s Guide to Objective Chance,” in Richard Jeffrey (ed.), Studies in Inductive Logic and Probability, vol. ii. Berkeley: University of California Press, pp. 263–93. 1981. Philosophical Papers, vol. i. New York: Oxford University Press. 1983. “New Work for a Theory of Universals,” Australasian Journal of Philosophy 61: 343–77. 1984. “Putnam’s Paradox,” Australasian Journal of Philosophy 62: 221–36. 1986. On the Plurality of Worlds. Oxford: Blackwell. 1997. “New Work for a Theory of Universals,” in D. H. Mellor and Alex Oliver (eds.), Properties. Oxford University Press, pp. 173–87. Locke, John 1975 [1690]. An Essay Concerning Human Understanding, ed. P. H. Nidditch, Oxford: Clarendon Press. Lowe, E. J. 1988. “Substance, Identity and Time,” Proceedings of the Aristotelian Society 62: 61–100. 1989a. “Impredicative Identity Criteria and Davidson’s Criterion of Event Identity,” Analysis 49: 178–81. 1989b. Kinds of Being. Oxford: Blackwell. 1989c. “What is a Criterion of Identity?,” Philosophical Quarterly 39: 1–21. 1994. “Primitive Substances,” Philosophy and Phenomenological Research 54: 531–52. 1995. Locke on Human Understanding. London and New York: Routledge. 1996. Subjects of Experience. Cambridge University Press. 1997. “Objects and Criteria of Identity,” in Bob Hale and Crispin Wright (eds.), A Companion to the Philosophy of Language. Oxford: Blackwell, pp. 613–33. 2000. An Introduction to the Philosophy of Mind. Cambridge University Press. 2001. “Identity, Composition and the Simplicity of the Self,” in Kevin Corcoran (ed.), Soul, Body, and Survival: Essays on the Metaphysics of Human Persons. Ithaca, NY: Cornell University Press, pp. 139–58. 2003. “Substance Causation, Persons, and Free Will,” in Christian Kanzian, Josef Quitterer and Edmund Runggaldier (eds.), Persons. Vienna: OeBV-hpt, pp. 76–88. 2009. More Kinds of Being: A Further Study of Individuation, Identity and the Logic of Sortal Terms. Malden, MA: Blackwell. Lund, David 2005. The Conscious Self. Amherst, MA: Humanity Books. Lutz-Bachmann, Matthias 1983. “Person und Natur in den Opuscula Sacra des A. M. S. Boethius,” Theologie und Philosophie 58: 48–70. Madell, Geoffrey 1981. The Identity of the Self. Edinburgh University Press. Markosian, Ned 1993. “How Fast Does Time Pass?,” Philosophy and Phenomenological Research 53: 829–44. Mavrodes, George 1977. “The Life Everlasting and the Bodily Criterion of Identity,” Noûs 11: 27–39. Merricks, Trenton 1997. “Fission and Personal Identity over Time,” Philosophical Studies 88: 163–86. 1998a. “Against the Doctrine of Microphysical Supervenience,” Mind 107: 59–71.

References

253

1998b. “There Are No Criteria of Identity over Time,” Noûs 32: 106–24. 1999. “Endurance, Psychological Continuity, and the Importance of Personal Identity,” Philosophy and Phenomenological Research 59: 983–97. 2001a. “How to Live Forever without Saving Your Soul,” in Kevin Corcoran (ed.), Soul, Body, and Survival: Essays on the Metaphysics of Human Persons. Ithaca, NY: Cornell University Press, pp. 183–200. 2001b. “Realism About Personal Identity over Time,” in “Metaphysics,” special issue of Philosophical Perspectives 15: 173–87. 2003. “Maximality and Consciousness,” Philosophy and Phenomenological Research 66: 150–8. 2006. “Goodbye Growing Block,” Oxford Studies in Metaphysics 2: 103–10. Mithen, Steve 2004. “Review Symposium of Andy Clark’s Natural-Born Cyborgs,” Metascience 13: 163–9. Nagel, Thomas 1965. “Physicalism,” Philosophical Review 74: 339–56. 1986. The View from Nowhere. New York: Oxford University Press. Nerlich, Graham 1994. What Spacetime Explains. Cambridge University Press. Nida-Rümelin, Martine 2006. Der Blick von Innen. Frankfurt am Main: Suhrkamp. 2010. “An Argument from Transtemporal Identity for Subject−Body Dualism,” in Robert C. Koons and George Bealer (eds.), The Waning of Materialism. Oxford University Press, pp. 190–211. 2012. “The Conceptual Origin of Subject−Body Dualism,” in Annalisa Colliva (ed.), Self and Self-Knowledge. Oxford University Press, pp. 39−73. Noonan, Harold W. 1989. Personal Identity. London: Routledge. 2003. Personal Identity, 2nd edn. London: Routledge. 2011. “The Complex and Simple Views of Personal Identity,” Analysis 71: 72–7. Nozick, Robert 1981. Philosophical Explanations. Oxford University Press. O’Connor, Timothy and Jacobs, Jonathan D. 2010. “Emergent Individuals and the Resurrection,” European Journal for Philosophy of Religion 2: 69–88. Oderberg, David S. 1993. The Metaphysics of Identity over Time. New York: St. Martin’s Press. Olson, Eric T. 1995. “Human People or Human Animals?,” Philosophical Studies 80 (2): 159−81. 1997a. “Relativism and Persistence,” Philosophical Studies 88: 141–62. 1997b. The Human Animal: Personal Identity Without Psychology. New York: Oxford University Press. 2002. “What Does Functionalism Tell Us about Personal Identity?,” Noûs 36: 682–98. 2007. What Are We? A Study in Personal Ontology. Oxford University Press. 2010. “Personal Identity,” in Edward N. Zalta (ed.), Stanford Encyclopedia of Philosophy, available at: http://plato.stanford.edu/entries/identity-personal/. Parfit, Derek 1982. “Personal Identity and Rationality,” Synthese 53: 227–41. 1984. Reasons and Persons. Oxford University Press. Perry, John 1972. “Can the Self Divide?,” Journal of Philosophy 69: 463–88. 1979. “The Problem of the Essential Indexical,” Noûs 3: 3–21.

254

References

2008 [1975]. “The Problem of Personal Identity,” in John Perry (ed.), Personal Identity. Berkeley: University of California Press, pp. 3–30. Pinker, Steven 1977. How the Mind Works. New York: W. W. Norton. Putnam, Hilary 1975. “The Meaning of ‘Meaning’,” in Hilary Putnam (ed.), Mind, Language and Reality. Philosophical Papers, vol. ii. Cambridge University Press, pp. 215–71. Quinton, Anthony 1962. “Spaces and Times,” Philosophy 37: 130–47. Reid, Thomas 1975 [1785]. “Of Mr. Locke’s Account of Personal Identity,” in John Perry (ed.), Personal Identity, 1st edn. Berkeley and Los Angeles: University of California Press, pp. 113–18. 2002 [1785]. Essays on the Intellectual Powers of Man, ed. D. Brookes. University Park: Pennsylvania State University Press. Rorty, Amelie (ed.) 1976. The Identities of Persons. Berkeley: University of California Press. Schaffer, Jonathan 2009. “On What Grounds What,” in David Chalmers, David Manley and Ryan Wasserman (eds.), Metametaphysics. Oxford University Press, pp. 347–83. Shoemaker, Sydney 1963. Self-Knowledge and Self-Identity. Ithaca, NY: Cornell University Press. 1970. “Persons and their Pasts,” American Philosophical Quarterly 7: 269–85. 1979. “Identity, Properties, and Causality,” Midwest Studies in Philosophy 4: 321–42; reprinted in his Identity, Cause, and Mind, expanded edn. Oxford: Clarendon Press, 2003. 1984. “Personal Identity: A Materialist’s Account,” in Sydney Shoemaker, and Richard Swinburne, Personal Identity. Oxford: Blackwell, pp. 67–132. 1985. “Critical Notice of Reasons and Persons,” Mind 94: 443–53; reprinted as “Parfit on Identity,” in J. Dancy (ed.), Reading Parfit, Oxford: Blackwell, 1997. 1994. “Introspection and the Self,” in Quassim Cassam (ed.), Self-Knowledge. Oxford University Press, pp. 118–39. 1999. “Self, Body, and Coincidence,” Proceedings of the Aristotelian Society, supplementary vol. 73: 287–306. 2007. Physical Realization. Oxford University Press. Shoemaker, Sydney and Swinburne, Richard 1984. Personal Identity. Oxford: Blackwell. Sider, Theodore 1996. “Naturalness and Arbitrariness,” Philosophical Studies 81: 283–301. 2000. “Criteria of Personal Identity and the Limits of Conceptual Analysis,” Philosophical Perspectives 15: 189–209. 2001. Four-Dimensionalism: An Ontology of Persistence and Time. Oxford University Press. 2003. “Maximality and Microphysical Supervenience,” Philosophy and Phenomenological Research 66: 139–49. 2012. Writing the Book of the World. Oxford University Press. Simons, Peter 2006. “Real Wholes, Real Parts: Mereology without Algebra,” Journal of Philosophy 103: 597–613.

References

255

Singer, Peter 1978. Practical Ethics. Cambridge University Press. Skow, Bradford 2007. “What Makes Time Different From Space?,” Noûs 41: 227–52. Slote, Michael 1985. Common-Sense Morality and Consequentialism. London: Routledge. Smart, J. J. C. 1949. “The River of Time,” Mind 58: 483–94. 1998. “The Space-Time World: An Excerpt from Philosophy and Scientific Realism,” in Peter van Inwagen and Dean Zimmerman (eds.), Metaphysics: The Big Questions. Oxford: Blackwell, pp. 94–101. Sorenson, Roy 2001. Vagueness and Contradiction. Oxford University Press. Strawson, Peter F. 1959. Individuals: An Essay in Descriptive Metaphysics. London: Methuen. Suarez-Nani, Tiziana 2009. “Singularité et individualité selon Pierre Auriol,” in Stephen F. Brown, Thomas Dewender and Theo Kobusch (eds.), Philosophical Debates at Paris in the Early Fourteenth Century. Leiden: Brill, pp. 339–57. Swinburne, Richard 1973–4. “Personal Identity,” Proceedings of the Aristotelian Society 74: 231–47. 1984. “Personal Identity: The Dualist Theory,” in Sydney Shoemaker and Richard Swinburne, Personal Identity. Oxford: Blackwell, pp. 3–66. 1986. The Evolution of the Soul. Oxford: Clarendon Press. 1995. “Thisness,” Australasian Journal of Philosophy 73: 389–400. 1997. The Evolution of the Soul, rev. edn. Oxford University Press. 2007. “From Mental/Physical Identity to Substance Dualism,” in Peter van Inwagen and Dean Zimmerman (eds.), Persons: Human and Divine. Oxford University Press, pp. 142–65. 2010. “In Defence of Logical Nominalism: Reply to Leftow,” Religious Studies 46: 311–30. Thomson, Judith Jarvis 2008. “Persons and their Bodies,” in Theodore Sider, John Hawthorne and Dean Zimmerman (eds.), Contemporary Debates in Metaphysics. Oxford: Blackwell, pp. 155–76. Tooley, Michael 1997. Time, Tense and Causation. Oxford University Press. Unger, P. 1990. Identity, Consciousness, and Value. Oxford University Press. Van Cleve, James 1987. “Right, Left and the Fourth Dimension,” Philosophical Review 96: 33–68. van Inwagen, Peter 1990a. “Four-Dimensional Objects,” Noûs 24: 245–55. 1990b. Material Beings. Ithaca, NY: Cornell University Press. 2006. “Can Mereological Sums Change their Parts?,” Journal of Philosophy 103: 614–30. Wasserman, Ryan 2004a. “Framing the Debate over Persistence,” Metaphysica 5: 67–80. 2004b. “Indeterminacy, Ignorance and the Possibility of Parity,” Philosophical Perspectives 14: 391–403. Wiggins, David 1980. Sameness and Substance. Oxford: Blackwell. 2001. Sameness and Substance Renewed. Cambridge University Press.

256

References

Wilkes, Kathleen 1988. Real People: Personal Identity without Thought Experiments. Oxford University Press. Williams, Bernard 1970. “The Self and the Future,” Philosophical Review 79: 161–80. Williamson, Timothy 1994. Vagueness. London: Routledge. Wright, Crispin 1987. “Further Reflections on the Sorites Paradox,” Philosophical Topics 15: 227–90. Zimmerman, Dean 1991. “Two Cartesian Arguments for the Simplicity of the Soul,” American Philosophical Quarterly 28: 217–26. 1997. “Immanent Causation,” in “Mind, Causation, and World,” special issue of Philosophical Perspectives 11: 433–71. 1998. “Criteria of Identity and the ‘Identity Mystics’,” Erkenntnis 48: 281–301. 1999. “The Compatibility of Materialism and Survival: The ‘Falling Elevator’ Model,” Faith and Philosophy 16: 194–212. 2003. “Material People,” in Michael Loux and Dean Zimmerman (eds.), The Oxford Handbook of Metaphysics. Oxford University Press, pp. 491–526. 2004. “Should a Christian Be a Mind-Body Dualist?,” in Michael L. Peterson and Raymond J. VanArragon (eds.), Contemporary Debates in Philosophy of Religion. Malden, MA: Blackwell, pp. 315–27. 2009. “Properties, Minds, and Bodies: An Examination of Sydney Shoemaker’s Metaphysics,” Philosophy and Phenomenological Research 78: 673–738. 2010. “From Property Dualism to Substance Dualism,” Proceedings of the Aristotelian Society, supplementary vol. 84: 119–50.

Index

agency 174 amnesia 6, 28 animal human 4–5, 88, 89, 90, 92, 94, 238, 243 non-human 190 animalism 4–6, 7, 91, 131, 145–6 artifacts 108 Baker, Lynne Rudder 13, 59, 179, 181, 182, 184, 191 Barnett, David 17, 73 Benovski, Jiri 193 block theory of time growing 238–40, 241, 242 morphing 236, 241, 242–3, 244, 247–8 shrinking 238–40, 242 Boethius 195 boundary between simple and complex view 44, 50, 55 brain damage 5, 48 identity 151, 152 transplantation of 6, 7, 88, 95, 151 Brown and Brownson 6, 88, 90, 95, 100 Butler, Joseph 60, 85, 123, 148, 158, 207, 218 Cartesian Ego 210, 211 Cartesianism see dualism causation holistic 227–9 immanent 209, 231–4 Chisholm, Roderick 15, 54, 60, 63, 64, 85, 158, 180, 185, 187, 207–11, 218–21 circularity of personal identity 125, 134–6, 143, 148, 149, 151–5 co-consciousness 86, 87 coincident entities 98, 100, 101 complex view 3–10, 15, 44, 46–7, 49, 50, 51, 52, 54, 56–8, 60, 63, 64–6, 71, 73, 82, 85–8, 95, 108, 123, 124, 127, 132–3, 134, 135, 136, 153, 179, 180, 192–3, 204, 206–7, 211, 212, 236–8, 243 composition 17, 35, 186, 188, 201, 203, 205

compound of body and soul 51, 52 conceivability 164, 234, 235 conditions of identity see identity connectedness physical 65 psychological 63, 65, 238, 243 constitution relation 181, 187, 191 constitution view 191 constitutional basis 158–65, 173 continuer 94 psychological 91–3 continuity biological 90, 135 bodily/material 3, 8, 12, 13, 44, 45, 51, 52, 55, 56, 58, 65, 87, 90, 108, 114, 115, 120, 179, 206, 207, 220, 221 of brain 106–8, 119, 120 of form 203 of life 5, 203 of memory 106–8, 114, 115, 132 of stuff 56, 87 psychological 3, 6–9, 12, 13, 44, 51, 53, 55–6, 60, 63, 65, 86, 87, 89–92, 95, 108, 114, 132, 179, 206, 207, 218, 220, 221, 238, 243 spatio-temporal 127, 128, 129, 130, 200 counterparts 125 perfect 159, 161–73, 174, 175 criteria of identity see identity Davidson, Donald 143, 149, 150, 152 Descartes, René 144 dualism Cartesian 49–50, 52, 55–6, 57, 58 mind–body 123 duplication problem 184, 185 embryo 45 emergentism 190, 204 and materialism see materialism endurantism 237, 238 epistemicism 71–4

257

258

Index

eternalism 238, 239 Evans, Gareth 72, 89 experiencing subjects 175 first-person memory 130, 131, 176 first-person perspective 10, 15, 16, 144, 180–2, 184, 185, 188–91 first-person reference 89, 90, 91, 92, 182 fission and fuzziness 222, 223, 224, 227, 233, 234, 235 fission argument 222, 224, 229 fission case 11, 15, 16, 148, 224 fission scenario see fission case form individual 192, 200, 202–3, 205 Foster, John 58, 180, 183 four-dimensionalism 11, 13, 53, 133 Frege, Gottlob 140, 141, 142, 155 French, Steven 119 fusion 76, 148, 171, 185, 225–7, 240, 241 fuzziness 221, 223–4 Gallois, André 121 Geach, Peter 207 Gilbert of Poitiers 194 Goodman, Nelson 185 graduality 8, 13, 15 Hasker, William 204–5 Hauser, Marc 190 Hick, John 59 Hirsch, Eli 201 Howard-Snyder, Frances 78 human vegetable 7 Hume, David 56, 57, 85, 193, 206 identity anti-criterionalism of 49–52 biological criterion of 198, 238 constitutive criterion of 47, 131, 133, 153, 154 informative criterion of 3, 14, 17, 134, 137, 151, 152, 153, 156 Locke’s criterion of 89–93, 135, 146–54, 155 non-circular criterion of see identity, informative criterion of one-level criterion of 141, 142, 147, 151, 152, 155 psychological criterion of 16–17, 198, 238 specific criterion of 61 transitivity of 11, 13, 127, 140 two-level criterion of 87, 141–2, 151–2, 155, 156 immaterial substance 144, 145, 180, 183, 191, 208 immaterialism 180

impersonal facts 63, 221, 224 indeterminacy epistemic 97 ontic 97 individuation of conscious states 132, 153 Jacobs, Jonathan D. 227, 229 Kant, Immanuel 58 Kripke, Saul 116, 118 laws of nature 31, 39, 242 Leibniz’s law 138, 140 Leonard, Henry S. 185 Lewis, David 8, 54, 77, 82, 85, 87, 89, 92, 94, 180, 206 living being 4–7, 45, 52, 88, 135, 181, 182, 189, 190, 194, 196, 198, 201, 202, 217, 233, 238 location 241 Locke, John 52, 132, 137, 139, 143–6, 149, 151, 152, 153, 154, 193, 198, 206 Lowe, E. J. 3, 14, 17, 132–4, 141, 143, 144, 146, 150, 152, 153, 156, 158 Lund, David 10 Lutz-Bachmann, Matthias 196 Madell, Geoffrey 15 materialism 237 and dualism 206 emergent 220, 223–7, 229, 233, 234 mereology 185–7 Merricks, Trenton 47, 50, 60, 226–7, 229 Mithen, Steve 189 Nagel, Thomas 4 neo-Lockean account of personal identity see identity Nida-Rümelin, Martine 16 Non-reductionist account of personal identity 207–8, 210 Noonan, Harold 8, 10, 13, 56–7, 179, 183, 184 Nozick, Robert 10 occupation see location O’Connor, Timothy 227, 229 Oderberg, David S. 201 Olson, Eric 1, 4, 89, 145, 236 only x and y principle 10 organism see living being human 2, 4, 6, 145, 171, 181, 189, 196, 202, 227, 237, 243, 246 Parfit, Derek 6, 9, 11, 63–4, 70, 73, 85, 89, 96, 148, 179, 180, 206–9, 213, 216, 233, 234

Index parts essential 120 material 202 mereological 185–8, 191 temporal 2, 11–13, 14, 53, 54, 75–6, 133, 209, 225–7, 230–4, 237, 244–7 part–whole relation 185 Perry, John 12, 13, 93, 206 personal identity see identity personality change 6, 32, 38, 39, 40, 42 personhood 2, 4, 46, 54, 84–9, 97, 144, 146, 180, 188, 191, 194, 195, 196, 197, 201, 226 person-stages 7, 8, 54, 76, 87, 133, 134 Pinker, Steven 190 possibility logical 108, 112, 115, 120 metaphysical 15, 109, 115, 164, 166, 243, 248 presentism 238, 239 primary kind 181 properties emergent 189, 220, 229 essential 122, 234 intrinsic 226, 228 mental 15, 16, 49, 53, 54, 94, 105, 107, 108, 120, 121, 122, 131, 135, 159, 172, 208, 209, 210, 216, 229 physical 16, 105, 106, 108, 120, 121, 122, 128, 131, 135 sortal 127, 128 thin and thick 234 Putnam, Hilary 117 quasi-memory 132, 148, 216 reductionist account of personal identity 63, 136, 206–7, 211, 213–14, 216, 231, 233 Reid, Thomas 60, 63, 68, 85, 114, 123, 147, 158, 183, 207, 218 relations causal 50, 59, 119, 124–7, 129, 131, 210, 217, 226, 227, 228, 231, 232, 234, 242 spatial 37, 127, 213, 214 temporal 121, 127, 213, 239 replacement of brain 115 of parts 115 of person 244 replacement machine 65, 66, 70, 73, 74, 75 rigid designators 116, 117 Rorty, Amelie 2

259

sameness of brain 146 of person 211 of soul 50, 55 Shoemaker, Sydney 6, 7, 10, 53, 54, 60, 85, 86, 88, 89, 91, 92, 93–5, 96, 100, 119, 132, 134, 153–5, 180, 207, 209, 234 Sider, Theodore 11, 12, 77, 226 similarity 55, 65, 130, 141, 212, 218 Simons, Peter 187 simple view 1, 3, 14–15, 17, 44–6, 47, 48, 49, 50–2, 56–8, 60, 62, 85–6, 88–9, 105, 107, 108, 120, 123, 124–7, 128, 129, 130, 136, 158, 179–80, 182–3, 184, 185, 188, 191, 193, 204, 205, 206, 207, 208, 218, 248 simples 188, 229, 230, 231, 232, 233 Singer, Peter 196, 197 skepticism 14, 130, 153, 247 sortal term 85, 192, 194–6, 198, 199–200 Strawson, Peter F. 149, 153 substance dualism 51, 120, 122, 196, 220 supervenience 212–13, 214–16, 219, 220, 221, 223, 226, 230–4 Swinburne, Richard 15, 49, 50, 51, 52, 54, 55, 56, 64, 85–7, 88, 96, 152, 158, 180, 207, 218–21 symmetry 16, 100, 140, 222 teletransportation 27–8, 32, 38, 39, 40, 41–2 temporal gap 238 theories of time 238, 240, 242 thinking animal problem 4–5, 89, 94 Tibbles and Tib 99 too many thinkers problem see thinking animal problem Unger, Peter 52, 54 unity of consciousness 204–5 unity relation 124, 126, 127, 129, 156 vague objects 96–101 vagueness 54, 71, 72, 76, 95, 98, 100, 111 van Inwagen, Peter 52, 54, 88, 187 Wiggins, David 46, 148 Wilkes, Kathleen 11 Williams, Bernard 10 Williamson, Timothy 71–2, 74, 75 Wright, Crispin 72 Zimmerman, Dean 50, 59, 209, 234